Annotation of imach/src/imach.c, revision 1.262
1.262 ! brouard 1: /* $Id: imach.c,v 1.261 2017/04/05 10:14:09 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.262 ! brouard 4: Revision 1.261 2017/04/05 10:14:09 brouard
! 5: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1
! 6:
1.261 brouard 7: Revision 1.260 2017/04/04 17:46:59 brouard
8: Summary: Gnuplot indexations fixed (humm)
9:
1.260 brouard 10: Revision 1.259 2017/04/04 13:01:16 brouard
11: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3
12:
1.259 brouard 13: Revision 1.258 2017/04/03 10:17:47 brouard
14: Summary: Version 0.99r12
15:
16: Some cleanings, conformed with updated documentation.
17:
1.258 brouard 18: Revision 1.257 2017/03/29 16:53:30 brouard
19: Summary: Temp
20:
1.257 brouard 21: Revision 1.256 2017/03/27 05:50:23 brouard
22: Summary: Temporary
23:
1.256 brouard 24: Revision 1.255 2017/03/08 16:02:28 brouard
25: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed
26:
1.255 brouard 27: Revision 1.254 2017/03/08 07:13:00 brouard
28: Summary: Fixing data parameter line
29:
1.254 brouard 30: Revision 1.253 2016/12/15 11:59:41 brouard
31: Summary: 0.99 in progress
32:
1.253 brouard 33: Revision 1.252 2016/09/15 21:15:37 brouard
34: *** empty log message ***
35:
1.252 brouard 36: Revision 1.251 2016/09/15 15:01:13 brouard
37: Summary: not working
38:
1.251 brouard 39: Revision 1.250 2016/09/08 16:07:27 brouard
40: Summary: continue
41:
1.250 brouard 42: Revision 1.249 2016/09/07 17:14:18 brouard
43: Summary: Starting values from frequencies
44:
1.249 brouard 45: Revision 1.248 2016/09/07 14:10:18 brouard
46: *** empty log message ***
47:
1.248 brouard 48: Revision 1.247 2016/09/02 11:11:21 brouard
49: *** empty log message ***
50:
1.247 brouard 51: Revision 1.246 2016/09/02 08:49:22 brouard
52: *** empty log message ***
53:
1.246 brouard 54: Revision 1.245 2016/09/02 07:25:01 brouard
55: *** empty log message ***
56:
1.245 brouard 57: Revision 1.244 2016/09/02 07:17:34 brouard
58: *** empty log message ***
59:
1.244 brouard 60: Revision 1.243 2016/09/02 06:45:35 brouard
61: *** empty log message ***
62:
1.243 brouard 63: Revision 1.242 2016/08/30 15:01:20 brouard
64: Summary: Fixing a lots
65:
1.242 brouard 66: Revision 1.241 2016/08/29 17:17:25 brouard
67: Summary: gnuplot problem in Back projection to fix
68:
1.241 brouard 69: Revision 1.240 2016/08/29 07:53:18 brouard
70: Summary: Better
71:
1.240 brouard 72: Revision 1.239 2016/08/26 15:51:03 brouard
73: Summary: Improvement in Powell output in order to copy and paste
74:
75: Author:
76:
1.239 brouard 77: Revision 1.238 2016/08/26 14:23:35 brouard
78: Summary: Starting tests of 0.99
79:
1.238 brouard 80: Revision 1.237 2016/08/26 09:20:19 brouard
81: Summary: to valgrind
82:
1.237 brouard 83: Revision 1.236 2016/08/25 10:50:18 brouard
84: *** empty log message ***
85:
1.236 brouard 86: Revision 1.235 2016/08/25 06:59:23 brouard
87: *** empty log message ***
88:
1.235 brouard 89: Revision 1.234 2016/08/23 16:51:20 brouard
90: *** empty log message ***
91:
1.234 brouard 92: Revision 1.233 2016/08/23 07:40:50 brouard
93: Summary: not working
94:
1.233 brouard 95: Revision 1.232 2016/08/22 14:20:21 brouard
96: Summary: not working
97:
1.232 brouard 98: Revision 1.231 2016/08/22 07:17:15 brouard
99: Summary: not working
100:
1.231 brouard 101: Revision 1.230 2016/08/22 06:55:53 brouard
102: Summary: Not working
103:
1.230 brouard 104: Revision 1.229 2016/07/23 09:45:53 brouard
105: Summary: Completing for func too
106:
1.229 brouard 107: Revision 1.228 2016/07/22 17:45:30 brouard
108: Summary: Fixing some arrays, still debugging
109:
1.227 brouard 110: Revision 1.226 2016/07/12 18:42:34 brouard
111: Summary: temp
112:
1.226 brouard 113: Revision 1.225 2016/07/12 08:40:03 brouard
114: Summary: saving but not running
115:
1.225 brouard 116: Revision 1.224 2016/07/01 13:16:01 brouard
117: Summary: Fixes
118:
1.224 brouard 119: Revision 1.223 2016/02/19 09:23:35 brouard
120: Summary: temporary
121:
1.223 brouard 122: Revision 1.222 2016/02/17 08:14:50 brouard
123: Summary: Probably last 0.98 stable version 0.98r6
124:
1.222 brouard 125: Revision 1.221 2016/02/15 23:35:36 brouard
126: Summary: minor bug
127:
1.220 brouard 128: Revision 1.219 2016/02/15 00:48:12 brouard
129: *** empty log message ***
130:
1.219 brouard 131: Revision 1.218 2016/02/12 11:29:23 brouard
132: Summary: 0.99 Back projections
133:
1.218 brouard 134: Revision 1.217 2015/12/23 17:18:31 brouard
135: Summary: Experimental backcast
136:
1.217 brouard 137: Revision 1.216 2015/12/18 17:32:11 brouard
138: Summary: 0.98r4 Warning and status=-2
139:
140: Version 0.98r4 is now:
141: - displaying an error when status is -1, date of interview unknown and date of death known;
142: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
143: Older changes concerning s=-2, dating from 2005 have been supersed.
144:
1.216 brouard 145: Revision 1.215 2015/12/16 08:52:24 brouard
146: Summary: 0.98r4 working
147:
1.215 brouard 148: Revision 1.214 2015/12/16 06:57:54 brouard
149: Summary: temporary not working
150:
1.214 brouard 151: Revision 1.213 2015/12/11 18:22:17 brouard
152: Summary: 0.98r4
153:
1.213 brouard 154: Revision 1.212 2015/11/21 12:47:24 brouard
155: Summary: minor typo
156:
1.212 brouard 157: Revision 1.211 2015/11/21 12:41:11 brouard
158: Summary: 0.98r3 with some graph of projected cross-sectional
159:
160: Author: Nicolas Brouard
161:
1.211 brouard 162: Revision 1.210 2015/11/18 17:41:20 brouard
1.252 brouard 163: Summary: Start working on projected prevalences Revision 1.209 2015/11/17 22:12:03 brouard
1.210 brouard 164: Summary: Adding ftolpl parameter
165: Author: N Brouard
166:
167: We had difficulties to get smoothed confidence intervals. It was due
168: to the period prevalence which wasn't computed accurately. The inner
169: parameter ftolpl is now an outer parameter of the .imach parameter
170: file after estepm. If ftolpl is small 1.e-4 and estepm too,
171: computation are long.
172:
1.209 brouard 173: Revision 1.208 2015/11/17 14:31:57 brouard
174: Summary: temporary
175:
1.208 brouard 176: Revision 1.207 2015/10/27 17:36:57 brouard
177: *** empty log message ***
178:
1.207 brouard 179: Revision 1.206 2015/10/24 07:14:11 brouard
180: *** empty log message ***
181:
1.206 brouard 182: Revision 1.205 2015/10/23 15:50:53 brouard
183: Summary: 0.98r3 some clarification for graphs on likelihood contributions
184:
1.205 brouard 185: Revision 1.204 2015/10/01 16:20:26 brouard
186: Summary: Some new graphs of contribution to likelihood
187:
1.204 brouard 188: Revision 1.203 2015/09/30 17:45:14 brouard
189: Summary: looking at better estimation of the hessian
190:
191: Also a better criteria for convergence to the period prevalence And
192: therefore adding the number of years needed to converge. (The
193: prevalence in any alive state shold sum to one
194:
1.203 brouard 195: Revision 1.202 2015/09/22 19:45:16 brouard
196: Summary: Adding some overall graph on contribution to likelihood. Might change
197:
1.202 brouard 198: Revision 1.201 2015/09/15 17:34:58 brouard
199: Summary: 0.98r0
200:
201: - Some new graphs like suvival functions
202: - Some bugs fixed like model=1+age+V2.
203:
1.201 brouard 204: Revision 1.200 2015/09/09 16:53:55 brouard
205: Summary: Big bug thanks to Flavia
206:
207: Even model=1+age+V2. did not work anymore
208:
1.200 brouard 209: Revision 1.199 2015/09/07 14:09:23 brouard
210: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
211:
1.199 brouard 212: Revision 1.198 2015/09/03 07:14:39 brouard
213: Summary: 0.98q5 Flavia
214:
1.198 brouard 215: Revision 1.197 2015/09/01 18:24:39 brouard
216: *** empty log message ***
217:
1.197 brouard 218: Revision 1.196 2015/08/18 23:17:52 brouard
219: Summary: 0.98q5
220:
1.196 brouard 221: Revision 1.195 2015/08/18 16:28:39 brouard
222: Summary: Adding a hack for testing purpose
223:
224: After reading the title, ftol and model lines, if the comment line has
225: a q, starting with #q, the answer at the end of the run is quit. It
226: permits to run test files in batch with ctest. The former workaround was
227: $ echo q | imach foo.imach
228:
1.195 brouard 229: Revision 1.194 2015/08/18 13:32:00 brouard
230: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
231:
1.194 brouard 232: Revision 1.193 2015/08/04 07:17:42 brouard
233: Summary: 0.98q4
234:
1.193 brouard 235: Revision 1.192 2015/07/16 16:49:02 brouard
236: Summary: Fixing some outputs
237:
1.192 brouard 238: Revision 1.191 2015/07/14 10:00:33 brouard
239: Summary: Some fixes
240:
1.191 brouard 241: Revision 1.190 2015/05/05 08:51:13 brouard
242: Summary: Adding digits in output parameters (7 digits instead of 6)
243:
244: Fix 1+age+.
245:
1.190 brouard 246: Revision 1.189 2015/04/30 14:45:16 brouard
247: Summary: 0.98q2
248:
1.189 brouard 249: Revision 1.188 2015/04/30 08:27:53 brouard
250: *** empty log message ***
251:
1.188 brouard 252: Revision 1.187 2015/04/29 09:11:15 brouard
253: *** empty log message ***
254:
1.187 brouard 255: Revision 1.186 2015/04/23 12:01:52 brouard
256: Summary: V1*age is working now, version 0.98q1
257:
258: Some codes had been disabled in order to simplify and Vn*age was
259: working in the optimization phase, ie, giving correct MLE parameters,
260: but, as usual, outputs were not correct and program core dumped.
261:
1.186 brouard 262: Revision 1.185 2015/03/11 13:26:42 brouard
263: Summary: Inclusion of compile and links command line for Intel Compiler
264:
1.185 brouard 265: Revision 1.184 2015/03/11 11:52:39 brouard
266: Summary: Back from Windows 8. Intel Compiler
267:
1.184 brouard 268: Revision 1.183 2015/03/10 20:34:32 brouard
269: Summary: 0.98q0, trying with directest, mnbrak fixed
270:
271: We use directest instead of original Powell test; probably no
272: incidence on the results, but better justifications;
273: We fixed Numerical Recipes mnbrak routine which was wrong and gave
274: wrong results.
275:
1.183 brouard 276: Revision 1.182 2015/02/12 08:19:57 brouard
277: Summary: Trying to keep directest which seems simpler and more general
278: Author: Nicolas Brouard
279:
1.182 brouard 280: Revision 1.181 2015/02/11 23:22:24 brouard
281: Summary: Comments on Powell added
282:
283: Author:
284:
1.181 brouard 285: Revision 1.180 2015/02/11 17:33:45 brouard
286: Summary: Finishing move from main to function (hpijx and prevalence_limit)
287:
1.180 brouard 288: Revision 1.179 2015/01/04 09:57:06 brouard
289: Summary: back to OS/X
290:
1.179 brouard 291: Revision 1.178 2015/01/04 09:35:48 brouard
292: *** empty log message ***
293:
1.178 brouard 294: Revision 1.177 2015/01/03 18:40:56 brouard
295: Summary: Still testing ilc32 on OSX
296:
1.177 brouard 297: Revision 1.176 2015/01/03 16:45:04 brouard
298: *** empty log message ***
299:
1.176 brouard 300: Revision 1.175 2015/01/03 16:33:42 brouard
301: *** empty log message ***
302:
1.175 brouard 303: Revision 1.174 2015/01/03 16:15:49 brouard
304: Summary: Still in cross-compilation
305:
1.174 brouard 306: Revision 1.173 2015/01/03 12:06:26 brouard
307: Summary: trying to detect cross-compilation
308:
1.173 brouard 309: Revision 1.172 2014/12/27 12:07:47 brouard
310: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
311:
1.172 brouard 312: Revision 1.171 2014/12/23 13:26:59 brouard
313: Summary: Back from Visual C
314:
315: Still problem with utsname.h on Windows
316:
1.171 brouard 317: Revision 1.170 2014/12/23 11:17:12 brouard
318: Summary: Cleaning some \%% back to %%
319:
320: The escape was mandatory for a specific compiler (which one?), but too many warnings.
321:
1.170 brouard 322: Revision 1.169 2014/12/22 23:08:31 brouard
323: Summary: 0.98p
324:
325: Outputs some informations on compiler used, OS etc. Testing on different platforms.
326:
1.169 brouard 327: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 328: Summary: update
1.169 brouard 329:
1.168 brouard 330: Revision 1.167 2014/12/22 13:50:56 brouard
331: Summary: Testing uname and compiler version and if compiled 32 or 64
332:
333: Testing on Linux 64
334:
1.167 brouard 335: Revision 1.166 2014/12/22 11:40:47 brouard
336: *** empty log message ***
337:
1.166 brouard 338: Revision 1.165 2014/12/16 11:20:36 brouard
339: Summary: After compiling on Visual C
340:
341: * imach.c (Module): Merging 1.61 to 1.162
342:
1.165 brouard 343: Revision 1.164 2014/12/16 10:52:11 brouard
344: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
345:
346: * imach.c (Module): Merging 1.61 to 1.162
347:
1.164 brouard 348: Revision 1.163 2014/12/16 10:30:11 brouard
349: * imach.c (Module): Merging 1.61 to 1.162
350:
1.163 brouard 351: Revision 1.162 2014/09/25 11:43:39 brouard
352: Summary: temporary backup 0.99!
353:
1.162 brouard 354: Revision 1.1 2014/09/16 11:06:58 brouard
355: Summary: With some code (wrong) for nlopt
356:
357: Author:
358:
359: Revision 1.161 2014/09/15 20:41:41 brouard
360: Summary: Problem with macro SQR on Intel compiler
361:
1.161 brouard 362: Revision 1.160 2014/09/02 09:24:05 brouard
363: *** empty log message ***
364:
1.160 brouard 365: Revision 1.159 2014/09/01 10:34:10 brouard
366: Summary: WIN32
367: Author: Brouard
368:
1.159 brouard 369: Revision 1.158 2014/08/27 17:11:51 brouard
370: *** empty log message ***
371:
1.158 brouard 372: Revision 1.157 2014/08/27 16:26:55 brouard
373: Summary: Preparing windows Visual studio version
374: Author: Brouard
375:
376: In order to compile on Visual studio, time.h is now correct and time_t
377: and tm struct should be used. difftime should be used but sometimes I
378: just make the differences in raw time format (time(&now).
379: Trying to suppress #ifdef LINUX
380: Add xdg-open for __linux in order to open default browser.
381:
1.157 brouard 382: Revision 1.156 2014/08/25 20:10:10 brouard
383: *** empty log message ***
384:
1.156 brouard 385: Revision 1.155 2014/08/25 18:32:34 brouard
386: Summary: New compile, minor changes
387: Author: Brouard
388:
1.155 brouard 389: Revision 1.154 2014/06/20 17:32:08 brouard
390: Summary: Outputs now all graphs of convergence to period prevalence
391:
1.154 brouard 392: Revision 1.153 2014/06/20 16:45:46 brouard
393: Summary: If 3 live state, convergence to period prevalence on same graph
394: Author: Brouard
395:
1.153 brouard 396: Revision 1.152 2014/06/18 17:54:09 brouard
397: Summary: open browser, use gnuplot on same dir than imach if not found in the path
398:
1.152 brouard 399: Revision 1.151 2014/06/18 16:43:30 brouard
400: *** empty log message ***
401:
1.151 brouard 402: Revision 1.150 2014/06/18 16:42:35 brouard
403: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
404: Author: brouard
405:
1.150 brouard 406: Revision 1.149 2014/06/18 15:51:14 brouard
407: Summary: Some fixes in parameter files errors
408: Author: Nicolas Brouard
409:
1.149 brouard 410: Revision 1.148 2014/06/17 17:38:48 brouard
411: Summary: Nothing new
412: Author: Brouard
413:
414: Just a new packaging for OS/X version 0.98nS
415:
1.148 brouard 416: Revision 1.147 2014/06/16 10:33:11 brouard
417: *** empty log message ***
418:
1.147 brouard 419: Revision 1.146 2014/06/16 10:20:28 brouard
420: Summary: Merge
421: Author: Brouard
422:
423: Merge, before building revised version.
424:
1.146 brouard 425: Revision 1.145 2014/06/10 21:23:15 brouard
426: Summary: Debugging with valgrind
427: Author: Nicolas Brouard
428:
429: Lot of changes in order to output the results with some covariates
430: After the Edimburgh REVES conference 2014, it seems mandatory to
431: improve the code.
432: No more memory valgrind error but a lot has to be done in order to
433: continue the work of splitting the code into subroutines.
434: Also, decodemodel has been improved. Tricode is still not
435: optimal. nbcode should be improved. Documentation has been added in
436: the source code.
437:
1.144 brouard 438: Revision 1.143 2014/01/26 09:45:38 brouard
439: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
440:
441: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
442: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
443:
1.143 brouard 444: Revision 1.142 2014/01/26 03:57:36 brouard
445: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
446:
447: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
448:
1.142 brouard 449: Revision 1.141 2014/01/26 02:42:01 brouard
450: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
451:
1.141 brouard 452: Revision 1.140 2011/09/02 10:37:54 brouard
453: Summary: times.h is ok with mingw32 now.
454:
1.140 brouard 455: Revision 1.139 2010/06/14 07:50:17 brouard
456: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
457: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
458:
1.139 brouard 459: Revision 1.138 2010/04/30 18:19:40 brouard
460: *** empty log message ***
461:
1.138 brouard 462: Revision 1.137 2010/04/29 18:11:38 brouard
463: (Module): Checking covariates for more complex models
464: than V1+V2. A lot of change to be done. Unstable.
465:
1.137 brouard 466: Revision 1.136 2010/04/26 20:30:53 brouard
467: (Module): merging some libgsl code. Fixing computation
468: of likelione (using inter/intrapolation if mle = 0) in order to
469: get same likelihood as if mle=1.
470: Some cleaning of code and comments added.
471:
1.136 brouard 472: Revision 1.135 2009/10/29 15:33:14 brouard
473: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
474:
1.135 brouard 475: Revision 1.134 2009/10/29 13:18:53 brouard
476: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
477:
1.134 brouard 478: Revision 1.133 2009/07/06 10:21:25 brouard
479: just nforces
480:
1.133 brouard 481: Revision 1.132 2009/07/06 08:22:05 brouard
482: Many tings
483:
1.132 brouard 484: Revision 1.131 2009/06/20 16:22:47 brouard
485: Some dimensions resccaled
486:
1.131 brouard 487: Revision 1.130 2009/05/26 06:44:34 brouard
488: (Module): Max Covariate is now set to 20 instead of 8. A
489: lot of cleaning with variables initialized to 0. Trying to make
490: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
491:
1.130 brouard 492: Revision 1.129 2007/08/31 13:49:27 lievre
493: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
494:
1.129 lievre 495: Revision 1.128 2006/06/30 13:02:05 brouard
496: (Module): Clarifications on computing e.j
497:
1.128 brouard 498: Revision 1.127 2006/04/28 18:11:50 brouard
499: (Module): Yes the sum of survivors was wrong since
500: imach-114 because nhstepm was no more computed in the age
501: loop. Now we define nhstepma in the age loop.
502: (Module): In order to speed up (in case of numerous covariates) we
503: compute health expectancies (without variances) in a first step
504: and then all the health expectancies with variances or standard
505: deviation (needs data from the Hessian matrices) which slows the
506: computation.
507: In the future we should be able to stop the program is only health
508: expectancies and graph are needed without standard deviations.
509:
1.127 brouard 510: Revision 1.126 2006/04/28 17:23:28 brouard
511: (Module): Yes the sum of survivors was wrong since
512: imach-114 because nhstepm was no more computed in the age
513: loop. Now we define nhstepma in the age loop.
514: Version 0.98h
515:
1.126 brouard 516: Revision 1.125 2006/04/04 15:20:31 lievre
517: Errors in calculation of health expectancies. Age was not initialized.
518: Forecasting file added.
519:
520: Revision 1.124 2006/03/22 17:13:53 lievre
521: Parameters are printed with %lf instead of %f (more numbers after the comma).
522: The log-likelihood is printed in the log file
523:
524: Revision 1.123 2006/03/20 10:52:43 brouard
525: * imach.c (Module): <title> changed, corresponds to .htm file
526: name. <head> headers where missing.
527:
528: * imach.c (Module): Weights can have a decimal point as for
529: English (a comma might work with a correct LC_NUMERIC environment,
530: otherwise the weight is truncated).
531: Modification of warning when the covariates values are not 0 or
532: 1.
533: Version 0.98g
534:
535: Revision 1.122 2006/03/20 09:45:41 brouard
536: (Module): Weights can have a decimal point as for
537: English (a comma might work with a correct LC_NUMERIC environment,
538: otherwise the weight is truncated).
539: Modification of warning when the covariates values are not 0 or
540: 1.
541: Version 0.98g
542:
543: Revision 1.121 2006/03/16 17:45:01 lievre
544: * imach.c (Module): Comments concerning covariates added
545:
546: * imach.c (Module): refinements in the computation of lli if
547: status=-2 in order to have more reliable computation if stepm is
548: not 1 month. Version 0.98f
549:
550: Revision 1.120 2006/03/16 15:10:38 lievre
551: (Module): refinements in the computation of lli if
552: status=-2 in order to have more reliable computation if stepm is
553: not 1 month. Version 0.98f
554:
555: Revision 1.119 2006/03/15 17:42:26 brouard
556: (Module): Bug if status = -2, the loglikelihood was
557: computed as likelihood omitting the logarithm. Version O.98e
558:
559: Revision 1.118 2006/03/14 18:20:07 brouard
560: (Module): varevsij Comments added explaining the second
561: table of variances if popbased=1 .
562: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
563: (Module): Function pstamp added
564: (Module): Version 0.98d
565:
566: Revision 1.117 2006/03/14 17:16:22 brouard
567: (Module): varevsij Comments added explaining the second
568: table of variances if popbased=1 .
569: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
570: (Module): Function pstamp added
571: (Module): Version 0.98d
572:
573: Revision 1.116 2006/03/06 10:29:27 brouard
574: (Module): Variance-covariance wrong links and
575: varian-covariance of ej. is needed (Saito).
576:
577: Revision 1.115 2006/02/27 12:17:45 brouard
578: (Module): One freematrix added in mlikeli! 0.98c
579:
580: Revision 1.114 2006/02/26 12:57:58 brouard
581: (Module): Some improvements in processing parameter
582: filename with strsep.
583:
584: Revision 1.113 2006/02/24 14:20:24 brouard
585: (Module): Memory leaks checks with valgrind and:
586: datafile was not closed, some imatrix were not freed and on matrix
587: allocation too.
588:
589: Revision 1.112 2006/01/30 09:55:26 brouard
590: (Module): Back to gnuplot.exe instead of wgnuplot.exe
591:
592: Revision 1.111 2006/01/25 20:38:18 brouard
593: (Module): Lots of cleaning and bugs added (Gompertz)
594: (Module): Comments can be added in data file. Missing date values
595: can be a simple dot '.'.
596:
597: Revision 1.110 2006/01/25 00:51:50 brouard
598: (Module): Lots of cleaning and bugs added (Gompertz)
599:
600: Revision 1.109 2006/01/24 19:37:15 brouard
601: (Module): Comments (lines starting with a #) are allowed in data.
602:
603: Revision 1.108 2006/01/19 18:05:42 lievre
604: Gnuplot problem appeared...
605: To be fixed
606:
607: Revision 1.107 2006/01/19 16:20:37 brouard
608: Test existence of gnuplot in imach path
609:
610: Revision 1.106 2006/01/19 13:24:36 brouard
611: Some cleaning and links added in html output
612:
613: Revision 1.105 2006/01/05 20:23:19 lievre
614: *** empty log message ***
615:
616: Revision 1.104 2005/09/30 16:11:43 lievre
617: (Module): sump fixed, loop imx fixed, and simplifications.
618: (Module): If the status is missing at the last wave but we know
619: that the person is alive, then we can code his/her status as -2
620: (instead of missing=-1 in earlier versions) and his/her
621: contributions to the likelihood is 1 - Prob of dying from last
622: health status (= 1-p13= p11+p12 in the easiest case of somebody in
623: the healthy state at last known wave). Version is 0.98
624:
625: Revision 1.103 2005/09/30 15:54:49 lievre
626: (Module): sump fixed, loop imx fixed, and simplifications.
627:
628: Revision 1.102 2004/09/15 17:31:30 brouard
629: Add the possibility to read data file including tab characters.
630:
631: Revision 1.101 2004/09/15 10:38:38 brouard
632: Fix on curr_time
633:
634: Revision 1.100 2004/07/12 18:29:06 brouard
635: Add version for Mac OS X. Just define UNIX in Makefile
636:
637: Revision 1.99 2004/06/05 08:57:40 brouard
638: *** empty log message ***
639:
640: Revision 1.98 2004/05/16 15:05:56 brouard
641: New version 0.97 . First attempt to estimate force of mortality
642: directly from the data i.e. without the need of knowing the health
643: state at each age, but using a Gompertz model: log u =a + b*age .
644: This is the basic analysis of mortality and should be done before any
645: other analysis, in order to test if the mortality estimated from the
646: cross-longitudinal survey is different from the mortality estimated
647: from other sources like vital statistic data.
648:
649: The same imach parameter file can be used but the option for mle should be -3.
650:
1.133 brouard 651: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 652: former routines in order to include the new code within the former code.
653:
654: The output is very simple: only an estimate of the intercept and of
655: the slope with 95% confident intervals.
656:
657: Current limitations:
658: A) Even if you enter covariates, i.e. with the
659: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
660: B) There is no computation of Life Expectancy nor Life Table.
661:
662: Revision 1.97 2004/02/20 13:25:42 lievre
663: Version 0.96d. Population forecasting command line is (temporarily)
664: suppressed.
665:
666: Revision 1.96 2003/07/15 15:38:55 brouard
667: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
668: rewritten within the same printf. Workaround: many printfs.
669:
670: Revision 1.95 2003/07/08 07:54:34 brouard
671: * imach.c (Repository):
672: (Repository): Using imachwizard code to output a more meaningful covariance
673: matrix (cov(a12,c31) instead of numbers.
674:
675: Revision 1.94 2003/06/27 13:00:02 brouard
676: Just cleaning
677:
678: Revision 1.93 2003/06/25 16:33:55 brouard
679: (Module): On windows (cygwin) function asctime_r doesn't
680: exist so I changed back to asctime which exists.
681: (Module): Version 0.96b
682:
683: Revision 1.92 2003/06/25 16:30:45 brouard
684: (Module): On windows (cygwin) function asctime_r doesn't
685: exist so I changed back to asctime which exists.
686:
687: Revision 1.91 2003/06/25 15:30:29 brouard
688: * imach.c (Repository): Duplicated warning errors corrected.
689: (Repository): Elapsed time after each iteration is now output. It
690: helps to forecast when convergence will be reached. Elapsed time
691: is stamped in powell. We created a new html file for the graphs
692: concerning matrix of covariance. It has extension -cov.htm.
693:
694: Revision 1.90 2003/06/24 12:34:15 brouard
695: (Module): Some bugs corrected for windows. Also, when
696: mle=-1 a template is output in file "or"mypar.txt with the design
697: of the covariance matrix to be input.
698:
699: Revision 1.89 2003/06/24 12:30:52 brouard
700: (Module): Some bugs corrected for windows. Also, when
701: mle=-1 a template is output in file "or"mypar.txt with the design
702: of the covariance matrix to be input.
703:
704: Revision 1.88 2003/06/23 17:54:56 brouard
705: * imach.c (Repository): Create a sub-directory where all the secondary files are. Only imach, htm, gp and r(imach) are on the main directory. Correct time and other things.
706:
707: Revision 1.87 2003/06/18 12:26:01 brouard
708: Version 0.96
709:
710: Revision 1.86 2003/06/17 20:04:08 brouard
711: (Module): Change position of html and gnuplot routines and added
712: routine fileappend.
713:
714: Revision 1.85 2003/06/17 13:12:43 brouard
715: * imach.c (Repository): Check when date of death was earlier that
716: current date of interview. It may happen when the death was just
717: prior to the death. In this case, dh was negative and likelihood
718: was wrong (infinity). We still send an "Error" but patch by
719: assuming that the date of death was just one stepm after the
720: interview.
721: (Repository): Because some people have very long ID (first column)
722: we changed int to long in num[] and we added a new lvector for
723: memory allocation. But we also truncated to 8 characters (left
724: truncation)
725: (Repository): No more line truncation errors.
726:
727: Revision 1.84 2003/06/13 21:44:43 brouard
728: * imach.c (Repository): Replace "freqsummary" at a correct
729: place. It differs from routine "prevalence" which may be called
730: many times. Probs is memory consuming and must be used with
731: parcimony.
732: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
733:
734: Revision 1.83 2003/06/10 13:39:11 lievre
735: *** empty log message ***
736:
737: Revision 1.82 2003/06/05 15:57:20 brouard
738: Add log in imach.c and fullversion number is now printed.
739:
740: */
741: /*
742: Interpolated Markov Chain
743:
744: Short summary of the programme:
745:
1.227 brouard 746: This program computes Healthy Life Expectancies or State-specific
747: (if states aren't health statuses) Expectancies from
748: cross-longitudinal data. Cross-longitudinal data consist in:
749:
750: -1- a first survey ("cross") where individuals from different ages
751: are interviewed on their health status or degree of disability (in
752: the case of a health survey which is our main interest)
753:
754: -2- at least a second wave of interviews ("longitudinal") which
755: measure each change (if any) in individual health status. Health
756: expectancies are computed from the time spent in each health state
757: according to a model. More health states you consider, more time is
758: necessary to reach the Maximum Likelihood of the parameters involved
759: in the model. The simplest model is the multinomial logistic model
760: where pij is the probability to be observed in state j at the second
761: wave conditional to be observed in state i at the first
762: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
763: etc , where 'age' is age and 'sex' is a covariate. If you want to
764: have a more complex model than "constant and age", you should modify
765: the program where the markup *Covariates have to be included here
766: again* invites you to do it. More covariates you add, slower the
1.126 brouard 767: convergence.
768:
769: The advantage of this computer programme, compared to a simple
770: multinomial logistic model, is clear when the delay between waves is not
771: identical for each individual. Also, if a individual missed an
772: intermediate interview, the information is lost, but taken into
773: account using an interpolation or extrapolation.
774:
775: hPijx is the probability to be observed in state i at age x+h
776: conditional to the observed state i at age x. The delay 'h' can be
777: split into an exact number (nh*stepm) of unobserved intermediate
778: states. This elementary transition (by month, quarter,
779: semester or year) is modelled as a multinomial logistic. The hPx
780: matrix is simply the matrix product of nh*stepm elementary matrices
781: and the contribution of each individual to the likelihood is simply
782: hPijx.
783:
784: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 785: of the life expectancies. It also computes the period (stable) prevalence.
786:
787: Back prevalence and projections:
1.227 brouard 788:
789: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
790: double agemaxpar, double ftolpl, int *ncvyearp, double
791: dateprev1,double dateprev2, int firstpass, int lastpass, int
792: mobilavproj)
793:
794: Computes the back prevalence limit for any combination of
795: covariate values k at any age between ageminpar and agemaxpar and
796: returns it in **bprlim. In the loops,
797:
798: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
799: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
800:
801: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 802: Computes for any combination of covariates k and any age between bage and fage
803: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
804: oldm=oldms;savm=savms;
1.227 brouard 805:
806: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 807: Computes the transition matrix starting at age 'age' over
808: 'nhstepm*hstepm*stepm' months (i.e. until
809: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 810: nhstepm*hstepm matrices.
811:
812: Returns p3mat[i][j][h] after calling
813: p3mat[i][j][h]=matprod2(newm,
814: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
815: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
816: oldm);
1.226 brouard 817:
818: Important routines
819:
820: - func (or funcone), computes logit (pij) distinguishing
821: o fixed variables (single or product dummies or quantitative);
822: o varying variables by:
823: (1) wave (single, product dummies, quantitative),
824: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
825: % fixed dummy (treated) or quantitative (not done because time-consuming);
826: % varying dummy (not done) or quantitative (not done);
827: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
828: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
829: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
830: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
831: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 832:
1.226 brouard 833:
834:
1.133 brouard 835: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
836: Institut national d'études démographiques, Paris.
1.126 brouard 837: This software have been partly granted by Euro-REVES, a concerted action
838: from the European Union.
839: It is copyrighted identically to a GNU software product, ie programme and
840: software can be distributed freely for non commercial use. Latest version
841: can be accessed at http://euroreves.ined.fr/imach .
842:
843: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
844: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
845:
846: **********************************************************************/
847: /*
848: main
849: read parameterfile
850: read datafile
851: concatwav
852: freqsummary
853: if (mle >= 1)
854: mlikeli
855: print results files
856: if mle==1
857: computes hessian
858: read end of parameter file: agemin, agemax, bage, fage, estepm
859: begin-prev-date,...
860: open gnuplot file
861: open html file
1.145 brouard 862: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
863: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
864: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
865: freexexit2 possible for memory heap.
866:
867: h Pij x | pij_nom ficrestpij
868: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
869: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
870: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
871:
872: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
873: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
874: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
875: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
876: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
877:
1.126 brouard 878: forecasting if prevfcast==1 prevforecast call prevalence()
879: health expectancies
880: Variance-covariance of DFLE
881: prevalence()
882: movingaverage()
883: varevsij()
884: if popbased==1 varevsij(,popbased)
885: total life expectancies
886: Variance of period (stable) prevalence
887: end
888: */
889:
1.187 brouard 890: /* #define DEBUG */
891: /* #define DEBUGBRENT */
1.203 brouard 892: /* #define DEBUGLINMIN */
893: /* #define DEBUGHESS */
894: #define DEBUGHESSIJ
1.224 brouard 895: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 896: #define POWELL /* Instead of NLOPT */
1.224 brouard 897: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 898: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
899: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 900:
901: #include <math.h>
902: #include <stdio.h>
903: #include <stdlib.h>
904: #include <string.h>
1.226 brouard 905: #include <ctype.h>
1.159 brouard 906:
907: #ifdef _WIN32
908: #include <io.h>
1.172 brouard 909: #include <windows.h>
910: #include <tchar.h>
1.159 brouard 911: #else
1.126 brouard 912: #include <unistd.h>
1.159 brouard 913: #endif
1.126 brouard 914:
915: #include <limits.h>
916: #include <sys/types.h>
1.171 brouard 917:
918: #if defined(__GNUC__)
919: #include <sys/utsname.h> /* Doesn't work on Windows */
920: #endif
921:
1.126 brouard 922: #include <sys/stat.h>
923: #include <errno.h>
1.159 brouard 924: /* extern int errno; */
1.126 brouard 925:
1.157 brouard 926: /* #ifdef LINUX */
927: /* #include <time.h> */
928: /* #include "timeval.h" */
929: /* #else */
930: /* #include <sys/time.h> */
931: /* #endif */
932:
1.126 brouard 933: #include <time.h>
934:
1.136 brouard 935: #ifdef GSL
936: #include <gsl/gsl_errno.h>
937: #include <gsl/gsl_multimin.h>
938: #endif
939:
1.167 brouard 940:
1.162 brouard 941: #ifdef NLOPT
942: #include <nlopt.h>
943: typedef struct {
944: double (* function)(double [] );
945: } myfunc_data ;
946: #endif
947:
1.126 brouard 948: /* #include <libintl.h> */
949: /* #define _(String) gettext (String) */
950:
1.251 brouard 951: #define MAXLINE 2048 /* Was 256 and 1024. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 952:
953: #define GNUPLOTPROGRAM "gnuplot"
954: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
955: #define FILENAMELENGTH 132
956:
957: #define GLOCK_ERROR_NOPATH -1 /* empty path */
958: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
959:
1.144 brouard 960: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
961: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 962:
963: #define NINTERVMAX 8
1.144 brouard 964: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
965: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
966: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 967: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 968: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
969: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 970: #define MAXN 20000
1.144 brouard 971: #define YEARM 12. /**< Number of months per year */
1.218 brouard 972: /* #define AGESUP 130 */
973: #define AGESUP 150
974: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 975: #define AGEBASE 40
1.194 brouard 976: #define AGEOVERFLOW 1.e20
1.164 brouard 977: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 978: #ifdef _WIN32
979: #define DIRSEPARATOR '\\'
980: #define CHARSEPARATOR "\\"
981: #define ODIRSEPARATOR '/'
982: #else
1.126 brouard 983: #define DIRSEPARATOR '/'
984: #define CHARSEPARATOR "/"
985: #define ODIRSEPARATOR '\\'
986: #endif
987:
1.262 ! brouard 988: /* $Id: imach.c,v 1.261 2017/04/05 10:14:09 brouard Exp $ */
1.126 brouard 989: /* $State: Exp $ */
1.196 brouard 990: #include "version.h"
991: char version[]=__IMACH_VERSION__;
1.224 brouard 992: char copyright[]="February 2016,INED-EUROREVES-Institut de longevite-Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research 25293121), Intel Software 2015-2018";
1.262 ! brouard 993: char fullversion[]="$Revision: 1.261 $ $Date: 2017/04/05 10:14:09 $";
1.126 brouard 994: char strstart[80];
995: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 996: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 997: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 998: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
999: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
1000: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 1001: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
1002: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 1003: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
1004: int cptcovprodnoage=0; /**< Number of covariate products without age */
1005: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 1006: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
1007: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 1008: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 1009: int nsd=0; /**< Total number of single dummy variables (output) */
1010: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 1011: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 1012: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 1013: int ntveff=0; /**< ntveff number of effective time varying variables */
1014: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 1015: int cptcov=0; /* Working variable */
1.218 brouard 1016: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 1017: int npar=NPARMAX;
1018: int nlstate=2; /* Number of live states */
1019: int ndeath=1; /* Number of dead states */
1.130 brouard 1020: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 1021: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 1022: int popbased=0;
1023:
1024: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 1025: int maxwav=0; /* Maxim number of waves */
1026: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
1027: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
1028: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 1029: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 1030: int mle=1, weightopt=0;
1.126 brouard 1031: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
1032: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
1033: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
1034: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 1035: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 1036: int selected(int kvar); /* Is covariate kvar selected for printing results */
1037:
1.130 brouard 1038: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 1039: double **matprod2(); /* test */
1.126 brouard 1040: double **oldm, **newm, **savm; /* Working pointers to matrices */
1041: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 1042: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
1043:
1.136 brouard 1044: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 1045: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 1046: FILE *ficlog, *ficrespow;
1.130 brouard 1047: int globpr=0; /* Global variable for printing or not */
1.126 brouard 1048: double fretone; /* Only one call to likelihood */
1.130 brouard 1049: long ipmx=0; /* Number of contributions */
1.126 brouard 1050: double sw; /* Sum of weights */
1051: char filerespow[FILENAMELENGTH];
1052: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
1053: FILE *ficresilk;
1054: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
1055: FILE *ficresprobmorprev;
1056: FILE *fichtm, *fichtmcov; /* Html File */
1057: FILE *ficreseij;
1058: char filerese[FILENAMELENGTH];
1059: FILE *ficresstdeij;
1060: char fileresstde[FILENAMELENGTH];
1061: FILE *ficrescveij;
1062: char filerescve[FILENAMELENGTH];
1063: FILE *ficresvij;
1064: char fileresv[FILENAMELENGTH];
1065: FILE *ficresvpl;
1066: char fileresvpl[FILENAMELENGTH];
1067: char title[MAXLINE];
1.234 brouard 1068: char model[MAXLINE]; /**< The model line */
1.217 brouard 1069: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 1070: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
1071: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
1072: char command[FILENAMELENGTH];
1073: int outcmd=0;
1074:
1.217 brouard 1075: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 1076: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 1077: char filelog[FILENAMELENGTH]; /* Log file */
1078: char filerest[FILENAMELENGTH];
1079: char fileregp[FILENAMELENGTH];
1080: char popfile[FILENAMELENGTH];
1081:
1082: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1083:
1.157 brouard 1084: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1085: /* struct timezone tzp; */
1086: /* extern int gettimeofday(); */
1087: struct tm tml, *gmtime(), *localtime();
1088:
1089: extern time_t time();
1090:
1091: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1092: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1093: struct tm tm;
1094:
1.126 brouard 1095: char strcurr[80], strfor[80];
1096:
1097: char *endptr;
1098: long lval;
1099: double dval;
1100:
1101: #define NR_END 1
1102: #define FREE_ARG char*
1103: #define FTOL 1.0e-10
1104:
1105: #define NRANSI
1.240 brouard 1106: #define ITMAX 200
1107: #define ITPOWMAX 20 /* This is now multiplied by the number of parameters */
1.126 brouard 1108:
1109: #define TOL 2.0e-4
1110:
1111: #define CGOLD 0.3819660
1112: #define ZEPS 1.0e-10
1113: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1114:
1115: #define GOLD 1.618034
1116: #define GLIMIT 100.0
1117: #define TINY 1.0e-20
1118:
1119: static double maxarg1,maxarg2;
1120: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1121: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1122:
1123: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1124: #define rint(a) floor(a+0.5)
1.166 brouard 1125: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1126: #define mytinydouble 1.0e-16
1.166 brouard 1127: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1128: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1129: /* static double dsqrarg; */
1130: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1131: static double sqrarg;
1132: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1133: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1134: int agegomp= AGEGOMP;
1135:
1136: int imx;
1137: int stepm=1;
1138: /* Stepm, step in month: minimum step interpolation*/
1139:
1140: int estepm;
1141: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1142:
1143: int m,nb;
1144: long *num;
1.197 brouard 1145: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1146: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1147: covariate for which somebody answered excluding
1148: undefined. Usually 2: 0 and 1. */
1149: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1150: covariate for which somebody answered including
1151: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1152: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1153: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1154: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1155: double *ageexmed,*agecens;
1156: double dateintmean=0;
1157:
1158: double *weight;
1159: int **s; /* Status */
1.141 brouard 1160: double *agedc;
1.145 brouard 1161: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1162: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1163: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1164: double **coqvar; /* Fixed quantitative covariate iqv */
1165: double ***cotvar; /* Time varying covariate itv */
1166: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1167: double idx;
1168: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1169: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1170: /*k 1 2 3 4 5 6 7 8 9 */
1171: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1172: /* Tndvar[k] 1 2 3 4 5 */
1173: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1174: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1175: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1176: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1177: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1178: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1179: /* Tprod[i]=k 4 7 */
1180: /* Tage[i]=k 5 8 */
1181: /* */
1182: /* Type */
1183: /* V 1 2 3 4 5 */
1184: /* F F V V V */
1185: /* D Q D D Q */
1186: /* */
1187: int *TvarsD;
1188: int *TvarsDind;
1189: int *TvarsQ;
1190: int *TvarsQind;
1191:
1.235 brouard 1192: #define MAXRESULTLINES 10
1193: int nresult=0;
1.258 brouard 1194: int parameterline=0; /* # of the parameter (type) line */
1.235 brouard 1195: int TKresult[MAXRESULTLINES];
1.237 brouard 1196: int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1197: int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1.235 brouard 1198: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1199: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.237 brouard 1200: double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.235 brouard 1201: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1202:
1.234 brouard 1203: /* int *TDvar; /\**< TDvar[1]=4, TDvarF[2]=3, TDvar[3]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 *\/ */
1.232 brouard 1204: int *TvarF; /**< TvarF[1]=Tvar[6]=2, TvarF[2]=Tvar[7]=7, TvarF[3]=Tvar[9]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1205: int *TvarFind; /**< TvarFind[1]=6, TvarFind[2]=7, Tvarind[3]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1206: int *TvarV; /**< TvarV[1]=Tvar[1]=5, TvarV[2]=Tvar[2]=4 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1207: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1208: int *TvarA; /**< TvarA[1]=Tvar[5]=5, TvarA[2]=Tvar[8]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1209: int *TvarAind; /**< TvarindA[1]=5, TvarAind[2]=8 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 1210: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1211: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1212: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1213: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1214: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1215: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1216: int *TvarVQ; /* TvarVQ[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1217: int *TvarVQind; /* TvarVQind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1218:
1.230 brouard 1219: int *Tvarsel; /**< Selected covariates for output */
1220: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1221: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1222: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1223: int *Dummy; /** Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product */
1.238 brouard 1224: int *DummyV; /** Dummy[v] 0=dummy (0 1), 1 quantitative */
1225: int *FixedV; /** FixedV[v] 0 fixed, 1 varying */
1.197 brouard 1226: int *Tage;
1.227 brouard 1227: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1228: int *Tmodelind; /** Tmodelind[Tvaraff[3]]=9 for V1 position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.230 brouard 1229: int *TmodelInvind; /** Tmodelind[Tvaraff[3]]=9 for V1 position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1230: int *TmodelInvQind; /** Tmodelqind[1]=1 for V5(quantitative varying) position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.145 brouard 1231: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1232: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1233: int **Tvard;
1234: int *Tprod;/**< Gives the k position of the k1 product */
1.238 brouard 1235: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3 */
1.227 brouard 1236: int *Tposprod; /**< Gives the k1 product from the k position */
1.238 brouard 1237: /* if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2) */
1238: /* Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5(V3*V2)]=2 (2nd product without age) */
1.227 brouard 1239: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1240: double *lsurv, *lpop, *tpop;
1241:
1.231 brouard 1242: #define FD 1; /* Fixed dummy covariate */
1243: #define FQ 2; /* Fixed quantitative covariate */
1244: #define FP 3; /* Fixed product covariate */
1245: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1246: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1247: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1248: #define VD 10; /* Varying dummy covariate */
1249: #define VQ 11; /* Varying quantitative covariate */
1250: #define VP 12; /* Varying product covariate */
1251: #define VPDD 13; /* Varying product dummy*dummy covariate */
1252: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1253: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1254: #define APFD 16; /* Age product * fixed dummy covariate */
1255: #define APFQ 17; /* Age product * fixed quantitative covariate */
1256: #define APVD 18; /* Age product * varying dummy covariate */
1257: #define APVQ 19; /* Age product * varying quantitative covariate */
1258:
1259: #define FTYPE 1; /* Fixed covariate */
1260: #define VTYPE 2; /* Varying covariate (loop in wave) */
1261: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1262:
1263: struct kmodel{
1264: int maintype; /* main type */
1265: int subtype; /* subtype */
1266: };
1267: struct kmodel modell[NCOVMAX];
1268:
1.143 brouard 1269: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1270: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1271:
1272: /**************** split *************************/
1273: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1274: {
1275: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1276: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1277: */
1278: char *ss; /* pointer */
1.186 brouard 1279: int l1=0, l2=0; /* length counters */
1.126 brouard 1280:
1281: l1 = strlen(path ); /* length of path */
1282: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1283: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1284: if ( ss == NULL ) { /* no directory, so determine current directory */
1285: strcpy( name, path ); /* we got the fullname name because no directory */
1286: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1287: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1288: /* get current working directory */
1289: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1290: #ifdef WIN32
1291: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1292: #else
1293: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1294: #endif
1.126 brouard 1295: return( GLOCK_ERROR_GETCWD );
1296: }
1297: /* got dirc from getcwd*/
1298: printf(" DIRC = %s \n",dirc);
1.205 brouard 1299: } else { /* strip directory from path */
1.126 brouard 1300: ss++; /* after this, the filename */
1301: l2 = strlen( ss ); /* length of filename */
1302: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1303: strcpy( name, ss ); /* save file name */
1304: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1305: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1306: printf(" DIRC2 = %s \n",dirc);
1307: }
1308: /* We add a separator at the end of dirc if not exists */
1309: l1 = strlen( dirc ); /* length of directory */
1310: if( dirc[l1-1] != DIRSEPARATOR ){
1311: dirc[l1] = DIRSEPARATOR;
1312: dirc[l1+1] = 0;
1313: printf(" DIRC3 = %s \n",dirc);
1314: }
1315: ss = strrchr( name, '.' ); /* find last / */
1316: if (ss >0){
1317: ss++;
1318: strcpy(ext,ss); /* save extension */
1319: l1= strlen( name);
1320: l2= strlen(ss)+1;
1321: strncpy( finame, name, l1-l2);
1322: finame[l1-l2]= 0;
1323: }
1324:
1325: return( 0 ); /* we're done */
1326: }
1327:
1328:
1329: /******************************************/
1330:
1331: void replace_back_to_slash(char *s, char*t)
1332: {
1333: int i;
1334: int lg=0;
1335: i=0;
1336: lg=strlen(t);
1337: for(i=0; i<= lg; i++) {
1338: (s[i] = t[i]);
1339: if (t[i]== '\\') s[i]='/';
1340: }
1341: }
1342:
1.132 brouard 1343: char *trimbb(char *out, char *in)
1.137 brouard 1344: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1345: char *s;
1346: s=out;
1347: while (*in != '\0'){
1.137 brouard 1348: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1349: in++;
1350: }
1351: *out++ = *in++;
1352: }
1353: *out='\0';
1354: return s;
1355: }
1356:
1.187 brouard 1357: /* char *substrchaine(char *out, char *in, char *chain) */
1358: /* { */
1359: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1360: /* char *s, *t; */
1361: /* t=in;s=out; */
1362: /* while ((*in != *chain) && (*in != '\0')){ */
1363: /* *out++ = *in++; */
1364: /* } */
1365:
1366: /* /\* *in matches *chain *\/ */
1367: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1368: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1369: /* } */
1370: /* in--; chain--; */
1371: /* while ( (*in != '\0')){ */
1372: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1373: /* *out++ = *in++; */
1374: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1375: /* } */
1376: /* *out='\0'; */
1377: /* out=s; */
1378: /* return out; */
1379: /* } */
1380: char *substrchaine(char *out, char *in, char *chain)
1381: {
1382: /* Substract chain 'chain' from 'in', return and output 'out' */
1383: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1384:
1385: char *strloc;
1386:
1387: strcpy (out, in);
1388: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1389: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1390: if(strloc != NULL){
1391: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1392: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1393: /* strcpy (strloc, strloc +strlen(chain));*/
1394: }
1395: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1396: return out;
1397: }
1398:
1399:
1.145 brouard 1400: char *cutl(char *blocc, char *alocc, char *in, char occ)
1401: {
1.187 brouard 1402: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1403: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1404: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1405: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1406: */
1.160 brouard 1407: char *s, *t;
1.145 brouard 1408: t=in;s=in;
1409: while ((*in != occ) && (*in != '\0')){
1410: *alocc++ = *in++;
1411: }
1412: if( *in == occ){
1413: *(alocc)='\0';
1414: s=++in;
1415: }
1416:
1417: if (s == t) {/* occ not found */
1418: *(alocc-(in-s))='\0';
1419: in=s;
1420: }
1421: while ( *in != '\0'){
1422: *blocc++ = *in++;
1423: }
1424:
1425: *blocc='\0';
1426: return t;
1427: }
1.137 brouard 1428: char *cutv(char *blocc, char *alocc, char *in, char occ)
1429: {
1.187 brouard 1430: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1431: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1432: gives blocc="abcdef2ghi" and alocc="j".
1433: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1434: */
1435: char *s, *t;
1436: t=in;s=in;
1437: while (*in != '\0'){
1438: while( *in == occ){
1439: *blocc++ = *in++;
1440: s=in;
1441: }
1442: *blocc++ = *in++;
1443: }
1444: if (s == t) /* occ not found */
1445: *(blocc-(in-s))='\0';
1446: else
1447: *(blocc-(in-s)-1)='\0';
1448: in=s;
1449: while ( *in != '\0'){
1450: *alocc++ = *in++;
1451: }
1452:
1453: *alocc='\0';
1454: return s;
1455: }
1456:
1.126 brouard 1457: int nbocc(char *s, char occ)
1458: {
1459: int i,j=0;
1460: int lg=20;
1461: i=0;
1462: lg=strlen(s);
1463: for(i=0; i<= lg; i++) {
1.234 brouard 1464: if (s[i] == occ ) j++;
1.126 brouard 1465: }
1466: return j;
1467: }
1468:
1.137 brouard 1469: /* void cutv(char *u,char *v, char*t, char occ) */
1470: /* { */
1471: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1472: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1473: /* gives u="abcdef2ghi" and v="j" *\/ */
1474: /* int i,lg,j,p=0; */
1475: /* i=0; */
1476: /* lg=strlen(t); */
1477: /* for(j=0; j<=lg-1; j++) { */
1478: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1479: /* } */
1.126 brouard 1480:
1.137 brouard 1481: /* for(j=0; j<p; j++) { */
1482: /* (u[j] = t[j]); */
1483: /* } */
1484: /* u[p]='\0'; */
1.126 brouard 1485:
1.137 brouard 1486: /* for(j=0; j<= lg; j++) { */
1487: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1488: /* } */
1489: /* } */
1.126 brouard 1490:
1.160 brouard 1491: #ifdef _WIN32
1492: char * strsep(char **pp, const char *delim)
1493: {
1494: char *p, *q;
1495:
1496: if ((p = *pp) == NULL)
1497: return 0;
1498: if ((q = strpbrk (p, delim)) != NULL)
1499: {
1500: *pp = q + 1;
1501: *q = '\0';
1502: }
1503: else
1504: *pp = 0;
1505: return p;
1506: }
1507: #endif
1508:
1.126 brouard 1509: /********************** nrerror ********************/
1510:
1511: void nrerror(char error_text[])
1512: {
1513: fprintf(stderr,"ERREUR ...\n");
1514: fprintf(stderr,"%s\n",error_text);
1515: exit(EXIT_FAILURE);
1516: }
1517: /*********************** vector *******************/
1518: double *vector(int nl, int nh)
1519: {
1520: double *v;
1521: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1522: if (!v) nrerror("allocation failure in vector");
1523: return v-nl+NR_END;
1524: }
1525:
1526: /************************ free vector ******************/
1527: void free_vector(double*v, int nl, int nh)
1528: {
1529: free((FREE_ARG)(v+nl-NR_END));
1530: }
1531:
1532: /************************ivector *******************************/
1533: int *ivector(long nl,long nh)
1534: {
1535: int *v;
1536: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1537: if (!v) nrerror("allocation failure in ivector");
1538: return v-nl+NR_END;
1539: }
1540:
1541: /******************free ivector **************************/
1542: void free_ivector(int *v, long nl, long nh)
1543: {
1544: free((FREE_ARG)(v+nl-NR_END));
1545: }
1546:
1547: /************************lvector *******************************/
1548: long *lvector(long nl,long nh)
1549: {
1550: long *v;
1551: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1552: if (!v) nrerror("allocation failure in ivector");
1553: return v-nl+NR_END;
1554: }
1555:
1556: /******************free lvector **************************/
1557: void free_lvector(long *v, long nl, long nh)
1558: {
1559: free((FREE_ARG)(v+nl-NR_END));
1560: }
1561:
1562: /******************* imatrix *******************************/
1563: int **imatrix(long nrl, long nrh, long ncl, long nch)
1564: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1565: {
1566: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1567: int **m;
1568:
1569: /* allocate pointers to rows */
1570: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1571: if (!m) nrerror("allocation failure 1 in matrix()");
1572: m += NR_END;
1573: m -= nrl;
1574:
1575:
1576: /* allocate rows and set pointers to them */
1577: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1578: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1579: m[nrl] += NR_END;
1580: m[nrl] -= ncl;
1581:
1582: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1583:
1584: /* return pointer to array of pointers to rows */
1585: return m;
1586: }
1587:
1588: /****************** free_imatrix *************************/
1589: void free_imatrix(m,nrl,nrh,ncl,nch)
1590: int **m;
1591: long nch,ncl,nrh,nrl;
1592: /* free an int matrix allocated by imatrix() */
1593: {
1594: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1595: free((FREE_ARG) (m+nrl-NR_END));
1596: }
1597:
1598: /******************* matrix *******************************/
1599: double **matrix(long nrl, long nrh, long ncl, long nch)
1600: {
1601: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1602: double **m;
1603:
1604: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1605: if (!m) nrerror("allocation failure 1 in matrix()");
1606: m += NR_END;
1607: m -= nrl;
1608:
1609: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1610: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1611: m[nrl] += NR_END;
1612: m[nrl] -= ncl;
1613:
1614: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1615: return m;
1.145 brouard 1616: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1617: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1618: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1619: */
1620: }
1621:
1622: /*************************free matrix ************************/
1623: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1624: {
1625: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1626: free((FREE_ARG)(m+nrl-NR_END));
1627: }
1628:
1629: /******************* ma3x *******************************/
1630: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1631: {
1632: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1633: double ***m;
1634:
1635: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1636: if (!m) nrerror("allocation failure 1 in matrix()");
1637: m += NR_END;
1638: m -= nrl;
1639:
1640: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1641: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1642: m[nrl] += NR_END;
1643: m[nrl] -= ncl;
1644:
1645: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1646:
1647: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1648: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1649: m[nrl][ncl] += NR_END;
1650: m[nrl][ncl] -= nll;
1651: for (j=ncl+1; j<=nch; j++)
1652: m[nrl][j]=m[nrl][j-1]+nlay;
1653:
1654: for (i=nrl+1; i<=nrh; i++) {
1655: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1656: for (j=ncl+1; j<=nch; j++)
1657: m[i][j]=m[i][j-1]+nlay;
1658: }
1659: return m;
1660: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1661: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1662: */
1663: }
1664:
1665: /*************************free ma3x ************************/
1666: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1667: {
1668: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1669: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1670: free((FREE_ARG)(m+nrl-NR_END));
1671: }
1672:
1673: /*************** function subdirf ***********/
1674: char *subdirf(char fileres[])
1675: {
1676: /* Caution optionfilefiname is hidden */
1677: strcpy(tmpout,optionfilefiname);
1678: strcat(tmpout,"/"); /* Add to the right */
1679: strcat(tmpout,fileres);
1680: return tmpout;
1681: }
1682:
1683: /*************** function subdirf2 ***********/
1684: char *subdirf2(char fileres[], char *preop)
1685: {
1686:
1687: /* Caution optionfilefiname is hidden */
1688: strcpy(tmpout,optionfilefiname);
1689: strcat(tmpout,"/");
1690: strcat(tmpout,preop);
1691: strcat(tmpout,fileres);
1692: return tmpout;
1693: }
1694:
1695: /*************** function subdirf3 ***********/
1696: char *subdirf3(char fileres[], char *preop, char *preop2)
1697: {
1698:
1699: /* Caution optionfilefiname is hidden */
1700: strcpy(tmpout,optionfilefiname);
1701: strcat(tmpout,"/");
1702: strcat(tmpout,preop);
1703: strcat(tmpout,preop2);
1704: strcat(tmpout,fileres);
1705: return tmpout;
1706: }
1.213 brouard 1707:
1708: /*************** function subdirfext ***********/
1709: char *subdirfext(char fileres[], char *preop, char *postop)
1710: {
1711:
1712: strcpy(tmpout,preop);
1713: strcat(tmpout,fileres);
1714: strcat(tmpout,postop);
1715: return tmpout;
1716: }
1.126 brouard 1717:
1.213 brouard 1718: /*************** function subdirfext3 ***********/
1719: char *subdirfext3(char fileres[], char *preop, char *postop)
1720: {
1721:
1722: /* Caution optionfilefiname is hidden */
1723: strcpy(tmpout,optionfilefiname);
1724: strcat(tmpout,"/");
1725: strcat(tmpout,preop);
1726: strcat(tmpout,fileres);
1727: strcat(tmpout,postop);
1728: return tmpout;
1729: }
1730:
1.162 brouard 1731: char *asc_diff_time(long time_sec, char ascdiff[])
1732: {
1733: long sec_left, days, hours, minutes;
1734: days = (time_sec) / (60*60*24);
1735: sec_left = (time_sec) % (60*60*24);
1736: hours = (sec_left) / (60*60) ;
1737: sec_left = (sec_left) %(60*60);
1738: minutes = (sec_left) /60;
1739: sec_left = (sec_left) % (60);
1740: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1741: return ascdiff;
1742: }
1743:
1.126 brouard 1744: /***************** f1dim *************************/
1745: extern int ncom;
1746: extern double *pcom,*xicom;
1747: extern double (*nrfunc)(double []);
1748:
1749: double f1dim(double x)
1750: {
1751: int j;
1752: double f;
1753: double *xt;
1754:
1755: xt=vector(1,ncom);
1756: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1757: f=(*nrfunc)(xt);
1758: free_vector(xt,1,ncom);
1759: return f;
1760: }
1761:
1762: /*****************brent *************************/
1763: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1764: {
1765: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1766: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1767: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1768: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1769: * returned function value.
1770: */
1.126 brouard 1771: int iter;
1772: double a,b,d,etemp;
1.159 brouard 1773: double fu=0,fv,fw,fx;
1.164 brouard 1774: double ftemp=0.;
1.126 brouard 1775: double p,q,r,tol1,tol2,u,v,w,x,xm;
1776: double e=0.0;
1777:
1778: a=(ax < cx ? ax : cx);
1779: b=(ax > cx ? ax : cx);
1780: x=w=v=bx;
1781: fw=fv=fx=(*f)(x);
1782: for (iter=1;iter<=ITMAX;iter++) {
1783: xm=0.5*(a+b);
1784: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1785: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1786: printf(".");fflush(stdout);
1787: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1788: #ifdef DEBUGBRENT
1.126 brouard 1789: printf("br %d,x=%.10e xm=%.10e b=%.10e a=%.10e tol=%.10e tol1=%.10e tol2=%.10e x-xm=%.10e fx=%.12e fu=%.12e,fw=%.12e,ftemp=%.12e,ftol=%.12e\n",iter,x,xm,b,a,tol,tol1,tol2,(x-xm),fx,fu,fw,ftemp,ftol);
1790: fprintf(ficlog,"br %d,x=%.10e xm=%.10e b=%.10e a=%.10e tol=%.10e tol1=%.10e tol2=%.10e x-xm=%.10e fx=%.12e fu=%.12e,fw=%.12e,ftemp=%.12e,ftol=%.12e\n",iter,x,xm,b,a,tol,tol1,tol2,(x-xm),fx,fu,fw,ftemp,ftol);
1791: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1792: #endif
1793: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1794: *xmin=x;
1795: return fx;
1796: }
1797: ftemp=fu;
1798: if (fabs(e) > tol1) {
1799: r=(x-w)*(fx-fv);
1800: q=(x-v)*(fx-fw);
1801: p=(x-v)*q-(x-w)*r;
1802: q=2.0*(q-r);
1803: if (q > 0.0) p = -p;
1804: q=fabs(q);
1805: etemp=e;
1806: e=d;
1807: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1808: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1809: else {
1.224 brouard 1810: d=p/q;
1811: u=x+d;
1812: if (u-a < tol2 || b-u < tol2)
1813: d=SIGN(tol1,xm-x);
1.126 brouard 1814: }
1815: } else {
1816: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1817: }
1818: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1819: fu=(*f)(u);
1820: if (fu <= fx) {
1821: if (u >= x) a=x; else b=x;
1822: SHFT(v,w,x,u)
1.183 brouard 1823: SHFT(fv,fw,fx,fu)
1824: } else {
1825: if (u < x) a=u; else b=u;
1826: if (fu <= fw || w == x) {
1.224 brouard 1827: v=w;
1828: w=u;
1829: fv=fw;
1830: fw=fu;
1.183 brouard 1831: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1832: v=u;
1833: fv=fu;
1.183 brouard 1834: }
1835: }
1.126 brouard 1836: }
1837: nrerror("Too many iterations in brent");
1838: *xmin=x;
1839: return fx;
1840: }
1841:
1842: /****************** mnbrak ***********************/
1843:
1844: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1845: double (*func)(double))
1.183 brouard 1846: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1847: the downhill direction (defined by the function as evaluated at the initial points) and returns
1848: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1849: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1850: */
1.126 brouard 1851: double ulim,u,r,q, dum;
1852: double fu;
1.187 brouard 1853:
1854: double scale=10.;
1855: int iterscale=0;
1856:
1857: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1858: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1859:
1860:
1861: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1862: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1863: /* *bx = *ax - (*ax - *bx)/scale; */
1864: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1865: /* } */
1866:
1.126 brouard 1867: if (*fb > *fa) {
1868: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1869: SHFT(dum,*fb,*fa,dum)
1870: }
1.126 brouard 1871: *cx=(*bx)+GOLD*(*bx-*ax);
1872: *fc=(*func)(*cx);
1.183 brouard 1873: #ifdef DEBUG
1.224 brouard 1874: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1875: fprintf(ficlog,"mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1.183 brouard 1876: #endif
1.224 brouard 1877: while (*fb > *fc) { /* Declining a,b,c with fa> fb > fc. If fc=inf it exits and if flat fb=fc it exits too.*/
1.126 brouard 1878: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1879: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1880: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1881: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1882: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1883: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1884: fu=(*func)(u);
1.163 brouard 1885: #ifdef DEBUG
1886: /* f(x)=A(x-u)**2+f(u) */
1887: double A, fparabu;
1888: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1889: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1890: printf("\nmnbrak (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf), (*u=%.12f, fu=%.12lf, fparabu=%.12f, q=%lf < %lf=r)\n",*ax,*fa,*bx,*fb,*cx,*fc,u,fu, fparabu,q,r);
1891: fprintf(ficlog,"\nmnbrak (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf), (*u=%.12f, fu=%.12lf, fparabu=%.12f, q=%lf < %lf=r)\n",*ax,*fa,*bx,*fb,*cx,*fc,u,fu, fparabu,q,r);
1.183 brouard 1892: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1893: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1894: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1895: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1896: #endif
1.184 brouard 1897: #ifdef MNBRAKORIGINAL
1.183 brouard 1898: #else
1.191 brouard 1899: /* if (fu > *fc) { */
1900: /* #ifdef DEBUG */
1901: /* printf("mnbrak4 fu > fc \n"); */
1902: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1903: /* #endif */
1904: /* /\* SHFT(u,*cx,*cx,u) /\\* ie a=c, c=u and u=c; in that case, next SHFT(a,b,c,u) will give a=b=b, b=c=u, c=u=c and *\\/ *\/ */
1905: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1906: /* dum=u; /\* Shifting c and u *\/ */
1907: /* u = *cx; */
1908: /* *cx = dum; */
1909: /* dum = fu; */
1910: /* fu = *fc; */
1911: /* *fc =dum; */
1912: /* } else { /\* end *\/ */
1913: /* #ifdef DEBUG */
1914: /* printf("mnbrak3 fu < fc \n"); */
1915: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1916: /* #endif */
1917: /* dum=u; /\* Shifting c and u *\/ */
1918: /* u = *cx; */
1919: /* *cx = dum; */
1920: /* dum = fu; */
1921: /* fu = *fc; */
1922: /* *fc =dum; */
1923: /* } */
1.224 brouard 1924: #ifdef DEBUGMNBRAK
1925: double A, fparabu;
1926: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1927: fparabu= *fa - A*(*ax-u)*(*ax-u);
1928: printf("\nmnbrak35 ax=%lf fa=%lf bx=%lf fb=%lf, u=%lf fp=%lf fu=%lf < or >= fc=%lf cx=%lf, q=%lf < %lf=r \n",*ax, *fa, *bx,*fb,u,fparabu,fu,*fc,*cx,q,r);
1929: fprintf(ficlog,"\nmnbrak35 ax=%lf fa=%lf bx=%lf fb=%lf, u=%lf fp=%lf fu=%lf < or >= fc=%lf cx=%lf, q=%lf < %lf=r \n",*ax, *fa, *bx,*fb,u,fparabu,fu,*fc,*cx,q,r);
1.183 brouard 1930: #endif
1.191 brouard 1931: dum=u; /* Shifting c and u */
1932: u = *cx;
1933: *cx = dum;
1934: dum = fu;
1935: fu = *fc;
1936: *fc =dum;
1.183 brouard 1937: #endif
1.162 brouard 1938: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1939: #ifdef DEBUG
1.224 brouard 1940: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1941: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1942: #endif
1.126 brouard 1943: fu=(*func)(u);
1944: if (fu < *fc) {
1.183 brouard 1945: #ifdef DEBUG
1.224 brouard 1946: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1947: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1948: #endif
1949: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1950: SHFT(*fb,*fc,fu,(*func)(u))
1951: #ifdef DEBUG
1952: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1953: #endif
1954: }
1.162 brouard 1955: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1956: #ifdef DEBUG
1.224 brouard 1957: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1958: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1959: #endif
1.126 brouard 1960: u=ulim;
1961: fu=(*func)(u);
1.183 brouard 1962: } else { /* u could be left to b (if r > q parabola has a maximum) */
1963: #ifdef DEBUG
1.224 brouard 1964: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1965: fprintf(ficlog,"\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1.183 brouard 1966: #endif
1.126 brouard 1967: u=(*cx)+GOLD*(*cx-*bx);
1968: fu=(*func)(u);
1.224 brouard 1969: #ifdef DEBUG
1970: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1971: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1972: #endif
1.183 brouard 1973: } /* end tests */
1.126 brouard 1974: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1975: SHFT(*fa,*fb,*fc,fu)
1976: #ifdef DEBUG
1.224 brouard 1977: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1978: fprintf(ficlog, "\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1.183 brouard 1979: #endif
1980: } /* end while; ie return (a, b, c, fa, fb, fc) such that a < b < c with f(a) > f(b) and fb < f(c) */
1.126 brouard 1981: }
1982:
1983: /*************** linmin ************************/
1.162 brouard 1984: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1985: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1986: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1987: the value of func at the returned location p . This is actually all accomplished by calling the
1988: routines mnbrak and brent .*/
1.126 brouard 1989: int ncom;
1990: double *pcom,*xicom;
1991: double (*nrfunc)(double []);
1992:
1.224 brouard 1993: #ifdef LINMINORIGINAL
1.126 brouard 1994: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1995: #else
1996: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1997: #endif
1.126 brouard 1998: {
1999: double brent(double ax, double bx, double cx,
2000: double (*f)(double), double tol, double *xmin);
2001: double f1dim(double x);
2002: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
2003: double *fc, double (*func)(double));
2004: int j;
2005: double xx,xmin,bx,ax;
2006: double fx,fb,fa;
1.187 brouard 2007:
1.203 brouard 2008: #ifdef LINMINORIGINAL
2009: #else
2010: double scale=10., axs, xxs; /* Scale added for infinity */
2011: #endif
2012:
1.126 brouard 2013: ncom=n;
2014: pcom=vector(1,n);
2015: xicom=vector(1,n);
2016: nrfunc=func;
2017: for (j=1;j<=n;j++) {
2018: pcom[j]=p[j];
1.202 brouard 2019: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 2020: }
1.187 brouard 2021:
1.203 brouard 2022: #ifdef LINMINORIGINAL
2023: xx=1.;
2024: #else
2025: axs=0.0;
2026: xxs=1.;
2027: do{
2028: xx= xxs;
2029: #endif
1.187 brouard 2030: ax=0.;
2031: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
2032: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
2033: /* xt[x,j]=pcom[j]+x*xicom[j] f(ax) = f(xt(a,j=1,n)) = f(p(j) + 0 * xi(j)) and f(xx) = f(xt(x, j=1,n)) = f(p(j) + 1 * xi(j)) */
2034: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
2035: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
2036: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
2037: /* Find a bracket a,x,b in direction n=xi ie xicom, order may change. Scale is [0:xxs*xi[j]] et non plus [0:xi[j]]*/
1.203 brouard 2038: #ifdef LINMINORIGINAL
2039: #else
2040: if (fx != fx){
1.224 brouard 2041: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
2042: printf("|");
2043: fprintf(ficlog,"|");
1.203 brouard 2044: #ifdef DEBUGLINMIN
1.224 brouard 2045: printf("\nLinmin NAN : input [axs=%lf:xxs=%lf], mnbrak outputs fx=%lf <(fb=%lf and fa=%lf) with xx=%lf in [ax=%lf:bx=%lf] \n", axs, xxs, fx,fb, fa, xx, ax, bx);
1.203 brouard 2046: #endif
2047: }
1.224 brouard 2048: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 2049: #endif
2050:
1.191 brouard 2051: #ifdef DEBUGLINMIN
2052: printf("\nLinmin after mnbrak: ax=%12.7f xx=%12.7f bx=%12.7f fa=%12.2f fx=%12.2f fb=%12.2f\n", ax,xx,bx,fa,fx,fb);
1.202 brouard 2053: fprintf(ficlog,"\nLinmin after mnbrak: ax=%12.7f xx=%12.7f bx=%12.7f fa=%12.2f fx=%12.2f fb=%12.2f\n", ax,xx,bx,fa,fx,fb);
1.191 brouard 2054: #endif
1.224 brouard 2055: #ifdef LINMINORIGINAL
2056: #else
2057: if(fb == fx){ /* Flat function in the direction */
2058: xmin=xx;
2059: *flat=1;
2060: }else{
2061: *flat=0;
2062: #endif
2063: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 2064: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
2065: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
2066: /* fmin = f(p[j] + xmin * xi[j]) */
2067: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
2068: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 2069: #ifdef DEBUG
1.224 brouard 2070: printf("retour brent from bracket (a=%lf fa=%lf, xx=%lf fx=%lf, b=%lf fb=%lf): fret=%lf xmin=%lf\n",ax,fa,xx,fx,bx,fb,*fret,xmin);
2071: fprintf(ficlog,"retour brent from bracket (a=%lf fa=%lf, xx=%lf fx=%lf, b=%lf fb=%lf): fret=%lf xmin=%lf\n",ax,fa,xx,fx,bx,fb,*fret,xmin);
2072: #endif
2073: #ifdef LINMINORIGINAL
2074: #else
2075: }
1.126 brouard 2076: #endif
1.191 brouard 2077: #ifdef DEBUGLINMIN
2078: printf("linmin end ");
1.202 brouard 2079: fprintf(ficlog,"linmin end ");
1.191 brouard 2080: #endif
1.126 brouard 2081: for (j=1;j<=n;j++) {
1.203 brouard 2082: #ifdef LINMINORIGINAL
2083: xi[j] *= xmin;
2084: #else
2085: #ifdef DEBUGLINMIN
2086: if(xxs <1.0)
2087: printf(" before xi[%d]=%12.8f", j,xi[j]);
2088: #endif
2089: xi[j] *= xmin*xxs; /* xi rescaled by xmin and number of loops: if xmin=-1.237 and xi=(1,0,...,0) xi=(-1.237,0,...,0) */
2090: #ifdef DEBUGLINMIN
2091: if(xxs <1.0)
2092: printf(" after xi[%d]=%12.8f, xmin=%12.8f, ax=%12.8f, xx=%12.8f, bx=%12.8f, xxs=%12.8f", j,xi[j], xmin, ax, xx, bx,xxs );
2093: #endif
2094: #endif
1.187 brouard 2095: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2096: }
1.191 brouard 2097: #ifdef DEBUGLINMIN
1.203 brouard 2098: printf("\n");
1.191 brouard 2099: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2100: fprintf(ficlog,"Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.191 brouard 2101: for (j=1;j<=n;j++) {
1.202 brouard 2102: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2103: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2104: if(j % ncovmodel == 0){
1.191 brouard 2105: printf("\n");
1.202 brouard 2106: fprintf(ficlog,"\n");
2107: }
1.191 brouard 2108: }
1.203 brouard 2109: #else
1.191 brouard 2110: #endif
1.126 brouard 2111: free_vector(xicom,1,n);
2112: free_vector(pcom,1,n);
2113: }
2114:
2115:
2116: /*************** powell ************************/
1.162 brouard 2117: /*
2118: Minimization of a function func of n variables. Input consists of an initial starting point
2119: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2120: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2121: such that failure to decrease by more than this amount on one iteration signals doneness. On
2122: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2123: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2124: */
1.224 brouard 2125: #ifdef LINMINORIGINAL
2126: #else
2127: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2128: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2129: #endif
1.126 brouard 2130: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2131: double (*func)(double []))
2132: {
1.224 brouard 2133: #ifdef LINMINORIGINAL
2134: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2135: double (*func)(double []));
1.224 brouard 2136: #else
1.241 brouard 2137: void linmin(double p[], double xi[], int n, double *fret,
2138: double (*func)(double []),int *flat);
1.224 brouard 2139: #endif
1.239 brouard 2140: int i,ibig,j,jk,k;
1.126 brouard 2141: double del,t,*pt,*ptt,*xit;
1.181 brouard 2142: double directest;
1.126 brouard 2143: double fp,fptt;
2144: double *xits;
2145: int niterf, itmp;
1.224 brouard 2146: #ifdef LINMINORIGINAL
2147: #else
2148:
2149: flatdir=ivector(1,n);
2150: for (j=1;j<=n;j++) flatdir[j]=0;
2151: #endif
1.126 brouard 2152:
2153: pt=vector(1,n);
2154: ptt=vector(1,n);
2155: xit=vector(1,n);
2156: xits=vector(1,n);
2157: *fret=(*func)(p);
2158: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2159: rcurr_time = time(NULL);
1.126 brouard 2160: for (*iter=1;;++(*iter)) {
1.187 brouard 2161: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2162: ibig=0;
2163: del=0.0;
1.157 brouard 2164: rlast_time=rcurr_time;
2165: /* (void) gettimeofday(&curr_time,&tzp); */
2166: rcurr_time = time(NULL);
2167: curr_time = *localtime(&rcurr_time);
2168: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2169: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2170: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2171: for (i=1;i<=n;i++) {
1.126 brouard 2172: fprintf(ficrespow," %.12lf", p[i]);
2173: }
1.239 brouard 2174: fprintf(ficrespow,"\n");fflush(ficrespow);
2175: printf("\n#model= 1 + age ");
2176: fprintf(ficlog,"\n#model= 1 + age ");
2177: if(nagesqr==1){
1.241 brouard 2178: printf(" + age*age ");
2179: fprintf(ficlog," + age*age ");
1.239 brouard 2180: }
2181: for(j=1;j <=ncovmodel-2;j++){
2182: if(Typevar[j]==0) {
2183: printf(" + V%d ",Tvar[j]);
2184: fprintf(ficlog," + V%d ",Tvar[j]);
2185: }else if(Typevar[j]==1) {
2186: printf(" + V%d*age ",Tvar[j]);
2187: fprintf(ficlog," + V%d*age ",Tvar[j]);
2188: }else if(Typevar[j]==2) {
2189: printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2190: fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2191: }
2192: }
1.126 brouard 2193: printf("\n");
1.239 brouard 2194: /* printf("12 47.0114589 0.0154322 33.2424412 0.3279905 2.3731903 */
2195: /* 13 -21.5392400 0.1118147 1.2680506 1.2973408 -1.0663662 */
1.126 brouard 2196: fprintf(ficlog,"\n");
1.239 brouard 2197: for(i=1,jk=1; i <=nlstate; i++){
2198: for(k=1; k <=(nlstate+ndeath); k++){
2199: if (k != i) {
2200: printf("%d%d ",i,k);
2201: fprintf(ficlog,"%d%d ",i,k);
2202: for(j=1; j <=ncovmodel; j++){
2203: printf("%12.7f ",p[jk]);
2204: fprintf(ficlog,"%12.7f ",p[jk]);
2205: jk++;
2206: }
2207: printf("\n");
2208: fprintf(ficlog,"\n");
2209: }
2210: }
2211: }
1.241 brouard 2212: if(*iter <=3 && *iter >1){
1.157 brouard 2213: tml = *localtime(&rcurr_time);
2214: strcpy(strcurr,asctime(&tml));
2215: rforecast_time=rcurr_time;
1.126 brouard 2216: itmp = strlen(strcurr);
2217: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.241 brouard 2218: strcurr[itmp-1]='\0';
1.162 brouard 2219: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2220: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2221: for(niterf=10;niterf<=30;niterf+=10){
1.241 brouard 2222: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2223: forecast_time = *localtime(&rforecast_time);
2224: strcpy(strfor,asctime(&forecast_time));
2225: itmp = strlen(strfor);
2226: if(strfor[itmp-1]=='\n')
2227: strfor[itmp-1]='\0';
2228: printf(" - if your program needs %d iterations to converge, convergence will be \n reached in %s i.e.\n on %s (current time is %s);\n",niterf, asc_diff_time(rforecast_time-rcurr_time,tmpout),strfor,strcurr);
2229: fprintf(ficlog," - if your program needs %d iterations to converge, convergence will be \n reached in %s i.e.\n on %s (current time is %s);\n",niterf, asc_diff_time(rforecast_time-rcurr_time,tmpout),strfor,strcurr);
1.126 brouard 2230: }
2231: }
1.187 brouard 2232: for (i=1;i<=n;i++) { /* For each direction i */
2233: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2234: fptt=(*fret);
2235: #ifdef DEBUG
1.203 brouard 2236: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2237: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2238: #endif
1.203 brouard 2239: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2240: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2241: #ifdef LINMINORIGINAL
1.188 brouard 2242: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2243: #else
2244: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2245: flatdir[i]=flat; /* Function is vanishing in that direction i */
2246: #endif
2247: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2248: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2249: /* because that direction will be replaced unless the gain del is small */
2250: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2251: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2252: /* with the new direction. */
2253: del=fabs(fptt-(*fret));
2254: ibig=i;
1.126 brouard 2255: }
2256: #ifdef DEBUG
2257: printf("%d %.12e",i,(*fret));
2258: fprintf(ficlog,"%d %.12e",i,(*fret));
2259: for (j=1;j<=n;j++) {
1.224 brouard 2260: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2261: printf(" x(%d)=%.12e",j,xit[j]);
2262: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2263: }
2264: for(j=1;j<=n;j++) {
1.225 brouard 2265: printf(" p(%d)=%.12e",j,p[j]);
2266: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2267: }
2268: printf("\n");
2269: fprintf(ficlog,"\n");
2270: #endif
1.187 brouard 2271: } /* end loop on each direction i */
2272: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2273: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2274: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2275: for(j=1;j<=n;j++) {
1.225 brouard 2276: if(flatdir[j] >0){
2277: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2278: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2279: }
2280: /* printf("\n"); */
2281: /* fprintf(ficlog,"\n"); */
2282: }
1.243 brouard 2283: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */
2284: if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */
1.188 brouard 2285: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2286: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2287: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2288: /* decreased of more than 3.84 */
2289: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2290: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2291: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2292:
1.188 brouard 2293: /* Starting the program with initial values given by a former maximization will simply change */
2294: /* the scales of the directions and the directions, because the are reset to canonical directions */
2295: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2296: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2297: #ifdef DEBUG
2298: int k[2],l;
2299: k[0]=1;
2300: k[1]=-1;
2301: printf("Max: %.12e",(*func)(p));
2302: fprintf(ficlog,"Max: %.12e",(*func)(p));
2303: for (j=1;j<=n;j++) {
2304: printf(" %.12e",p[j]);
2305: fprintf(ficlog," %.12e",p[j]);
2306: }
2307: printf("\n");
2308: fprintf(ficlog,"\n");
2309: for(l=0;l<=1;l++) {
2310: for (j=1;j<=n;j++) {
2311: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2312: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2313: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2314: }
2315: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2316: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2317: }
2318: #endif
2319:
1.224 brouard 2320: #ifdef LINMINORIGINAL
2321: #else
2322: free_ivector(flatdir,1,n);
2323: #endif
1.126 brouard 2324: free_vector(xit,1,n);
2325: free_vector(xits,1,n);
2326: free_vector(ptt,1,n);
2327: free_vector(pt,1,n);
2328: return;
1.192 brouard 2329: } /* enough precision */
1.240 brouard 2330: if (*iter == ITMAX*n) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2331: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2332: ptt[j]=2.0*p[j]-pt[j];
2333: xit[j]=p[j]-pt[j];
2334: pt[j]=p[j];
2335: }
1.181 brouard 2336: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2337: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2338: if (*iter <=4) {
1.225 brouard 2339: #else
2340: #endif
1.224 brouard 2341: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2342: #else
1.161 brouard 2343: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2344: #endif
1.162 brouard 2345: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2346: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2347: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2348: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2349: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2350: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2351: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2352: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2353: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2354: /* Even if f3 <f1, directest can be negative and t >0 */
2355: /* mu² and del² are equal when f3=f1 */
2356: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2357: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2358: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2359: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2360: #ifdef NRCORIGINAL
2361: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2362: #else
2363: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del); /* Intel compiler doesn't work on one line; bug reported */
1.161 brouard 2364: t= t- del*SQR(fp-fptt);
1.183 brouard 2365: #endif
1.202 brouard 2366: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2367: #ifdef DEBUG
1.181 brouard 2368: printf("t1= %.12lf, t2= %.12lf, t=%.12lf directest=%.12lf\n", 2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del),del*SQR(fp-fptt),t,directest);
2369: fprintf(ficlog,"t1= %.12lf, t2= %.12lf, t=%.12lf directest=%.12lf\n", 2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del),del*SQR(fp-fptt),t,directest);
1.161 brouard 2370: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2371: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2372: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2373: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2374: printf("tt= %.12lf, t=%.12lf\n",2.0*(fp-2.0*(*fret)+fptt)*(fp-(*fret)-del)*(fp-(*fret)-del)-del*(fp-fptt)*(fp-fptt),t);
2375: fprintf(ficlog, "tt= %.12lf, t=%.12lf\n",2.0*(fp-2.0*(*fret)+fptt)*(fp-(*fret)-del)*(fp-(*fret)-del)-del*(fp-fptt)*(fp-fptt),t);
2376: #endif
1.183 brouard 2377: #ifdef POWELLORIGINAL
2378: if (t < 0.0) { /* Then we use it for new direction */
2379: #else
1.182 brouard 2380: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2381: printf("directest= %.12lf (if <0 we include P0 Pn as new direction), t= %.12lf, f1= %.12lf,f2= %.12lf,f3= %.12lf, del= %.12lf\n",directest, t, fp,(*fret),fptt,del);
1.192 brouard 2382: printf("f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
1.224 brouard 2383: fprintf(ficlog,"directest= %.12lf (if directest<0 or t<0 we include P0 Pn as new direction), t= %.12lf, f1= %.12lf,f2= %.12lf,f3= %.12lf, del= %.12lf\n",directest, t, fp,(*fret),fptt, del);
1.192 brouard 2384: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2385: }
1.181 brouard 2386: if (directest < 0.0) { /* Then we use it for new direction */
2387: #endif
1.191 brouard 2388: #ifdef DEBUGLINMIN
1.234 brouard 2389: printf("Before linmin in direction P%d-P0\n",n);
2390: for (j=1;j<=n;j++) {
2391: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2392: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2393: if(j % ncovmodel == 0){
2394: printf("\n");
2395: fprintf(ficlog,"\n");
2396: }
2397: }
1.224 brouard 2398: #endif
2399: #ifdef LINMINORIGINAL
1.234 brouard 2400: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2401: #else
1.234 brouard 2402: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2403: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2404: #endif
1.234 brouard 2405:
1.191 brouard 2406: #ifdef DEBUGLINMIN
1.234 brouard 2407: for (j=1;j<=n;j++) {
2408: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2409: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2410: if(j % ncovmodel == 0){
2411: printf("\n");
2412: fprintf(ficlog,"\n");
2413: }
2414: }
1.224 brouard 2415: #endif
1.234 brouard 2416: for (j=1;j<=n;j++) {
2417: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2418: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2419: }
1.224 brouard 2420: #ifdef LINMINORIGINAL
2421: #else
1.234 brouard 2422: for (j=1, flatd=0;j<=n;j++) {
2423: if(flatdir[j]>0)
2424: flatd++;
2425: }
2426: if(flatd >0){
1.255 brouard 2427: printf("%d flat directions: ",flatd);
2428: fprintf(ficlog,"%d flat directions :",flatd);
1.234 brouard 2429: for (j=1;j<=n;j++) {
2430: if(flatdir[j]>0){
2431: printf("%d ",j);
2432: fprintf(ficlog,"%d ",j);
2433: }
2434: }
2435: printf("\n");
2436: fprintf(ficlog,"\n");
2437: }
1.191 brouard 2438: #endif
1.234 brouard 2439: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2440: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2441:
1.126 brouard 2442: #ifdef DEBUG
1.234 brouard 2443: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2444: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2445: for(j=1;j<=n;j++){
2446: printf(" %lf",xit[j]);
2447: fprintf(ficlog," %lf",xit[j]);
2448: }
2449: printf("\n");
2450: fprintf(ficlog,"\n");
1.126 brouard 2451: #endif
1.192 brouard 2452: } /* end of t or directest negative */
1.224 brouard 2453: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2454: #else
1.234 brouard 2455: } /* end if (fptt < fp) */
1.192 brouard 2456: #endif
1.225 brouard 2457: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2458: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2459: #else
1.224 brouard 2460: #endif
1.234 brouard 2461: } /* loop iteration */
1.126 brouard 2462: }
1.234 brouard 2463:
1.126 brouard 2464: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2465:
1.235 brouard 2466: double **prevalim(double **prlim, int nlstate, double x[], double age, double **oldm, double **savm, double ftolpl, int *ncvyear, int ij, int nres)
1.234 brouard 2467: {
1.235 brouard 2468: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2469: (and selected quantitative values in nres)
2470: by left multiplying the unit
1.234 brouard 2471: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2472: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2473: /* Wx is row vector: population in state 1, population in state 2, population dead */
2474: /* or prevalence in state 1, prevalence in state 2, 0 */
2475: /* newm is the matrix after multiplications, its rows are identical at a factor */
2476: /* Initial matrix pimij */
2477: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2478: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2479: /* 0, 0 , 1} */
2480: /*
2481: * and after some iteration: */
2482: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2483: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2484: /* 0, 0 , 1} */
2485: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2486: /* {0.51571254859325999, 0.4842874514067399, */
2487: /* 0.51326036147820708, 0.48673963852179264} */
2488: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2489:
1.126 brouard 2490: int i, ii,j,k;
1.209 brouard 2491: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2492: /* double **matprod2(); */ /* test */
1.218 brouard 2493: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2494: double **newm;
1.209 brouard 2495: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2496: int ncvloop=0;
1.169 brouard 2497:
1.209 brouard 2498: min=vector(1,nlstate);
2499: max=vector(1,nlstate);
2500: meandiff=vector(1,nlstate);
2501:
1.218 brouard 2502: /* Starting with matrix unity */
1.126 brouard 2503: for (ii=1;ii<=nlstate+ndeath;ii++)
2504: for (j=1;j<=nlstate+ndeath;j++){
2505: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2506: }
1.169 brouard 2507:
2508: cov[1]=1.;
2509:
2510: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2511: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2512: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2513: ncvloop++;
1.126 brouard 2514: newm=savm;
2515: /* Covariates have to be included here again */
1.138 brouard 2516: cov[2]=agefin;
1.187 brouard 2517: if(nagesqr==1)
2518: cov[3]= agefin*agefin;;
1.234 brouard 2519: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2520: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2521: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2522: /* printf("prevalim Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
1.234 brouard 2523: }
2524: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2525: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2526: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2527: /* printf("prevalim Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
1.138 brouard 2528: }
1.237 brouard 2529: for (k=1; k<=cptcovage;k++){ /* For product with age */
1.234 brouard 2530: if(Dummy[Tvar[Tage[k]]]){
2531: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2532: } else{
1.235 brouard 2533: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2534: }
1.235 brouard 2535: /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
1.234 brouard 2536: }
1.237 brouard 2537: for (k=1; k<=cptcovprod;k++){ /* For product without age */
1.235 brouard 2538: /* printf("prevalim Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
1.237 brouard 2539: if(Dummy[Tvard[k][1]==0]){
2540: if(Dummy[Tvard[k][2]==0]){
2541: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2542: }else{
2543: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2544: }
2545: }else{
2546: if(Dummy[Tvard[k][2]==0]){
2547: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2548: }else{
2549: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2550: }
2551: }
1.234 brouard 2552: }
1.138 brouard 2553: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2554: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2555: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2556: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2557: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2558: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2559: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2560:
1.126 brouard 2561: savm=oldm;
2562: oldm=newm;
1.209 brouard 2563:
2564: for(j=1; j<=nlstate; j++){
2565: max[j]=0.;
2566: min[j]=1.;
2567: }
2568: for(i=1;i<=nlstate;i++){
2569: sumnew=0;
2570: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2571: for(j=1; j<=nlstate; j++){
2572: prlim[i][j]= newm[i][j]/(1-sumnew);
2573: max[j]=FMAX(max[j],prlim[i][j]);
2574: min[j]=FMIN(min[j],prlim[i][j]);
2575: }
2576: }
2577:
1.126 brouard 2578: maxmax=0.;
1.209 brouard 2579: for(j=1; j<=nlstate; j++){
2580: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2581: maxmax=FMAX(maxmax,meandiff[j]);
2582: /* printf(" age= %d meandiff[%d]=%f, agefin=%d max[%d]=%f min[%d]=%f maxmax=%f\n", (int)age, j, meandiff[j],(int)agefin, j, max[j], j, min[j],maxmax); */
1.169 brouard 2583: } /* j loop */
1.203 brouard 2584: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2585: /* printf("maxmax=%lf maxmin=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, maxmin, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.126 brouard 2586: if(maxmax < ftolpl){
1.209 brouard 2587: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2588: free_vector(min,1,nlstate);
2589: free_vector(max,1,nlstate);
2590: free_vector(meandiff,1,nlstate);
1.126 brouard 2591: return prlim;
2592: }
1.169 brouard 2593: } /* age loop */
1.208 brouard 2594: /* After some age loop it doesn't converge */
1.209 brouard 2595: printf("Warning: the stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. \n\
1.208 brouard 2596: Earliest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
1.209 brouard 2597: /* Try to lower 'ftol', for example from 1.e-8 to 6.e-9.\n", ftolpl, (int)age, (int)delaymax, (int)agefin, ncvloop, (int)age-(int)agefin); */
2598: free_vector(min,1,nlstate);
2599: free_vector(max,1,nlstate);
2600: free_vector(meandiff,1,nlstate);
1.208 brouard 2601:
1.169 brouard 2602: return prlim; /* should not reach here */
1.126 brouard 2603: }
2604:
1.217 brouard 2605:
2606: /**** Back Prevalence limit (stable or period prevalence) ****************/
2607:
1.218 brouard 2608: /* double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ageminpar, double agemaxpar, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, double ftolpl, int *ncvyear, int ij) */
2609: /* double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, double ftolpl, int *ncvyear, int ij) */
1.242 brouard 2610: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij, int nres)
1.217 brouard 2611: {
1.218 brouard 2612: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2613: matrix by transitions matrix until convergence is reached with precision ftolpl */
2614: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2615: /* Wx is row vector: population in state 1, population in state 2, population dead */
2616: /* or prevalence in state 1, prevalence in state 2, 0 */
2617: /* newm is the matrix after multiplications, its rows are identical at a factor */
2618: /* Initial matrix pimij */
2619: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2620: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2621: /* 0, 0 , 1} */
2622: /*
2623: * and after some iteration: */
2624: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2625: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2626: /* 0, 0 , 1} */
2627: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2628: /* {0.51571254859325999, 0.4842874514067399, */
2629: /* 0.51326036147820708, 0.48673963852179264} */
2630: /* If we start from prlim again, prlim tends to a constant matrix */
2631:
2632: int i, ii,j,k;
1.247 brouard 2633: int first=0;
1.217 brouard 2634: double *min, *max, *meandiff, maxmax,sumnew=0.;
2635: /* double **matprod2(); */ /* test */
2636: double **out, cov[NCOVMAX+1], **bmij();
2637: double **newm;
1.218 brouard 2638: double **dnewm, **doldm, **dsavm; /* for use */
2639: double **oldm, **savm; /* for use */
2640:
1.217 brouard 2641: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2642: int ncvloop=0;
2643:
2644: min=vector(1,nlstate);
2645: max=vector(1,nlstate);
2646: meandiff=vector(1,nlstate);
2647:
1.218 brouard 2648: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2649: oldm=oldms; savm=savms;
2650:
2651: /* Starting with matrix unity */
2652: for (ii=1;ii<=nlstate+ndeath;ii++)
2653: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2654: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2655: }
2656:
2657: cov[1]=1.;
2658:
2659: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2660: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2661: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2662: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2663: ncvloop++;
1.218 brouard 2664: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2665: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2666: /* Covariates have to be included here again */
2667: cov[2]=agefin;
2668: if(nagesqr==1)
2669: cov[3]= agefin*agefin;;
1.242 brouard 2670: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2671: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2672: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
2673: /* printf("bprevalim Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
2674: }
2675: /* for (k=1; k<=cptcovn;k++) { */
2676: /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */
2677: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2678: /* /\* printf("prevalim ij=%d k=%d Tvar[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, Tvar[k],nbcode[Tvar[k]][codtabm(ij,Tvar[k])],cov[2+k], ij, k, codtabm(ij,Tvar[k])]); *\/ */
2679: /* } */
2680: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2681: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2682: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2683: /* printf("prevalim Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
2684: }
2685: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2]; */
2686: /* for (k=1; k<=cptcovprod;k++) /\* Useless *\/ */
2687: /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */
2688: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */
2689: for (k=1; k<=cptcovage;k++){ /* For product with age */
2690: if(Dummy[Tvar[Tage[k]]]){
2691: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2692: } else{
2693: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2694: }
2695: /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
2696: }
2697: for (k=1; k<=cptcovprod;k++){ /* For product without age */
2698: /* printf("prevalim Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
2699: if(Dummy[Tvard[k][1]==0]){
2700: if(Dummy[Tvard[k][2]==0]){
2701: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2702: }else{
2703: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2704: }
2705: }else{
2706: if(Dummy[Tvard[k][2]==0]){
2707: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2708: }else{
2709: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2710: }
2711: }
1.217 brouard 2712: }
2713:
2714: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2715: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2716: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2717: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2718: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2719: /* ij should be linked to the correct index of cov */
2720: /* age and covariate values ij are in 'cov', but we need to pass
2721: * ij for the observed prevalence at age and status and covariate
2722: * number: prevacurrent[(int)agefin][ii][ij]
2723: */
2724: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, ageminpar, agemaxpar, dnewm, doldm, dsavm,ij)); /\* Bug Valgrind *\/ */
2725: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij)); /\* Bug Valgrind *\/ */
2726: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij)); /* Bug Valgrind */
1.217 brouard 2727: savm=oldm;
2728: oldm=newm;
2729: for(j=1; j<=nlstate; j++){
2730: max[j]=0.;
2731: min[j]=1.;
2732: }
2733: for(j=1; j<=nlstate; j++){
2734: for(i=1;i<=nlstate;i++){
1.234 brouard 2735: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2736: bprlim[i][j]= newm[i][j];
2737: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2738: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2739: }
2740: }
1.218 brouard 2741:
1.217 brouard 2742: maxmax=0.;
2743: for(i=1; i<=nlstate; i++){
2744: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2745: maxmax=FMAX(maxmax,meandiff[i]);
2746: /* printf("Back age= %d meandiff[%d]=%f, agefin=%d max[%d]=%f min[%d]=%f maxmax=%f\n", (int)age, i, meandiff[i],(int)agefin, i, max[i], i, min[i],maxmax); */
2747: } /* j loop */
2748: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2749: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2750: if(maxmax < ftolpl){
1.220 brouard 2751: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2752: free_vector(min,1,nlstate);
2753: free_vector(max,1,nlstate);
2754: free_vector(meandiff,1,nlstate);
2755: return bprlim;
2756: }
2757: } /* age loop */
2758: /* After some age loop it doesn't converge */
1.247 brouard 2759: if(first){
2760: first=1;
2761: printf("Warning: the back stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. Others in log file only...\n\
2762: Oldest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
2763: }
2764: fprintf(ficlog,"Warning: the back stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. \n\
1.217 brouard 2765: Oldest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
2766: /* Try to lower 'ftol', for example from 1.e-8 to 6.e-9.\n", ftolpl, (int)age, (int)delaymax, (int)agefin, ncvloop, (int)age-(int)agefin); */
2767: free_vector(min,1,nlstate);
2768: free_vector(max,1,nlstate);
2769: free_vector(meandiff,1,nlstate);
2770:
2771: return bprlim; /* should not reach here */
2772: }
2773:
1.126 brouard 2774: /*************** transition probabilities ***************/
2775:
2776: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2777: {
1.138 brouard 2778: /* According to parameters values stored in x and the covariate's values stored in cov,
2779: computes the probability to be observed in state j being in state i by appying the
2780: model to the ncovmodel covariates (including constant and age).
2781: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2782: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2783: ncth covariate in the global vector x is given by the formula:
2784: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2785: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2786: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2787: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2788: Outputs ps[i][j] the probability to be observed in j being in j according to
2789: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2790: */
2791: double s1, lnpijopii;
1.126 brouard 2792: /*double t34;*/
1.164 brouard 2793: int i,j, nc, ii, jj;
1.126 brouard 2794:
1.223 brouard 2795: for(i=1; i<= nlstate; i++){
2796: for(j=1; j<i;j++){
2797: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2798: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2799: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2800: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2801: }
2802: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2803: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2804: }
2805: for(j=i+1; j<=nlstate+ndeath;j++){
2806: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2807: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2808: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2809: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2810: }
2811: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2812: }
2813: }
1.218 brouard 2814:
1.223 brouard 2815: for(i=1; i<= nlstate; i++){
2816: s1=0;
2817: for(j=1; j<i; j++){
2818: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2819: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2820: }
2821: for(j=i+1; j<=nlstate+ndeath; j++){
2822: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2823: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2824: }
2825: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2826: ps[i][i]=1./(s1+1.);
2827: /* Computing other pijs */
2828: for(j=1; j<i; j++)
2829: ps[i][j]= exp(ps[i][j])*ps[i][i];
2830: for(j=i+1; j<=nlstate+ndeath; j++)
2831: ps[i][j]= exp(ps[i][j])*ps[i][i];
2832: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2833: } /* end i */
1.218 brouard 2834:
1.223 brouard 2835: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2836: for(jj=1; jj<= nlstate+ndeath; jj++){
2837: ps[ii][jj]=0;
2838: ps[ii][ii]=1;
2839: }
2840: }
1.218 brouard 2841:
2842:
1.223 brouard 2843: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2844: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2845: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2846: /* } */
2847: /* printf("\n "); */
2848: /* } */
2849: /* printf("\n ");printf("%lf ",cov[2]);*/
2850: /*
2851: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2852: goto end;*/
1.223 brouard 2853: return ps;
1.126 brouard 2854: }
2855:
1.218 brouard 2856: /*************** backward transition probabilities ***************/
2857:
2858: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ageminpar, double agemaxpar, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2859: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2860: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2861: {
1.222 brouard 2862: /* Computes the backward probability at age agefin and covariate ij
2863: * and returns in **ps as well as **bmij.
2864: */
1.218 brouard 2865: int i, ii, j,k;
1.222 brouard 2866:
2867: double **out, **pmij();
2868: double sumnew=0.;
1.218 brouard 2869: double agefin;
1.222 brouard 2870:
2871: double **dnewm, **dsavm, **doldm;
2872: double **bbmij;
2873:
1.218 brouard 2874: doldm=ddoldms; /* global pointers */
1.222 brouard 2875: dnewm=ddnewms;
2876: dsavm=ddsavms;
2877:
2878: agefin=cov[2];
2879: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2880: the observed prevalence (with this covariate ij) */
2881: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2882: /* We do have the matrix Px in savm and we need pij */
2883: for (j=1;j<=nlstate+ndeath;j++){
2884: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2885: for (ii=1;ii<=nlstate;ii++){
2886: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2887: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2888: for (ii=1;ii<=nlstate+ndeath;ii++){
2889: if(sumnew >= 1.e-10){
2890: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2891: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2892: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2893: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2894: /* }else */
2895: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2896: }else{
1.242 brouard 2897: ;
2898: /* printf("ii=%d, i=%d, doldm=%lf dsavm=%lf, probs=%lf, sumnew=%lf,agefin=%d\n",ii,j,doldm[ii][j],dsavm[ii][j],prevacurrent[(int)agefin][ii][ij],sumnew, (int)agefin); */
1.222 brouard 2899: }
2900: } /*End ii */
2901: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2902: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2903: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2904: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2905: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2906: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2907: /* left Product of this matrix by diag matrix of prevalences (savm) */
2908: for (j=1;j<=nlstate+ndeath;j++){
2909: for (ii=1;ii<=nlstate+ndeath;ii++){
2910: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2911: }
2912: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2913: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2914: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2915: /* end bmij */
2916: return ps;
1.218 brouard 2917: }
1.217 brouard 2918: /*************** transition probabilities ***************/
2919:
1.218 brouard 2920: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2921: {
2922: /* According to parameters values stored in x and the covariate's values stored in cov,
2923: computes the probability to be observed in state j being in state i by appying the
2924: model to the ncovmodel covariates (including constant and age).
2925: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2926: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2927: ncth covariate in the global vector x is given by the formula:
2928: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2929: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2930: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2931: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2932: Outputs ps[i][j] the probability to be observed in j being in j according to
2933: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2934: */
2935: double s1, lnpijopii;
2936: /*double t34;*/
2937: int i,j, nc, ii, jj;
2938:
1.234 brouard 2939: for(i=1; i<= nlstate; i++){
2940: for(j=1; j<i;j++){
2941: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2942: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2943: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2944: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2945: }
2946: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2947: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2948: }
2949: for(j=i+1; j<=nlstate+ndeath;j++){
2950: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2951: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2952: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2953: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2954: }
2955: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2956: }
2957: }
2958:
2959: for(i=1; i<= nlstate; i++){
2960: s1=0;
2961: for(j=1; j<i; j++){
2962: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2963: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2964: }
2965: for(j=i+1; j<=nlstate+ndeath; j++){
2966: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2967: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2968: }
2969: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2970: ps[i][i]=1./(s1+1.);
2971: /* Computing other pijs */
2972: for(j=1; j<i; j++)
2973: ps[i][j]= exp(ps[i][j])*ps[i][i];
2974: for(j=i+1; j<=nlstate+ndeath; j++)
2975: ps[i][j]= exp(ps[i][j])*ps[i][i];
2976: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2977: } /* end i */
2978:
2979: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2980: for(jj=1; jj<= nlstate+ndeath; jj++){
2981: ps[ii][jj]=0;
2982: ps[ii][ii]=1;
2983: }
2984: }
2985: /* Added for backcast */ /* Transposed matrix too */
2986: for(jj=1; jj<= nlstate+ndeath; jj++){
2987: s1=0.;
2988: for(ii=1; ii<= nlstate+ndeath; ii++){
2989: s1+=ps[ii][jj];
2990: }
2991: for(ii=1; ii<= nlstate; ii++){
2992: ps[ii][jj]=ps[ii][jj]/s1;
2993: }
2994: }
2995: /* Transposition */
2996: for(jj=1; jj<= nlstate+ndeath; jj++){
2997: for(ii=jj; ii<= nlstate+ndeath; ii++){
2998: s1=ps[ii][jj];
2999: ps[ii][jj]=ps[jj][ii];
3000: ps[jj][ii]=s1;
3001: }
3002: }
3003: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
3004: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
3005: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
3006: /* } */
3007: /* printf("\n "); */
3008: /* } */
3009: /* printf("\n ");printf("%lf ",cov[2]);*/
3010: /*
3011: for(i=1; i<= npar; i++) printf("%f ",x[i]);
3012: goto end;*/
3013: return ps;
1.217 brouard 3014: }
3015:
3016:
1.126 brouard 3017: /**************** Product of 2 matrices ******************/
3018:
1.145 brouard 3019: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 3020: {
3021: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
3022: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
3023: /* in, b, out are matrice of pointers which should have been initialized
3024: before: only the contents of out is modified. The function returns
3025: a pointer to pointers identical to out */
1.145 brouard 3026: int i, j, k;
1.126 brouard 3027: for(i=nrl; i<= nrh; i++)
1.145 brouard 3028: for(k=ncolol; k<=ncoloh; k++){
3029: out[i][k]=0.;
3030: for(j=ncl; j<=nch; j++)
3031: out[i][k] +=in[i][j]*b[j][k];
3032: }
1.126 brouard 3033: return out;
3034: }
3035:
3036:
3037: /************* Higher Matrix Product ***************/
3038:
1.235 brouard 3039: double ***hpxij(double ***po, int nhstepm, double age, int hstepm, double *x, int nlstate, int stepm, double **oldm, double **savm, int ij, int nres )
1.126 brouard 3040: {
1.218 brouard 3041: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 3042: 'nhstepm*hstepm*stepm' months (i.e. until
3043: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3044: nhstepm*hstepm matrices.
3045: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3046: (typically every 2 years instead of every month which is too big
3047: for the memory).
3048: Model is determined by parameters x and covariates have to be
3049: included manually here.
3050:
3051: */
3052:
3053: int i, j, d, h, k;
1.131 brouard 3054: double **out, cov[NCOVMAX+1];
1.126 brouard 3055: double **newm;
1.187 brouard 3056: double agexact;
1.214 brouard 3057: double agebegin, ageend;
1.126 brouard 3058:
3059: /* Hstepm could be zero and should return the unit matrix */
3060: for (i=1;i<=nlstate+ndeath;i++)
3061: for (j=1;j<=nlstate+ndeath;j++){
3062: oldm[i][j]=(i==j ? 1.0 : 0.0);
3063: po[i][j][0]=(i==j ? 1.0 : 0.0);
3064: }
3065: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3066: for(h=1; h <=nhstepm; h++){
3067: for(d=1; d <=hstepm; d++){
3068: newm=savm;
3069: /* Covariates have to be included here again */
3070: cov[1]=1.;
1.214 brouard 3071: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 3072: cov[2]=agexact;
3073: if(nagesqr==1)
1.227 brouard 3074: cov[3]= agexact*agexact;
1.235 brouard 3075: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
3076: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
3077: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
3078: /* printf("hpxij Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
3079: }
3080: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
3081: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
3082: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
3083: /* printf("hPxij Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
3084: }
3085: for (k=1; k<=cptcovage;k++){
3086: if(Dummy[Tvar[Tage[k]]]){
3087: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3088: } else{
3089: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
3090: }
3091: /* printf("hPxij Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
3092: }
3093: for (k=1; k<=cptcovprod;k++){ /* */
3094: /* printf("hPxij Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
3095: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
3096: }
3097: /* for (k=1; k<=cptcovn;k++) */
3098: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
3099: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
3100: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
3101: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
3102: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 3103:
3104:
1.126 brouard 3105: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3106: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3107: /* right multiplication of oldm by the current matrix */
1.126 brouard 3108: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
3109: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 3110: /* if((int)age == 70){ */
3111: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3112: /* for(i=1; i<=nlstate+ndeath; i++) { */
3113: /* printf("%d pmmij ",i); */
3114: /* for(j=1;j<=nlstate+ndeath;j++) { */
3115: /* printf("%f ",pmmij[i][j]); */
3116: /* } */
3117: /* printf(" oldm "); */
3118: /* for(j=1;j<=nlstate+ndeath;j++) { */
3119: /* printf("%f ",oldm[i][j]); */
3120: /* } */
3121: /* printf("\n"); */
3122: /* } */
3123: /* } */
1.126 brouard 3124: savm=oldm;
3125: oldm=newm;
3126: }
3127: for(i=1; i<=nlstate+ndeath; i++)
3128: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 3129: po[i][j][h]=newm[i][j];
3130: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 3131: }
1.128 brouard 3132: /*printf("h=%d ",h);*/
1.126 brouard 3133: } /* end h */
1.218 brouard 3134: /* printf("\n H=%d \n",h); */
1.126 brouard 3135: return po;
3136: }
3137:
1.217 brouard 3138: /************* Higher Back Matrix Product ***************/
1.218 brouard 3139: /* double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, int ij ) */
1.222 brouard 3140: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 3141: {
1.218 brouard 3142: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 3143: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 3144: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3145: nhstepm*hstepm matrices.
3146: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3147: (typically every 2 years instead of every month which is too big
1.217 brouard 3148: for the memory).
1.218 brouard 3149: Model is determined by parameters x and covariates have to be
3150: included manually here.
1.217 brouard 3151:
1.222 brouard 3152: */
1.217 brouard 3153:
3154: int i, j, d, h, k;
3155: double **out, cov[NCOVMAX+1];
3156: double **newm;
3157: double agexact;
3158: double agebegin, ageend;
1.222 brouard 3159: double **oldm, **savm;
1.217 brouard 3160:
1.222 brouard 3161: oldm=oldms;savm=savms;
1.217 brouard 3162: /* Hstepm could be zero and should return the unit matrix */
3163: for (i=1;i<=nlstate+ndeath;i++)
3164: for (j=1;j<=nlstate+ndeath;j++){
3165: oldm[i][j]=(i==j ? 1.0 : 0.0);
3166: po[i][j][0]=(i==j ? 1.0 : 0.0);
3167: }
3168: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3169: for(h=1; h <=nhstepm; h++){
3170: for(d=1; d <=hstepm; d++){
3171: newm=savm;
3172: /* Covariates have to be included here again */
3173: cov[1]=1.;
3174: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3175: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3176: cov[2]=agexact;
3177: if(nagesqr==1)
1.222 brouard 3178: cov[3]= agexact*agexact;
1.218 brouard 3179: for (k=1; k<=cptcovn;k++)
1.222 brouard 3180: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3181: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3182: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3183: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3184: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3185: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3186: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3187: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3188: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])]*nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
1.218 brouard 3189:
3190:
1.217 brouard 3191: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3192: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3193: /* Careful transposed matrix */
1.222 brouard 3194: /* age is in cov[2] */
1.218 brouard 3195: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3196: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3197: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3198: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3199: /* if((int)age == 70){ */
3200: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3201: /* for(i=1; i<=nlstate+ndeath; i++) { */
3202: /* printf("%d pmmij ",i); */
3203: /* for(j=1;j<=nlstate+ndeath;j++) { */
3204: /* printf("%f ",pmmij[i][j]); */
3205: /* } */
3206: /* printf(" oldm "); */
3207: /* for(j=1;j<=nlstate+ndeath;j++) { */
3208: /* printf("%f ",oldm[i][j]); */
3209: /* } */
3210: /* printf("\n"); */
3211: /* } */
3212: /* } */
3213: savm=oldm;
3214: oldm=newm;
3215: }
3216: for(i=1; i<=nlstate+ndeath; i++)
3217: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3218: po[i][j][h]=newm[i][j];
3219: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3220: }
3221: /*printf("h=%d ",h);*/
3222: } /* end h */
1.222 brouard 3223: /* printf("\n H=%d \n",h); */
1.217 brouard 3224: return po;
3225: }
3226:
3227:
1.162 brouard 3228: #ifdef NLOPT
3229: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3230: double fret;
3231: double *xt;
3232: int j;
3233: myfunc_data *d2 = (myfunc_data *) pd;
3234: /* xt = (p1-1); */
3235: xt=vector(1,n);
3236: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3237:
3238: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3239: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3240: printf("Function = %.12lf ",fret);
3241: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3242: printf("\n");
3243: free_vector(xt,1,n);
3244: return fret;
3245: }
3246: #endif
1.126 brouard 3247:
3248: /*************** log-likelihood *************/
3249: double func( double *x)
3250: {
1.226 brouard 3251: int i, ii, j, k, mi, d, kk;
3252: int ioffset=0;
3253: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3254: double **out;
3255: double lli; /* Individual log likelihood */
3256: int s1, s2;
1.228 brouard 3257: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
1.226 brouard 3258: double bbh, survp;
3259: long ipmx;
3260: double agexact;
3261: /*extern weight */
3262: /* We are differentiating ll according to initial status */
3263: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3264: /*for(i=1;i<imx;i++)
3265: printf(" %d\n",s[4][i]);
3266: */
1.162 brouard 3267:
1.226 brouard 3268: ++countcallfunc;
1.162 brouard 3269:
1.226 brouard 3270: cov[1]=1.;
1.126 brouard 3271:
1.226 brouard 3272: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3273: ioffset=0;
1.226 brouard 3274: if(mle==1){
3275: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3276: /* Computes the values of the ncovmodel covariates of the model
3277: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3278: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3279: to be observed in j being in i according to the model.
3280: */
1.243 brouard 3281: ioffset=2+nagesqr ;
1.233 brouard 3282: /* Fixed */
1.234 brouard 3283: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3284: cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (k=6)*/
3285: }
1.226 brouard 3286: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3287: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3288: has been calculated etc */
3289: /* For an individual i, wav[i] gives the number of effective waves */
3290: /* We compute the contribution to Likelihood of each effective transition
3291: mw[mi][i] is real wave of the mi th effectve wave */
3292: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3293: s2=s[mw[mi+1][i]][i];
3294: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3295: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3296: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3297: */
3298: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3299: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3300: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3301: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
1.234 brouard 3302: }
3303: for (ii=1;ii<=nlstate+ndeath;ii++)
3304: for (j=1;j<=nlstate+ndeath;j++){
3305: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3306: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3307: }
3308: for(d=0; d<dh[mi][i]; d++){
3309: newm=savm;
3310: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3311: cov[2]=agexact;
3312: if(nagesqr==1)
3313: cov[3]= agexact*agexact; /* Should be changed here */
3314: for (kk=1; kk<=cptcovage;kk++) {
1.242 brouard 3315: if(!FixedV[Tvar[Tage[kk]]])
1.234 brouard 3316: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
1.242 brouard 3317: else
3318: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
1.234 brouard 3319: }
3320: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3321: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3322: savm=oldm;
3323: oldm=newm;
3324: } /* end mult */
3325:
3326: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3327: /* But now since version 0.9 we anticipate for bias at large stepm.
3328: * If stepm is larger than one month (smallest stepm) and if the exact delay
3329: * (in months) between two waves is not a multiple of stepm, we rounded to
3330: * the nearest (and in case of equal distance, to the lowest) interval but now
3331: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3332: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3333: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3334: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3335: * -stepm/2 to stepm/2 .
3336: * For stepm=1 the results are the same as for previous versions of Imach.
3337: * For stepm > 1 the results are less biased than in previous versions.
3338: */
1.234 brouard 3339: s1=s[mw[mi][i]][i];
3340: s2=s[mw[mi+1][i]][i];
3341: bbh=(double)bh[mi][i]/(double)stepm;
3342: /* bias bh is positive if real duration
3343: * is higher than the multiple of stepm and negative otherwise.
3344: */
3345: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3346: if( s2 > nlstate){
3347: /* i.e. if s2 is a death state and if the date of death is known
3348: then the contribution to the likelihood is the probability to
3349: die between last step unit time and current step unit time,
3350: which is also equal to probability to die before dh
3351: minus probability to die before dh-stepm .
3352: In version up to 0.92 likelihood was computed
3353: as if date of death was unknown. Death was treated as any other
3354: health state: the date of the interview describes the actual state
3355: and not the date of a change in health state. The former idea was
3356: to consider that at each interview the state was recorded
3357: (healthy, disable or death) and IMaCh was corrected; but when we
3358: introduced the exact date of death then we should have modified
3359: the contribution of an exact death to the likelihood. This new
3360: contribution is smaller and very dependent of the step unit
3361: stepm. It is no more the probability to die between last interview
3362: and month of death but the probability to survive from last
3363: interview up to one month before death multiplied by the
3364: probability to die within a month. Thanks to Chris
3365: Jackson for correcting this bug. Former versions increased
3366: mortality artificially. The bad side is that we add another loop
3367: which slows down the processing. The difference can be up to 10%
3368: lower mortality.
3369: */
3370: /* If, at the beginning of the maximization mostly, the
3371: cumulative probability or probability to be dead is
3372: constant (ie = 1) over time d, the difference is equal to
3373: 0. out[s1][3] = savm[s1][3]: probability, being at state
3374: s1 at precedent wave, to be dead a month before current
3375: wave is equal to probability, being at state s1 at
3376: precedent wave, to be dead at mont of the current
3377: wave. Then the observed probability (that this person died)
3378: is null according to current estimated parameter. In fact,
3379: it should be very low but not zero otherwise the log go to
3380: infinity.
3381: */
1.183 brouard 3382: /* #ifdef INFINITYORIGINAL */
3383: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3384: /* #else */
3385: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3386: /* lli=log(mytinydouble); */
3387: /* else */
3388: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3389: /* #endif */
1.226 brouard 3390: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3391:
1.226 brouard 3392: } else if ( s2==-1 ) { /* alive */
3393: for (j=1,survp=0. ; j<=nlstate; j++)
3394: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3395: /*survp += out[s1][j]; */
3396: lli= log(survp);
3397: }
3398: else if (s2==-4) {
3399: for (j=3,survp=0. ; j<=nlstate; j++)
3400: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3401: lli= log(survp);
3402: }
3403: else if (s2==-5) {
3404: for (j=1,survp=0. ; j<=2; j++)
3405: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3406: lli= log(survp);
3407: }
3408: else{
3409: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3410: /* lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*(savm[s1][s2])):log((1.+bbh)*out[s1][s2]));*/ /* linear interpolation */
3411: }
3412: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3413: /*if(lli ==000.0)*/
3414: /*printf("bbh= %f lli=%f savm=%f out=%f %d\n",bbh,lli,savm[s1][s2], out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]],i); */
3415: ipmx +=1;
3416: sw += weight[i];
3417: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3418: /* if (lli < log(mytinydouble)){ */
3419: /* printf("Close to inf lli = %.10lf < %.10lf i= %d mi= %d, s[%d][i]=%d s1=%d s2=%d\n", lli,log(mytinydouble), i, mi,mw[mi][i], s[mw[mi][i]][i], s1,s2); */
3420: /* fprintf(ficlog,"Close to inf lli = %.10lf i= %d mi= %d, s[mw[mi][i]][i]=%d\n", lli, i, mi,s[mw[mi][i]][i]); */
3421: /* } */
3422: } /* end of wave */
3423: } /* end of individual */
3424: } else if(mle==2){
3425: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3426: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3427: for(mi=1; mi<= wav[i]-1; mi++){
3428: for (ii=1;ii<=nlstate+ndeath;ii++)
3429: for (j=1;j<=nlstate+ndeath;j++){
3430: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3431: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3432: }
3433: for(d=0; d<=dh[mi][i]; d++){
3434: newm=savm;
3435: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3436: cov[2]=agexact;
3437: if(nagesqr==1)
3438: cov[3]= agexact*agexact;
3439: for (kk=1; kk<=cptcovage;kk++) {
3440: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3441: }
3442: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3443: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3444: savm=oldm;
3445: oldm=newm;
3446: } /* end mult */
3447:
3448: s1=s[mw[mi][i]][i];
3449: s2=s[mw[mi+1][i]][i];
3450: bbh=(double)bh[mi][i]/(double)stepm;
3451: lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*(savm[s1][s2])):log((1.+bbh)*out[s1][s2])); /* linear interpolation */
3452: ipmx +=1;
3453: sw += weight[i];
3454: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3455: } /* end of wave */
3456: } /* end of individual */
3457: } else if(mle==3){ /* exponential inter-extrapolation */
3458: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3459: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3460: for(mi=1; mi<= wav[i]-1; mi++){
3461: for (ii=1;ii<=nlstate+ndeath;ii++)
3462: for (j=1;j<=nlstate+ndeath;j++){
3463: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3464: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3465: }
3466: for(d=0; d<dh[mi][i]; d++){
3467: newm=savm;
3468: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3469: cov[2]=agexact;
3470: if(nagesqr==1)
3471: cov[3]= agexact*agexact;
3472: for (kk=1; kk<=cptcovage;kk++) {
3473: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3474: }
3475: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3476: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3477: savm=oldm;
3478: oldm=newm;
3479: } /* end mult */
3480:
3481: s1=s[mw[mi][i]][i];
3482: s2=s[mw[mi+1][i]][i];
3483: bbh=(double)bh[mi][i]/(double)stepm;
3484: lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* exponential inter-extrapolation */
3485: ipmx +=1;
3486: sw += weight[i];
3487: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3488: } /* end of wave */
3489: } /* end of individual */
3490: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3491: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3492: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3493: for(mi=1; mi<= wav[i]-1; mi++){
3494: for (ii=1;ii<=nlstate+ndeath;ii++)
3495: for (j=1;j<=nlstate+ndeath;j++){
3496: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3497: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3498: }
3499: for(d=0; d<dh[mi][i]; d++){
3500: newm=savm;
3501: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3502: cov[2]=agexact;
3503: if(nagesqr==1)
3504: cov[3]= agexact*agexact;
3505: for (kk=1; kk<=cptcovage;kk++) {
3506: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3507: }
1.126 brouard 3508:
1.226 brouard 3509: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3510: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3511: savm=oldm;
3512: oldm=newm;
3513: } /* end mult */
3514:
3515: s1=s[mw[mi][i]][i];
3516: s2=s[mw[mi+1][i]][i];
3517: if( s2 > nlstate){
3518: lli=log(out[s1][s2] - savm[s1][s2]);
3519: } else if ( s2==-1 ) { /* alive */
3520: for (j=1,survp=0. ; j<=nlstate; j++)
3521: survp += out[s1][j];
3522: lli= log(survp);
3523: }else{
3524: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3525: }
3526: ipmx +=1;
3527: sw += weight[i];
3528: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3529: /* printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]); */
1.226 brouard 3530: } /* end of wave */
3531: } /* end of individual */
3532: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3533: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3534: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3535: for(mi=1; mi<= wav[i]-1; mi++){
3536: for (ii=1;ii<=nlstate+ndeath;ii++)
3537: for (j=1;j<=nlstate+ndeath;j++){
3538: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3539: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3540: }
3541: for(d=0; d<dh[mi][i]; d++){
3542: newm=savm;
3543: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3544: cov[2]=agexact;
3545: if(nagesqr==1)
3546: cov[3]= agexact*agexact;
3547: for (kk=1; kk<=cptcovage;kk++) {
3548: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3549: }
1.126 brouard 3550:
1.226 brouard 3551: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3552: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3553: savm=oldm;
3554: oldm=newm;
3555: } /* end mult */
3556:
3557: s1=s[mw[mi][i]][i];
3558: s2=s[mw[mi+1][i]][i];
3559: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3560: ipmx +=1;
3561: sw += weight[i];
3562: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3563: /*printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]);*/
3564: } /* end of wave */
3565: } /* end of individual */
3566: } /* End of if */
3567: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3568: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3569: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3570: return -l;
1.126 brouard 3571: }
3572:
3573: /*************** log-likelihood *************/
3574: double funcone( double *x)
3575: {
1.228 brouard 3576: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3577: int i, ii, j, k, mi, d, kk;
1.228 brouard 3578: int ioffset=0;
1.131 brouard 3579: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3580: double **out;
3581: double lli; /* Individual log likelihood */
3582: double llt;
3583: int s1, s2;
1.228 brouard 3584: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3585:
1.126 brouard 3586: double bbh, survp;
1.187 brouard 3587: double agexact;
1.214 brouard 3588: double agebegin, ageend;
1.126 brouard 3589: /*extern weight */
3590: /* We are differentiating ll according to initial status */
3591: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3592: /*for(i=1;i<imx;i++)
3593: printf(" %d\n",s[4][i]);
3594: */
3595: cov[1]=1.;
3596:
3597: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3598: ioffset=0;
3599: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.243 brouard 3600: /* ioffset=2+nagesqr+cptcovage; */
3601: ioffset=2+nagesqr;
1.232 brouard 3602: /* Fixed */
1.224 brouard 3603: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3604: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3605: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3606: cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (k=6)*/
3607: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3608: /* cov[2+6]=covar[Tvar[6]][i]; */
3609: /* cov[2+6]=covar[2][i]; V2 */
3610: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3611: /* cov[2+7]=covar[Tvar[7]][i]; */
3612: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3613: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3614: /* cov[2+9]=covar[Tvar[9]][i]; */
3615: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3616: }
1.232 brouard 3617: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3618: /* cov[++ioffset]=coqvar[TvarFQ[k]][i];/\* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V2 and V1*V2 is fixed (k=6 and 7?)*\/ */
3619: /* } */
1.231 brouard 3620: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3621: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3622: /* } */
1.225 brouard 3623:
1.233 brouard 3624:
3625: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3626: /* Wave varying (but not age varying) */
3627: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3628: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3629: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
3630: }
1.232 brouard 3631: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.242 brouard 3632: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3633: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
3634: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3635: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
3636: /* printf(" i=%d,mi=%d,itv=%d,TmodelInvind[itv]=%d,cotvar[mw[mi][i]][TmodelInvind[itv]][i]=%f\n", i, mi, itv, TmodelInvind[itv],cotvar[mw[mi][i]][TmodelInvind[itv]][i]); */
1.232 brouard 3637: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
1.242 brouard 3638: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3639: /* /\* printf(" i=%d,mi=%d,iqtv=%d,TmodelInvQind[iqtv]=%d,cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]=%f\n", i, mi, iqtv, TmodelInvQind[iqtv],cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]); *\/ */
3640: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
1.232 brouard 3641: /* } */
1.126 brouard 3642: for (ii=1;ii<=nlstate+ndeath;ii++)
1.242 brouard 3643: for (j=1;j<=nlstate+ndeath;j++){
3644: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3645: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3646: }
1.214 brouard 3647:
3648: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3649: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3650: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.247 brouard 3651: /* for(d=0; d<=0; d++){ /\* Delay between two effective waves Only one matrix to speed up*\/ */
1.242 brouard 3652: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3653: and mw[mi+1][i]. dh depends on stepm.*/
3654: newm=savm;
1.247 brouard 3655: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM; /* Here d is needed */
1.242 brouard 3656: cov[2]=agexact;
3657: if(nagesqr==1)
3658: cov[3]= agexact*agexact;
3659: for (kk=1; kk<=cptcovage;kk++) {
3660: if(!FixedV[Tvar[Tage[kk]]])
3661: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3662: else
3663: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
3664: }
3665: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3666: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3667: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3668: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3669: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3670: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3671: savm=oldm;
3672: oldm=newm;
1.126 brouard 3673: } /* end mult */
3674:
3675: s1=s[mw[mi][i]][i];
3676: s2=s[mw[mi+1][i]][i];
1.217 brouard 3677: /* if(s2==-1){ */
3678: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3679: /* /\* exit(1); *\/ */
3680: /* } */
1.126 brouard 3681: bbh=(double)bh[mi][i]/(double)stepm;
3682: /* bias is positive if real duration
3683: * is higher than the multiple of stepm and negative otherwise.
3684: */
3685: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.242 brouard 3686: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3687: } else if ( s2==-1 ) { /* alive */
1.242 brouard 3688: for (j=1,survp=0. ; j<=nlstate; j++)
3689: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3690: lli= log(survp);
1.126 brouard 3691: }else if (mle==1){
1.242 brouard 3692: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3693: } else if(mle==2){
1.242 brouard 3694: lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* linear interpolation */
1.126 brouard 3695: } else if(mle==3){ /* exponential inter-extrapolation */
1.242 brouard 3696: lli= (savm[s1][s2]>(double)1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* exponential inter-extrapolation */
1.126 brouard 3697: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.242 brouard 3698: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3699: } else{ /* mle=0 back to 1 */
1.242 brouard 3700: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3701: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3702: } /* End of if */
3703: ipmx +=1;
3704: sw += weight[i];
3705: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3706: /*printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]); */
1.126 brouard 3707: if(globpr){
1.246 brouard 3708: fprintf(ficresilk,"%09ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3709: %11.6f %11.6f %11.6f ", \
1.242 brouard 3710: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3711: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3712: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3713: llt +=ll[k]*gipmx/gsw;
3714: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3715: }
3716: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3717: }
1.232 brouard 3718: } /* end of wave */
3719: } /* end of individual */
3720: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3721: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3722: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3723: if(globpr==0){ /* First time we count the contributions and weights */
3724: gipmx=ipmx;
3725: gsw=sw;
3726: }
3727: return -l;
1.126 brouard 3728: }
3729:
3730:
3731: /*************** function likelione ***********/
3732: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3733: {
3734: /* This routine should help understanding what is done with
3735: the selection of individuals/waves and
3736: to check the exact contribution to the likelihood.
3737: Plotting could be done.
3738: */
3739: int k;
3740:
3741: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3742: strcpy(fileresilk,"ILK_");
1.202 brouard 3743: strcat(fileresilk,fileresu);
1.126 brouard 3744: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3745: printf("Problem with resultfile: %s\n", fileresilk);
3746: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3747: }
1.214 brouard 3748: fprintf(ficresilk, "#individual(line's_record) count ageb ageend s1 s2 wave# effective_wave# number_of_matrices_product pij weight weight/gpw -2ln(pij)*weight 0pij_x 0pij_(x-stepm) cumulating_loglikeli_by_health_state(reweighted=-2ll*weightXnumber_of_contribs/sum_of_weights) and_total\n");
3749: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3750: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3751: for(k=1; k<=nlstate; k++)
3752: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3753: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3754: }
3755:
3756: *fretone=(*funcone)(p);
3757: if(*globpri !=0){
3758: fclose(ficresilk);
1.205 brouard 3759: if (mle ==0)
3760: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3761: else if(mle >=1)
3762: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3763: fprintf(fichtm," You should at least run with mle >= 1 to get starting values corresponding to the optimized parameters in order to visualize the real contribution of each individual/wave: <a href=\"%s\">%s</a><br>\n",subdirf(fileresilk),subdirf(fileresilk));
1.207 brouard 3764:
1.208 brouard 3765:
3766: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3767: fprintf(fichtm,"<br>- Probability p<sub>%dj</sub> by origin %d and destination j. Dot's sizes are related to corresponding weight: <a href=\"%s-p%dj.png\">%s-p%dj.png</a><br> \
1.208 brouard 3768: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3769: }
1.207 brouard 3770: fprintf(fichtm,"<br>- The function drawn is -2Log(L) in Log scale: by state of origin <a href=\"%s-ori.png\">%s-ori.png</a><br> \
1.204 brouard 3771: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3772: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3773: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3774: fflush(fichtm);
1.205 brouard 3775: }
1.126 brouard 3776: return;
3777: }
3778:
3779:
3780: /*********** Maximum Likelihood Estimation ***************/
3781:
3782: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3783: {
1.165 brouard 3784: int i,j, iter=0;
1.126 brouard 3785: double **xi;
3786: double fret;
3787: double fretone; /* Only one call to likelihood */
3788: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3789:
3790: #ifdef NLOPT
3791: int creturn;
3792: nlopt_opt opt;
3793: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3794: double *lb;
3795: double minf; /* the minimum objective value, upon return */
3796: double * p1; /* Shifted parameters from 0 instead of 1 */
3797: myfunc_data dinst, *d = &dinst;
3798: #endif
3799:
3800:
1.126 brouard 3801: xi=matrix(1,npar,1,npar);
3802: for (i=1;i<=npar;i++)
3803: for (j=1;j<=npar;j++)
3804: xi[i][j]=(i==j ? 1.0 : 0.0);
3805: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3806: strcpy(filerespow,"POW_");
1.126 brouard 3807: strcat(filerespow,fileres);
3808: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3809: printf("Problem with resultfile: %s\n", filerespow);
3810: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3811: }
3812: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3813: for (i=1;i<=nlstate;i++)
3814: for(j=1;j<=nlstate+ndeath;j++)
3815: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3816: fprintf(ficrespow,"\n");
1.162 brouard 3817: #ifdef POWELL
1.126 brouard 3818: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3819: #endif
1.126 brouard 3820:
1.162 brouard 3821: #ifdef NLOPT
3822: #ifdef NEWUOA
3823: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3824: #else
3825: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3826: #endif
3827: lb=vector(0,npar-1);
3828: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3829: nlopt_set_lower_bounds(opt, lb);
3830: nlopt_set_initial_step1(opt, 0.1);
3831:
3832: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3833: d->function = func;
3834: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3835: nlopt_set_min_objective(opt, myfunc, d);
3836: nlopt_set_xtol_rel(opt, ftol);
3837: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3838: printf("nlopt failed! %d\n",creturn);
3839: }
3840: else {
3841: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3842: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3843: iter=1; /* not equal */
3844: }
3845: nlopt_destroy(opt);
3846: #endif
1.126 brouard 3847: free_matrix(xi,1,npar,1,npar);
3848: fclose(ficrespow);
1.203 brouard 3849: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3850: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3851: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3852:
3853: }
3854:
3855: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3856: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3857: {
3858: double **a,**y,*x,pd;
1.203 brouard 3859: /* double **hess; */
1.164 brouard 3860: int i, j;
1.126 brouard 3861: int *indx;
3862:
3863: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3864: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3865: void lubksb(double **a, int npar, int *indx, double b[]) ;
3866: void ludcmp(double **a, int npar, int *indx, double *d) ;
3867: double gompertz(double p[]);
1.203 brouard 3868: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3869:
3870: printf("\nCalculation of the hessian matrix. Wait...\n");
3871: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3872: for (i=1;i<=npar;i++){
1.203 brouard 3873: printf("%d-",i);fflush(stdout);
3874: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3875:
3876: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3877:
3878: /* printf(" %f ",p[i]);
3879: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3880: }
3881:
3882: for (i=1;i<=npar;i++) {
3883: for (j=1;j<=npar;j++) {
3884: if (j>i) {
1.203 brouard 3885: printf(".%d-%d",i,j);fflush(stdout);
3886: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3887: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3888:
3889: hess[j][i]=hess[i][j];
3890: /*printf(" %lf ",hess[i][j]);*/
3891: }
3892: }
3893: }
3894: printf("\n");
3895: fprintf(ficlog,"\n");
3896:
3897: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3898: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3899:
3900: a=matrix(1,npar,1,npar);
3901: y=matrix(1,npar,1,npar);
3902: x=vector(1,npar);
3903: indx=ivector(1,npar);
3904: for (i=1;i<=npar;i++)
3905: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3906: ludcmp(a,npar,indx,&pd);
3907:
3908: for (j=1;j<=npar;j++) {
3909: for (i=1;i<=npar;i++) x[i]=0;
3910: x[j]=1;
3911: lubksb(a,npar,indx,x);
3912: for (i=1;i<=npar;i++){
3913: matcov[i][j]=x[i];
3914: }
3915: }
3916:
3917: printf("\n#Hessian matrix#\n");
3918: fprintf(ficlog,"\n#Hessian matrix#\n");
3919: for (i=1;i<=npar;i++) {
3920: for (j=1;j<=npar;j++) {
1.203 brouard 3921: printf("%.6e ",hess[i][j]);
3922: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3923: }
3924: printf("\n");
3925: fprintf(ficlog,"\n");
3926: }
3927:
1.203 brouard 3928: /* printf("\n#Covariance matrix#\n"); */
3929: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3930: /* for (i=1;i<=npar;i++) { */
3931: /* for (j=1;j<=npar;j++) { */
3932: /* printf("%.6e ",matcov[i][j]); */
3933: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3934: /* } */
3935: /* printf("\n"); */
3936: /* fprintf(ficlog,"\n"); */
3937: /* } */
3938:
1.126 brouard 3939: /* Recompute Inverse */
1.203 brouard 3940: /* for (i=1;i<=npar;i++) */
3941: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3942: /* ludcmp(a,npar,indx,&pd); */
3943:
3944: /* printf("\n#Hessian matrix recomputed#\n"); */
3945:
3946: /* for (j=1;j<=npar;j++) { */
3947: /* for (i=1;i<=npar;i++) x[i]=0; */
3948: /* x[j]=1; */
3949: /* lubksb(a,npar,indx,x); */
3950: /* for (i=1;i<=npar;i++){ */
3951: /* y[i][j]=x[i]; */
3952: /* printf("%.3e ",y[i][j]); */
3953: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3954: /* } */
3955: /* printf("\n"); */
3956: /* fprintf(ficlog,"\n"); */
3957: /* } */
3958:
3959: /* Verifying the inverse matrix */
3960: #ifdef DEBUGHESS
3961: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3962:
1.203 brouard 3963: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3964: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3965:
3966: for (j=1;j<=npar;j++) {
3967: for (i=1;i<=npar;i++){
1.203 brouard 3968: printf("%.2f ",y[i][j]);
3969: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3970: }
3971: printf("\n");
3972: fprintf(ficlog,"\n");
3973: }
1.203 brouard 3974: #endif
1.126 brouard 3975:
3976: free_matrix(a,1,npar,1,npar);
3977: free_matrix(y,1,npar,1,npar);
3978: free_vector(x,1,npar);
3979: free_ivector(indx,1,npar);
1.203 brouard 3980: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3981:
3982:
3983: }
3984:
3985: /*************** hessian matrix ****************/
3986: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3987: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3988: int i;
3989: int l=1, lmax=20;
1.203 brouard 3990: double k1,k2, res, fx;
1.132 brouard 3991: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3992: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3993: int k=0,kmax=10;
3994: double l1;
3995:
3996: fx=func(x);
3997: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3998: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3999: l1=pow(10,l);
4000: delts=delt;
4001: for(k=1 ; k <kmax; k=k+1){
4002: delt = delta*(l1*k);
4003: p2[theta]=x[theta] +delt;
1.145 brouard 4004: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 4005: p2[theta]=x[theta]-delt;
4006: k2=func(p2)-fx;
4007: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 4008: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 4009:
1.203 brouard 4010: #ifdef DEBUGHESSII
1.126 brouard 4011: printf("%d %d k1=%.12e k2=%.12e xk1=%.12e xk2=%.12e delt=%.12e res=%.12e l=%d k=%d,fx=%.12e\n",theta,theta,k1,k2,x[theta]+delt,x[theta]-delt,delt,res, l, k,fx);
4012: fprintf(ficlog,"%d %d k1=%.12e k2=%.12e xk1=%.12e xk2=%.12e delt=%.12e res=%.12e l=%d k=%d,fx=%.12e\n",theta,theta,k1,k2,x[theta]+delt,x[theta]-delt,delt,res, l, k,fx);
4013: #endif
4014: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
4015: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
4016: k=kmax;
4017: }
4018: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 4019: k=kmax; l=lmax*10;
1.126 brouard 4020: }
4021: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
4022: delts=delt;
4023: }
1.203 brouard 4024: } /* End loop k */
1.126 brouard 4025: }
4026: delti[theta]=delts;
4027: return res;
4028:
4029: }
4030:
1.203 brouard 4031: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 4032: {
4033: int i;
1.164 brouard 4034: int l=1, lmax=20;
1.126 brouard 4035: double k1,k2,k3,k4,res,fx;
1.132 brouard 4036: double p2[MAXPARM+1];
1.203 brouard 4037: int k, kmax=1;
4038: double v1, v2, cv12, lc1, lc2;
1.208 brouard 4039:
4040: int firstime=0;
1.203 brouard 4041:
1.126 brouard 4042: fx=func(x);
1.203 brouard 4043: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 4044: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 4045: p2[thetai]=x[thetai]+delti[thetai]*k;
4046: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4047: k1=func(p2)-fx;
4048:
1.203 brouard 4049: p2[thetai]=x[thetai]+delti[thetai]*k;
4050: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4051: k2=func(p2)-fx;
4052:
1.203 brouard 4053: p2[thetai]=x[thetai]-delti[thetai]*k;
4054: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4055: k3=func(p2)-fx;
4056:
1.203 brouard 4057: p2[thetai]=x[thetai]-delti[thetai]*k;
4058: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4059: k4=func(p2)-fx;
1.203 brouard 4060: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
4061: if(k1*k2*k3*k4 <0.){
1.208 brouard 4062: firstime=1;
1.203 brouard 4063: kmax=kmax+10;
1.208 brouard 4064: }
4065: if(kmax >=10 || firstime ==1){
1.246 brouard 4066: printf("Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you could increase ftol=%.2e\n",thetai,thetaj, ftol);
4067: fprintf(ficlog,"Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you could increase ftol=%.2e\n",thetai,thetaj, ftol);
1.203 brouard 4068: printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti*k=%.12e deltj*k=%.12e, xi-de*k=%.12e xj-de*k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4069: fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti*k=%.12e deltj*k=%.12e, xi-de*k=%.12e xj-de*k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4070: }
4071: #ifdef DEBUGHESSIJ
4072: v1=hess[thetai][thetai];
4073: v2=hess[thetaj][thetaj];
4074: cv12=res;
4075: /* Computing eigen value of Hessian matrix */
4076: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4077: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4078: if ((lc2 <0) || (lc1 <0) ){
4079: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4080: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4081: printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4082: fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4083: }
1.126 brouard 4084: #endif
4085: }
4086: return res;
4087: }
4088:
1.203 brouard 4089: /* Not done yet: Was supposed to fix if not exactly at the maximum */
4090: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
4091: /* { */
4092: /* int i; */
4093: /* int l=1, lmax=20; */
4094: /* double k1,k2,k3,k4,res,fx; */
4095: /* double p2[MAXPARM+1]; */
4096: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
4097: /* int k=0,kmax=10; */
4098: /* double l1; */
4099:
4100: /* fx=func(x); */
4101: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
4102: /* l1=pow(10,l); */
4103: /* delts=delt; */
4104: /* for(k=1 ; k <kmax; k=k+1){ */
4105: /* delt = delti*(l1*k); */
4106: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
4107: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4108: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4109: /* k1=func(p2)-fx; */
4110:
4111: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4112: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4113: /* k2=func(p2)-fx; */
4114:
4115: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4116: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4117: /* k3=func(p2)-fx; */
4118:
4119: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4120: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4121: /* k4=func(p2)-fx; */
4122: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
4123: /* #ifdef DEBUGHESSIJ */
4124: /* printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4); */
4125: /* fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4); */
4126: /* #endif */
4127: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
4128: /* k=kmax; */
4129: /* } */
4130: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
4131: /* k=kmax; l=lmax*10; */
4132: /* } */
4133: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
4134: /* delts=delt; */
4135: /* } */
4136: /* } /\* End loop k *\/ */
4137: /* } */
4138: /* delti[theta]=delts; */
4139: /* return res; */
4140: /* } */
4141:
4142:
1.126 brouard 4143: /************** Inverse of matrix **************/
4144: void ludcmp(double **a, int n, int *indx, double *d)
4145: {
4146: int i,imax,j,k;
4147: double big,dum,sum,temp;
4148: double *vv;
4149:
4150: vv=vector(1,n);
4151: *d=1.0;
4152: for (i=1;i<=n;i++) {
4153: big=0.0;
4154: for (j=1;j<=n;j++)
4155: if ((temp=fabs(a[i][j])) > big) big=temp;
1.256 brouard 4156: if (big == 0.0){
4157: printf(" Singular Hessian matrix at row %d:\n",i);
4158: for (j=1;j<=n;j++) {
4159: printf(" a[%d][%d]=%f,",i,j,a[i][j]);
4160: fprintf(ficlog," a[%d][%d]=%f,",i,j,a[i][j]);
4161: }
4162: fflush(ficlog);
4163: fclose(ficlog);
4164: nrerror("Singular matrix in routine ludcmp");
4165: }
1.126 brouard 4166: vv[i]=1.0/big;
4167: }
4168: for (j=1;j<=n;j++) {
4169: for (i=1;i<j;i++) {
4170: sum=a[i][j];
4171: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
4172: a[i][j]=sum;
4173: }
4174: big=0.0;
4175: for (i=j;i<=n;i++) {
4176: sum=a[i][j];
4177: for (k=1;k<j;k++)
4178: sum -= a[i][k]*a[k][j];
4179: a[i][j]=sum;
4180: if ( (dum=vv[i]*fabs(sum)) >= big) {
4181: big=dum;
4182: imax=i;
4183: }
4184: }
4185: if (j != imax) {
4186: for (k=1;k<=n;k++) {
4187: dum=a[imax][k];
4188: a[imax][k]=a[j][k];
4189: a[j][k]=dum;
4190: }
4191: *d = -(*d);
4192: vv[imax]=vv[j];
4193: }
4194: indx[j]=imax;
4195: if (a[j][j] == 0.0) a[j][j]=TINY;
4196: if (j != n) {
4197: dum=1.0/(a[j][j]);
4198: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4199: }
4200: }
4201: free_vector(vv,1,n); /* Doesn't work */
4202: ;
4203: }
4204:
4205: void lubksb(double **a, int n, int *indx, double b[])
4206: {
4207: int i,ii=0,ip,j;
4208: double sum;
4209:
4210: for (i=1;i<=n;i++) {
4211: ip=indx[i];
4212: sum=b[ip];
4213: b[ip]=b[i];
4214: if (ii)
4215: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4216: else if (sum) ii=i;
4217: b[i]=sum;
4218: }
4219: for (i=n;i>=1;i--) {
4220: sum=b[i];
4221: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4222: b[i]=sum/a[i][i];
4223: }
4224: }
4225:
4226: void pstamp(FILE *fichier)
4227: {
1.196 brouard 4228: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4229: }
4230:
1.253 brouard 4231: int linreg(int ifi, int ila, int *no, const double x[], const double y[], double* a, double* b, double* r, double* sa, double * sb) {
4232:
4233: /* y=a+bx regression */
4234: double sumx = 0.0; /* sum of x */
4235: double sumx2 = 0.0; /* sum of x**2 */
4236: double sumxy = 0.0; /* sum of x * y */
4237: double sumy = 0.0; /* sum of y */
4238: double sumy2 = 0.0; /* sum of y**2 */
4239: double sume2; /* sum of square or residuals */
4240: double yhat;
4241:
4242: double denom=0;
4243: int i;
4244: int ne=*no;
4245:
4246: for ( i=ifi, ne=0;i<=ila;i++) {
4247: if(!isfinite(x[i]) || !isfinite(y[i])){
4248: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4249: continue;
4250: }
4251: ne=ne+1;
4252: sumx += x[i];
4253: sumx2 += x[i]*x[i];
4254: sumxy += x[i] * y[i];
4255: sumy += y[i];
4256: sumy2 += y[i]*y[i];
4257: denom = (ne * sumx2 - sumx*sumx);
4258: /* printf("ne=%d, i=%d,x[%d]=%f, y[%d]=%f sumx=%f, sumx2=%f, sumxy=%f, sumy=%f, sumy2=%f, denom=%f\n",ne,i,i,x[i],i,y[i], sumx, sumx2,sumxy, sumy, sumy2,denom); */
4259: }
4260:
4261: denom = (ne * sumx2 - sumx*sumx);
4262: if (denom == 0) {
4263: // vertical, slope m is infinity
4264: *b = INFINITY;
4265: *a = 0;
4266: if (r) *r = 0;
4267: return 1;
4268: }
4269:
4270: *b = (ne * sumxy - sumx * sumy) / denom;
4271: *a = (sumy * sumx2 - sumx * sumxy) / denom;
4272: if (r!=NULL) {
4273: *r = (sumxy - sumx * sumy / ne) / /* compute correlation coeff */
4274: sqrt((sumx2 - sumx*sumx/ne) *
4275: (sumy2 - sumy*sumy/ne));
4276: }
4277: *no=ne;
4278: for ( i=ifi, ne=0;i<=ila;i++) {
4279: if(!isfinite(x[i]) || !isfinite(y[i])){
4280: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4281: continue;
4282: }
4283: ne=ne+1;
4284: yhat = y[i] - *a -*b* x[i];
4285: sume2 += yhat * yhat ;
4286:
4287: denom = (ne * sumx2 - sumx*sumx);
4288: /* printf("ne=%d, i=%d,x[%d]=%f, y[%d]=%f sumx=%f, sumx2=%f, sumxy=%f, sumy=%f, sumy2=%f, denom=%f\n",ne,i,i,x[i],i,y[i], sumx, sumx2,sumxy, sumy, sumy2,denom); */
4289: }
4290: *sb = sqrt(sume2/(ne-2)/(sumx2 - sumx * sumx /ne));
4291: *sa= *sb * sqrt(sumx2/ne);
4292:
4293: return 0;
4294: }
4295:
1.126 brouard 4296: /************ Frequencies ********************/
1.251 brouard 4297: void freqsummary(char fileres[], double p[], double pstart[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
1.226 brouard 4298: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4299: int firstpass, int lastpass, int stepm, int weightopt, char model[])
1.250 brouard 4300: { /* Some frequencies as well as proposing some starting values */
1.226 brouard 4301:
1.253 brouard 4302: int i, m, jk, j1, bool, z1,j, nj, nl, k, iv, jj=0;
1.226 brouard 4303: int iind=0, iage=0;
4304: int mi; /* Effective wave */
4305: int first;
4306: double ***freq; /* Frequencies */
1.253 brouard 4307: double *x, *y, a,b,r, sa, sb; /* for regression, y=b+m*x and r is the correlation coefficient */
4308: int no;
1.226 brouard 4309: double *meanq;
4310: double **meanqt;
4311: double *pp, **prop, *posprop, *pospropt;
4312: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4313: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4314: double agebegin, ageend;
4315:
4316: pp=vector(1,nlstate);
1.251 brouard 4317: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4318: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4319: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4320: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4321: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4322: meanqt=matrix(1,lastpass,1,nqtveff);
4323: strcpy(fileresp,"P_");
4324: strcat(fileresp,fileresu);
4325: /*strcat(fileresphtm,fileresu);*/
4326: if((ficresp=fopen(fileresp,"w"))==NULL) {
4327: printf("Problem with prevalence resultfile: %s\n", fileresp);
4328: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4329: exit(0);
4330: }
1.240 brouard 4331:
1.226 brouard 4332: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4333: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4334: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4335: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4336: fflush(ficlog);
4337: exit(70);
4338: }
4339: else{
4340: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4341: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4342: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4343: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4344: }
1.237 brouard 4345: fprintf(ficresphtm,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies and prevalence by age at begin of transition and dummy covariate value at beginning of transition</h4>\n",fileresphtm, fileresphtm);
1.240 brouard 4346:
1.226 brouard 4347: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4348: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4349: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4350: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4351: fflush(ficlog);
4352: exit(70);
1.240 brouard 4353: } else{
1.226 brouard 4354: fprintf(ficresphtmfr,"<html><head>\n<title>IMaCh PHTM_Frequency table %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4355: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4356: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4357: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4358: }
1.240 brouard 4359: fprintf(ficresphtmfr,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies of all effective transitions of the model, by age at begin of transition, and covariate value at the begin of transition (if the covariate is a varying covariate) </h4>Unknown status is -1<br/>\n",fileresphtmfr, fileresphtmfr);
4360:
1.253 brouard 4361: y= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
4362: x= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.251 brouard 4363: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4364: j1=0;
1.126 brouard 4365:
1.227 brouard 4366: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4367: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4368: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.240 brouard 4369:
4370:
1.226 brouard 4371: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4372: reference=low_education V1=0,V2=0
4373: med_educ V1=1 V2=0,
4374: high_educ V1=0 V2=1
4375: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4376: */
1.249 brouard 4377: dateintsum=0;
4378: k2cpt=0;
4379:
1.253 brouard 4380: if(cptcoveff == 0 )
4381: nl=1; /* Constant model only */
4382: else
4383: nl=2;
4384: for (nj = 1; nj <= nl; nj++){ /* nj= 1 constant model, nl number of loops. */
4385: if(nj==1)
4386: j=0; /* First pass for the constant */
4387: else
4388: j=cptcoveff; /* Other passes for the covariate values */
1.251 brouard 4389: first=1;
4390: for (j1 = 1; j1 <= (int) pow(2,j); j1++){ /* Loop on covariates combination in order of model, excluding quantitatives, V4=0, V3=0 for example, fixed or varying covariates */
4391: posproptt=0.;
4392: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4393: scanf("%d", i);*/
4394: for (i=-5; i<=nlstate+ndeath; i++)
4395: for (jk=-5; jk<=nlstate+ndeath; jk++)
4396: for(m=iagemin; m <= iagemax+3; m++)
4397: freq[i][jk][m]=0;
4398:
4399: for (i=1; i<=nlstate; i++) {
1.240 brouard 4400: for(m=iagemin; m <= iagemax+3; m++)
1.251 brouard 4401: prop[i][m]=0;
4402: posprop[i]=0;
4403: pospropt[i]=0;
4404: }
4405: /* for (z1=1; z1<= nqfveff; z1++) { */
4406: /* meanq[z1]+=0.; */
4407: /* for(m=1;m<=lastpass;m++){ */
4408: /* meanqt[m][z1]=0.; */
4409: /* } */
4410: /* } */
4411:
4412: /* dateintsum=0; */
4413: /* k2cpt=0; */
4414:
4415: /* For that combination of covariate j1, we count and print the frequencies in one pass */
4416: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4417: bool=1;
4418: if(j !=0){
4419: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
4420: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
4421: /* for (z1=1; z1<= nqfveff; z1++) { */
4422: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4423: /* } */
4424: for (z1=1; z1<=cptcoveff; z1++) { /* loops on covariates in the model */
4425: /* if(Tvaraff[z1] ==-20){ */
4426: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4427: /* }else if(Tvaraff[z1] ==-10){ */
4428: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4429: /* }else */
4430: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){ /* for combination j1 of covariates */
4431: /* Tests if this individual iind responded to combination j1 (V4=1 V3=0) */
4432: bool=0; /* bool should be equal to 1 to be selected, one covariate value failed */
4433: /* printf("bool=%d i=%d, z1=%d, Tvaraff[%d]=%d, covar[Tvarff][%d]=%2f, codtabm(%d,%d)=%d, nbcode[Tvaraff][codtabm(%d,%d)=%d, j1=%d\n",
4434: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4435: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4436: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4437: } /* Onlyf fixed */
4438: } /* end z1 */
4439: } /* cptcovn > 0 */
4440: } /* end any */
4441: }/* end j==0 */
4442: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
4443: /* for(m=firstpass; m<=lastpass; m++){ */
4444: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4445: m=mw[mi][iind];
4446: if(j!=0){
4447: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4448: for (z1=1; z1<=cptcoveff; z1++) {
4449: if( Fixed[Tmodelind[z1]]==1){
4450: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4451: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality. If covariate's
4452: value is -1, we don't select. It differs from the
4453: constant and age model which counts them. */
4454: bool=0; /* not selected */
4455: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4456: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4457: bool=0;
4458: }
4459: }
4460: }
4461: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4462: } /* end j==0 */
4463: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4464: if(bool==1){
4465: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4466: and mw[mi+1][iind]. dh depends on stepm. */
4467: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4468: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4469: if(m >=firstpass && m <=lastpass){
4470: k2=anint[m][iind]+(mint[m][iind]/12.);
4471: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4472: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4473: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4474: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4475: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4476: if (m<lastpass) {
4477: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4478: /* printf(" num=%ld m=%d, iind=%d s1=%d s2=%d agev at m=%d\n", num[iind], m, iind,s[m][iind],s[m+1][iind], (int)agev[m][iind]); */
4479: if(s[m][iind]==-1)
4480: printf(" num=%ld m=%d, iind=%d s1=%d s2=%d agev at m=%d agebegin=%.2f ageend=%.2f, agemed=%d\n", num[iind], m, iind,s[m][iind],s[m+1][iind], (int)agev[m][iind],agebegin, ageend, (int)((agebegin+ageend)/2.));
4481: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4482: /* if((int)agev[m][iind] == 55) */
4483: /* printf("j=%d, j1=%d Age %d, iind=%d, num=%09ld m=%d\n",j,j1,(int)agev[m][iind],iind, num[iind],m); */
4484: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4485: freq[s[m][iind]][s[m+1][iind]][iagemax+3] += weight[iind]; /* Total is in iagemax+3 *//* At age of beginning of transition, where status is known */
1.234 brouard 4486: }
1.251 brouard 4487: } /* end if between passes */
4488: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99) && (j==0)) {
4489: dateintsum=dateintsum+k2; /* on all covariates ?*/
4490: k2cpt++;
4491: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
1.234 brouard 4492: }
1.251 brouard 4493: }else{
4494: bool=1;
4495: }/* end bool 2 */
4496: } /* end m */
4497: } /* end bool */
4498: } /* end iind = 1 to imx */
4499: /* prop[s][age] is feeded for any initial and valid live state as well as
4500: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
4501:
4502:
4503: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4504: pstamp(ficresp);
4505: if (cptcoveff>0 && j!=0){
4506: printf( "\n#********** Variable ");
4507: fprintf(ficresp, "\n#********** Variable ");
4508: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4509: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
4510: fprintf(ficlog, "\n#********** Variable ");
4511: for (z1=1; z1<=cptcoveff; z1++){
4512: if(!FixedV[Tvaraff[z1]]){
4513: printf( "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4514: fprintf(ficresp, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4515: fprintf(ficresphtm, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4516: fprintf(ficresphtmfr, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4517: fprintf(ficlog, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.250 brouard 4518: }else{
1.251 brouard 4519: printf( "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4520: fprintf(ficresp, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4521: fprintf(ficresphtm, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4522: fprintf(ficresphtmfr, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4523: fprintf(ficlog, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4524: }
4525: }
4526: printf( "**********\n#");
4527: fprintf(ficresp, "**********\n#");
4528: fprintf(ficresphtm, "**********</h3>\n");
4529: fprintf(ficresphtmfr, "**********</h3>\n");
4530: fprintf(ficlog, "**********\n");
4531: }
4532: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4533: for(i=1; i<=nlstate;i++) {
4534: fprintf(ficresp, " Age Prev(%d) N(%d) N ",i,i);
4535: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4536: }
4537: fprintf(ficresp, "\n");
4538: fprintf(ficresphtm, "\n");
4539:
4540: /* Header of frequency table by age */
4541: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4542: fprintf(ficresphtmfr,"<th>Age</th> ");
4543: for(jk=-1; jk <=nlstate+ndeath; jk++){
4544: for(m=-1; m <=nlstate+ndeath; m++){
4545: if(jk!=0 && m!=0)
4546: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.240 brouard 4547: }
1.226 brouard 4548: }
1.251 brouard 4549: fprintf(ficresphtmfr, "\n");
4550:
4551: /* For each age */
4552: for(iage=iagemin; iage <= iagemax+3; iage++){
4553: fprintf(ficresphtm,"<tr>");
4554: if(iage==iagemax+1){
4555: fprintf(ficlog,"1");
4556: fprintf(ficresphtmfr,"<tr><th>0</th> ");
4557: }else if(iage==iagemax+2){
4558: fprintf(ficlog,"0");
4559: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
4560: }else if(iage==iagemax+3){
4561: fprintf(ficlog,"Total");
4562: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
4563: }else{
1.240 brouard 4564: if(first==1){
1.251 brouard 4565: first=0;
4566: printf("See log file for details...\n");
4567: }
4568: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4569: fprintf(ficlog,"Age %d", iage);
4570: }
4571: for(jk=1; jk <=nlstate ; jk++){
4572: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4573: pp[jk] += freq[jk][m][iage];
4574: }
4575: for(jk=1; jk <=nlstate ; jk++){
4576: for(m=-1, pos=0; m <=0 ; m++)
4577: pos += freq[jk][m][iage];
4578: if(pp[jk]>=1.e-10){
4579: if(first==1){
4580: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4581: }
4582: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4583: }else{
4584: if(first==1)
4585: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4586: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
1.240 brouard 4587: }
4588: }
4589:
1.251 brouard 4590: for(jk=1; jk <=nlstate ; jk++){
4591: /* posprop[jk]=0; */
4592: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4593: pp[jk] += freq[jk][m][iage];
4594: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
4595:
4596: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
4597: pos += pp[jk]; /* pos is the total number of transitions until this age */
4598: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4599: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4600: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
1.240 brouard 4601: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4602: }
1.251 brouard 4603: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4604: if(pos>=1.e-5){
1.251 brouard 4605: if(first==1)
4606: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4607: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4608: }else{
4609: if(first==1)
4610: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4611: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4612: }
4613: if( iage <= iagemax){
4614: if(pos>=1.e-5){
4615: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4616: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4617: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4618: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4619: }
4620: else{
4621: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4622: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4623: }
1.240 brouard 4624: }
1.251 brouard 4625: pospropt[jk] +=posprop[jk];
4626: } /* end loop jk */
4627: /* pospropt=0.; */
4628: for(jk=-1; jk <=nlstate+ndeath; jk++){
4629: for(m=-1; m <=nlstate+ndeath; m++){
4630: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4631: if(first==1){
4632: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4633: }
1.253 brouard 4634: /* printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]); */
1.251 brouard 4635: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4636: }
4637: if(jk!=0 && m!=0)
4638: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
1.240 brouard 4639: }
1.251 brouard 4640: } /* end loop jk */
4641: posproptt=0.;
4642: for(jk=1; jk <=nlstate; jk++){
4643: posproptt += pospropt[jk];
4644: }
4645: fprintf(ficresphtmfr,"</tr>\n ");
4646: if(iage <= iagemax){
4647: fprintf(ficresp,"\n");
4648: fprintf(ficresphtm,"</tr>\n");
1.240 brouard 4649: }
1.251 brouard 4650: if(first==1)
4651: printf("Others in log...\n");
4652: fprintf(ficlog,"\n");
4653: } /* end loop age iage */
4654: fprintf(ficresphtm,"<tr><th>Tot</th>");
4655: for(jk=1; jk <=nlstate ; jk++){
4656: if(posproptt < 1.e-5){
4657: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
4658: }else{
4659: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.240 brouard 4660: }
1.226 brouard 4661: }
1.251 brouard 4662: fprintf(ficresphtm,"</tr>\n");
4663: fprintf(ficresphtm,"</table>\n");
4664: fprintf(ficresphtmfr,"</table>\n");
1.226 brouard 4665: if(posproptt < 1.e-5){
1.251 brouard 4666: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4667: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
1.260 brouard 4668: fprintf(ficlog,"# This combination (%d) is not valid and no result will be produced\n",j1);
4669: printf("# This combination (%d) is not valid and no result will be produced\n",j1);
1.251 brouard 4670: invalidvarcomb[j1]=1;
1.226 brouard 4671: }else{
1.251 brouard 4672: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4673: invalidvarcomb[j1]=0;
1.226 brouard 4674: }
1.251 brouard 4675: fprintf(ficresphtmfr,"</table>\n");
4676: fprintf(ficlog,"\n");
4677: if(j!=0){
4678: printf("#Freqsummary: Starting values for combination j1=%d:\n", j1);
4679: for(i=1,jk=1; i <=nlstate; i++){
4680: for(k=1; k <=(nlstate+ndeath); k++){
4681: if (k != i) {
4682: for(jj=1; jj <=ncovmodel; jj++){ /* For counting jk */
1.253 brouard 4683: if(jj==1){ /* Constant case (in fact cste + age) */
1.251 brouard 4684: if(j1==1){ /* All dummy covariates to zero */
4685: freq[i][k][iagemax+4]=freq[i][k][iagemax+3]; /* Stores case 0 0 0 */
4686: freq[i][i][iagemax+4]=freq[i][i][iagemax+3]; /* Stores case 0 0 0 */
1.252 brouard 4687: printf("%d%d ",i,k);
4688: fprintf(ficlog,"%d%d ",i,k);
4689: printf("%12.7f ln(%.0f/%.0f)= %f, OR=%f sd=%f \n",p[jk],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]),freq[i][k][iagemax+3]/freq[i][i][iagemax+3], sqrt(1/freq[i][k][iagemax+3]+1/freq[i][i][iagemax+3]));
4690: fprintf(ficlog,"%12.7f ln(%.0f/%.0f)= %12.7f \n",p[jk],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
4691: pstart[jk]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
1.251 brouard 4692: }
1.253 brouard 4693: }else if((j1==1) && (jj==2 || nagesqr==1)){ /* age or age*age parameter without covariate V4*age (to be done later) */
4694: for(iage=iagemin; iage <= iagemax+3; iage++){
4695: x[iage]= (double)iage;
4696: y[iage]= log(freq[i][k][iage]/freq[i][i][iage]);
4697: /* printf("i=%d, k=%d, jk=%d, j1=%d, jj=%d, y[%d]=%f\n",i,k,jk,j1,jj, iage, y[iage]); */
4698: }
4699: linreg(iagemin,iagemax,&no,x,y,&a,&b,&r, &sa, &sb ); /* y= a+b*x with standard errors */
4700: pstart[jk]=b;
4701: pstart[jk-1]=a;
1.252 brouard 4702: }else if( j1!=1 && (j1==2 || (log(j1-1.)/log(2.)-(int)(log(j1-1.)/log(2.))) <0.010) && ( TvarsDind[(int)(log(j1-1.)/log(2.))+1]+2+nagesqr == jj) && Dummy[jj-2-nagesqr]==0){ /* We want only if the position, jj, in model corresponds to unique covariate equal to 1 in j1 combination */
4703: printf("j1=%d, jj=%d, (int)(log(j1-1.)/log(2.))+1=%d, TvarsDind[(int)(log(j1-1.)/log(2.))+1]=%d\n",j1, jj,(int)(log(j1-1.)/log(2.))+1,TvarsDind[(int)(log(j1-1.)/log(2.))+1]);
4704: printf("j1=%d, jj=%d, (log(j1-1.)/log(2.))+1=%f, TvarsDind[(int)(log(j1-1.)/log(2.))+1]=%d\n",j1, jj,(log(j1-1.)/log(2.))+1,TvarsDind[(int)(log(j1-1.)/log(2.))+1]);
1.251 brouard 4705: pstart[jk]= log((freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4]));
1.252 brouard 4706: printf("%d%d ",i,k);
4707: fprintf(ficlog,"%d%d ",i,k);
1.251 brouard 4708: printf("jk=%d,i=%d,k=%d,p[%d]=%12.7f ln((%.0f/%.0f)/(%.0f/%.0f))= %f, OR=%f sd=%f \n",jk,i,k,jk,p[jk],freq[i][k][iagemax+3],freq[i][i][iagemax+3],freq[i][k][iagemax+4],freq[i][i][iagemax+4], log((freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4])),(freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4]), sqrt(1/freq[i][k][iagemax+3]+1/freq[i][i][iagemax+3]+1/freq[i][k][iagemax+4]+1/freq[i][i][iagemax+4]));
4709: }else{ /* Other cases, like quantitative fixed or varying covariates */
4710: ;
4711: }
4712: /* printf("%12.7f )", param[i][jj][k]); */
4713: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
4714: jk++;
4715: } /* end jj */
4716: } /* end k!= i */
4717: } /* end k */
4718: } /* end i, jk */
4719: } /* end j !=0 */
4720: } /* end selected combination of covariate j1 */
4721: if(j==0){ /* We can estimate starting values from the occurences in each case */
4722: printf("#Freqsummary: Starting values for the constants:\n");
4723: fprintf(ficlog,"\n");
4724: for(i=1,jk=1; i <=nlstate; i++){
4725: for(k=1; k <=(nlstate+ndeath); k++){
4726: if (k != i) {
4727: printf("%d%d ",i,k);
4728: fprintf(ficlog,"%d%d ",i,k);
4729: for(jj=1; jj <=ncovmodel; jj++){
1.253 brouard 4730: pstart[jk]=p[jk]; /* Setting pstart to p values by default */
4731: if(jj==1){ /* Age has to be done */
4732: pstart[jk]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
1.251 brouard 4733: printf("%12.7f ln(%.0f/%.0f)= %12.7f ",p[jk],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
4734: fprintf(ficlog,"%12.7f ln(%.0f/%.0f)= %12.7f ",p[jk],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
4735: }
4736: /* printf("%12.7f )", param[i][jj][k]); */
4737: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
4738: jk++;
1.250 brouard 4739: }
1.251 brouard 4740: printf("\n");
4741: fprintf(ficlog,"\n");
1.250 brouard 4742: }
4743: }
4744: }
1.251 brouard 4745: printf("#Freqsummary\n");
4746: fprintf(ficlog,"\n");
4747: for(jk=-1; jk <=nlstate+ndeath; jk++){
4748: for(m=-1; m <=nlstate+ndeath; m++){
4749: /* param[i]|j][k]= freq[jk][m][iagemax+3] */
1.250 brouard 4750: printf(" %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]);
4751: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]);
1.251 brouard 4752: /* if(freq[jk][m][iage] !=0 ) { /\* minimizing output *\/ */
4753: /* printf(" %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]); */
4754: /* fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]); */
4755: /* } */
4756: }
4757: } /* end loop jk */
4758:
4759: printf("\n");
4760: fprintf(ficlog,"\n");
4761: } /* end j=0 */
1.249 brouard 4762: } /* end j */
1.252 brouard 4763:
1.253 brouard 4764: if(mle == -2){ /* We want to use these values as starting values */
1.252 brouard 4765: for(i=1, jk=1; i <=nlstate; i++){
4766: for(j=1; j <=nlstate+ndeath; j++){
4767: if(j!=i){
4768: /*ca[0]= k+'a'-1;ca[1]='\0';*/
4769: printf("%1d%1d",i,j);
4770: fprintf(ficparo,"%1d%1d",i,j);
4771: for(k=1; k<=ncovmodel;k++){
4772: /* printf(" %lf",param[i][j][k]); */
4773: /* fprintf(ficparo," %lf",param[i][j][k]); */
4774: p[jk]=pstart[jk];
4775: printf(" %f ",pstart[jk]);
4776: fprintf(ficparo," %f ",pstart[jk]);
4777: jk++;
4778: }
4779: printf("\n");
4780: fprintf(ficparo,"\n");
4781: }
4782: }
4783: }
4784: } /* end mle=-2 */
1.226 brouard 4785: dateintmean=dateintsum/k2cpt;
1.240 brouard 4786:
1.226 brouard 4787: fclose(ficresp);
4788: fclose(ficresphtm);
4789: fclose(ficresphtmfr);
4790: free_vector(meanq,1,nqfveff);
4791: free_matrix(meanqt,1,lastpass,1,nqtveff);
1.253 brouard 4792: free_vector(x, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
4793: free_vector(y, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.251 brouard 4794: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4795: free_vector(pospropt,1,nlstate);
4796: free_vector(posprop,1,nlstate);
1.251 brouard 4797: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4798: free_vector(pp,1,nlstate);
4799: /* End of freqsummary */
4800: }
1.126 brouard 4801:
4802: /************ Prevalence ********************/
1.227 brouard 4803: void prevalence(double ***probs, double agemin, double agemax, int **s, double **agev, int nlstate, int imx, int *Tvar, int **nbcode, int *ncodemax,double **mint,double **anint, double dateprev1,double dateprev2, int firstpass, int lastpass)
4804: {
4805: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4806: in each health status at the date of interview (if between dateprev1 and dateprev2).
4807: We still use firstpass and lastpass as another selection.
4808: */
1.126 brouard 4809:
1.227 brouard 4810: int i, m, jk, j1, bool, z1,j, iv;
4811: int mi; /* Effective wave */
4812: int iage;
4813: double agebegin, ageend;
4814:
4815: double **prop;
4816: double posprop;
4817: double y2; /* in fractional years */
4818: int iagemin, iagemax;
4819: int first; /** to stop verbosity which is redirected to log file */
4820:
4821: iagemin= (int) agemin;
4822: iagemax= (int) agemax;
4823: /*pp=vector(1,nlstate);*/
1.251 brouard 4824: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4825: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4826: j1=0;
1.222 brouard 4827:
1.227 brouard 4828: /*j=cptcoveff;*/
4829: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4830:
1.227 brouard 4831: first=1;
4832: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4833: for (i=1; i<=nlstate; i++)
1.251 brouard 4834: for(iage=iagemin-AGEMARGE; iage <= iagemax+4+AGEMARGE; iage++)
1.227 brouard 4835: prop[i][iage]=0.0;
4836: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4837: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4838: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4839:
4840: for (i=1; i<=imx; i++) { /* Each individual */
4841: bool=1;
4842: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4843: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4844: m=mw[mi][i];
4845: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4846: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4847: for (z1=1; z1<=cptcoveff; z1++){
4848: if( Fixed[Tmodelind[z1]]==1){
4849: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4850: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4851: bool=0;
4852: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4853: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4854: bool=0;
4855: }
4856: }
4857: if(bool==1){ /* Otherwise we skip that wave/person */
4858: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4859: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4860: if(m >=firstpass && m <=lastpass){
4861: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4862: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4863: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4864: if(agev[m][i]==1) agev[m][i]=iagemax+2;
1.251 brouard 4865: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+4+AGEMARGE){
1.227 brouard 4866: printf("Error on individual # %d agev[m][i]=%f <%d-%d or > %d+3+%d m=%d; either change agemin or agemax or fix data\n",i, agev[m][i],iagemin,AGEMARGE, iagemax,AGEMARGE,m);
4867: exit(1);
4868: }
4869: if (s[m][i]>0 && s[m][i]<=nlstate) {
4870: /*if(i>4620) printf(" i=%d m=%d s[m][i]=%d (int)agev[m][i]=%d weight[i]=%f prop=%f\n",i,m,s[m][i],(int)agev[m][m],weight[i],prop[s[m][i]][(int)agev[m][i]]);*/
4871: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4872: prop[s[m][i]][iagemax+3] += weight[i];
4873: } /* end valid statuses */
4874: } /* end selection of dates */
4875: } /* end selection of waves */
4876: } /* end bool */
4877: } /* end wave */
4878: } /* end individual */
4879: for(i=iagemin; i <= iagemax+3; i++){
4880: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4881: posprop += prop[jk][i];
4882: }
4883:
4884: for(jk=1; jk <=nlstate ; jk++){
4885: if( i <= iagemax){
4886: if(posprop>=1.e-5){
4887: probs[i][jk][j1]= prop[jk][i]/posprop;
4888: } else{
4889: if(first==1){
4890: first=0;
4891: printf("Warning Observed prevalence probs[%d][%d][%d]=%lf because of lack of cases\nSee others in log file...\n",jk,i,j1,probs[i][jk][j1]);
4892: }
4893: }
4894: }
4895: }/* end jk */
4896: }/* end i */
1.222 brouard 4897: /*} *//* end i1 */
1.227 brouard 4898: } /* end j1 */
1.222 brouard 4899:
1.227 brouard 4900: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4901: /*free_vector(pp,1,nlstate);*/
1.251 brouard 4902: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4903: } /* End of prevalence */
1.126 brouard 4904:
4905: /************* Waves Concatenation ***************/
4906:
4907: void concatwav(int wav[], int **dh, int **bh, int **mw, int **s, double *agedc, double **agev, int firstpass, int lastpass, int imx, int nlstate, int stepm)
4908: {
4909: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4910: Death is a valid wave (if date is known).
4911: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4912: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4913: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4914: */
1.126 brouard 4915:
1.224 brouard 4916: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4917: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4918: double sum=0., jmean=0.;*/
1.224 brouard 4919: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4920: int j, k=0,jk, ju, jl;
4921: double sum=0.;
4922: first=0;
1.214 brouard 4923: firstwo=0;
1.217 brouard 4924: firsthree=0;
1.218 brouard 4925: firstfour=0;
1.164 brouard 4926: jmin=100000;
1.126 brouard 4927: jmax=-1;
4928: jmean=0.;
1.224 brouard 4929:
4930: /* Treating live states */
1.214 brouard 4931: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4932: mi=0; /* First valid wave */
1.227 brouard 4933: mli=0; /* Last valid wave */
1.126 brouard 4934: m=firstpass;
1.214 brouard 4935: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4936: if(m >firstpass && s[m][i]==s[m-1][i] && mint[m][i]==mint[m-1][i] && anint[m][i]==anint[m-1][i]){/* Two succesive identical information on wave m */
4937: mli=m-1;/* mw[++mi][i]=m-1; */
4938: }else if(s[m][i]>=1 || s[m][i]==-4 || s[m][i]==-5){ /* Since 0.98r4 if status=-2 vital status is really unknown, wave should be skipped */
4939: mw[++mi][i]=m;
4940: mli=m;
1.224 brouard 4941: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4942: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4943: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4944: }
1.227 brouard 4945: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4946: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4947: break;
1.224 brouard 4948: #else
1.227 brouard 4949: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4950: if(firsthree == 0){
1.262 ! brouard 4951: printf("Information! Unknown status for individual %ld line=%d occurred at last wave %d at known date %d/%d. Please, check if your unknown date of death %d/%d means a live state %d at wave %d. This case(%d)/wave(%d) contributes to the likelihood as 1-p%d%d .\nOthers in log file only\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], (int) moisdc[i], (int) andc[i], s[m][i], m, i, m, s[m][i], nlstate+ndeath);
1.227 brouard 4952: firsthree=1;
4953: }
1.262 ! brouard 4954: fprintf(ficlog,"Information! Unknown status for individual %ld line=%d occurred at last wave %d at known date %d/%d. Please, check if your unknown date of death %d/%d means a live state %d at wave %d. This case(%d)/wave(%d) contributes to the likelihood as 1-p%d%d .\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], (int) moisdc[i], (int) andc[i], s[m][i], m, i, m, s[m][i], nlstate+ndeath);
1.227 brouard 4955: mw[++mi][i]=m;
4956: mli=m;
4957: }
4958: if(s[m][i]==-2){ /* Vital status is really unknown */
4959: nbwarn++;
4960: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4961: printf("Warning! Vital status for individual %ld (line=%d) at last wave %d interviewed at date %d/%d is unknown %d. Please, check if the vital status and the date of death %d/%d are really unknown. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\nOthers in log file only\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], (int) moisdc[i], (int) andc[i], i, m);
4962: fprintf(ficlog,"Warning! Vital status for individual %ld (line=%d) at last wave %d interviewed at date %d/%d is unknown %d. Please, check if the vital status and the date of death %d/%d are really unknown. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], (int) moisdc[i], (int) andc[i], i, m);
4963: }
4964: break;
4965: }
4966: break;
1.224 brouard 4967: #endif
1.227 brouard 4968: }/* End m >= lastpass */
1.126 brouard 4969: }/* end while */
1.224 brouard 4970:
1.227 brouard 4971: /* mi is the last effective wave, m is lastpass, mw[j][i] gives the # of j-th effective wave for individual i */
1.216 brouard 4972: /* After last pass */
1.224 brouard 4973: /* Treating death states */
1.214 brouard 4974: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4975: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4976: /* } */
1.126 brouard 4977: mi++; /* Death is another wave */
4978: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4979: /* Only death is a correct wave */
1.126 brouard 4980: mw[mi][i]=m;
1.257 brouard 4981: } /* else not in a death state */
1.224 brouard 4982: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.257 brouard 4983: else if ((int) andc[i] != 9999) { /* Date of death is known */
1.218 brouard 4984: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4985: if((andc[i]+moisdc[i]/12.) <=(anint[m][i]+mint[m][i]/12.)){ /* death occured before last wave and status should have been death instead of -1 */
4986: nbwarn++;
4987: if(firstfiv==0){
4988: printf("Warning! Death for individual %ld line=%d occurred at %d/%d before last wave %d interviewed at %d/%d and should have been coded as death instead of '%d'. This case (%d)/wave (%d) is contributing to likelihood.\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], i,m );
4989: firstfiv=1;
4990: }else{
4991: fprintf(ficlog,"Warning! Death for individual %ld line=%d occurred at %d/%d before last wave %d interviewed at %d/%d and should have been coded as death instead of '%d'. This case (%d)/wave (%d) is contributing to likelihood.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], i,m );
4992: }
4993: }else{ /* Death occured afer last wave potential bias */
4994: nberr++;
4995: if(firstwo==0){
1.257 brouard 4996: printf("Error! Death for individual %ld line=%d occurred at %d/%d after last wave %d interviewed at %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood. Please add a new fictive wave at the date of last vital status scan, with a dead status or alive but unknown state status (-1). See documentation\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.227 brouard 4997: firstwo=1;
4998: }
1.257 brouard 4999: fprintf(ficlog,"Error! Death for individual %ld line=%d occurred at %d/%d after last wave %d interviewed at %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood. Please add a new fictive wave at the date of last vital status scan, with a dead status or alive but unknown state status (-1). See documentation\n\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.227 brouard 5000: }
1.257 brouard 5001: }else{ /* if date of interview is unknown */
1.227 brouard 5002: /* death is known but not confirmed by death status at any wave */
5003: if(firstfour==0){
5004: printf("Error! Death for individual %ld line=%d occurred %d/%d but not confirmed by any death status for any wave, including last wave %d at unknown date %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
5005: firstfour=1;
5006: }
5007: fprintf(ficlog,"Error! Death for individual %ld line=%d occurred %d/%d but not confirmed by any death status for any wave, including last wave %d at unknown date %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.214 brouard 5008: }
1.224 brouard 5009: } /* end if date of death is known */
5010: #endif
5011: wav[i]=mi; /* mi should be the last effective wave (or mli) */
5012: /* wav[i]=mw[mi][i]; */
1.126 brouard 5013: if(mi==0){
5014: nbwarn++;
5015: if(first==0){
1.227 brouard 5016: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
5017: first=1;
1.126 brouard 5018: }
5019: if(first==1){
1.227 brouard 5020: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 5021: }
5022: } /* end mi==0 */
5023: } /* End individuals */
1.214 brouard 5024: /* wav and mw are no more changed */
1.223 brouard 5025:
1.214 brouard 5026:
1.126 brouard 5027: for(i=1; i<=imx; i++){
5028: for(mi=1; mi<wav[i];mi++){
5029: if (stepm <=0)
1.227 brouard 5030: dh[mi][i]=1;
1.126 brouard 5031: else{
1.260 brouard 5032: if (s[mw[mi+1][i]][i] > nlstate) { /* A death, but what if date is unknown? */
1.227 brouard 5033: if (agedc[i] < 2*AGESUP) {
5034: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
5035: if(j==0) j=1; /* Survives at least one month after exam */
5036: else if(j<0){
5037: nberr++;
5038: printf("Error! Negative delay (%d to death) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5039: j=1; /* Temporary Dangerous patch */
5040: printf(" We assumed that the date of interview was correct (and not the date of death) and postponed the death %d month(s) (one stepm) after the interview. You MUST fix the contradiction between dates.\n",stepm);
5041: fprintf(ficlog,"Error! Negative delay (%d to death) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5042: fprintf(ficlog," We assumed that the date of interview was correct (and not the date of death) and postponed the death %d month(s) (one stepm) after the interview. You MUST fix the contradiction between dates.\n",stepm);
5043: }
5044: k=k+1;
5045: if (j >= jmax){
5046: jmax=j;
5047: ijmax=i;
5048: }
5049: if (j <= jmin){
5050: jmin=j;
5051: ijmin=i;
5052: }
5053: sum=sum+j;
5054: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
5055: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
5056: }
5057: }
5058: else{
5059: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 5060: /* if (j<0) printf("%d %lf %lf %d %d %d\n", i,agev[mw[mi+1][i]][i], agev[mw[mi][i]][i],j,s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]); */
1.223 brouard 5061:
1.227 brouard 5062: k=k+1;
5063: if (j >= jmax) {
5064: jmax=j;
5065: ijmax=i;
5066: }
5067: else if (j <= jmin){
5068: jmin=j;
5069: ijmin=i;
5070: }
5071: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
5072: /*printf("%d %lf %d %d %d\n", i,agev[mw[mi][i]][i],j,s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);*/
5073: if(j<0){
5074: nberr++;
5075: printf("Error! Negative delay (%d) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5076: fprintf(ficlog,"Error! Negative delay (%d) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5077: }
5078: sum=sum+j;
5079: }
5080: jk= j/stepm;
5081: jl= j -jk*stepm;
5082: ju= j -(jk+1)*stepm;
5083: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
5084: if(jl==0){
5085: dh[mi][i]=jk;
5086: bh[mi][i]=0;
5087: }else{ /* We want a negative bias in order to only have interpolation ie
5088: * to avoid the price of an extra matrix product in likelihood */
5089: dh[mi][i]=jk+1;
5090: bh[mi][i]=ju;
5091: }
5092: }else{
5093: if(jl <= -ju){
5094: dh[mi][i]=jk;
5095: bh[mi][i]=jl; /* bias is positive if real duration
5096: * is higher than the multiple of stepm and negative otherwise.
5097: */
5098: }
5099: else{
5100: dh[mi][i]=jk+1;
5101: bh[mi][i]=ju;
5102: }
5103: if(dh[mi][i]==0){
5104: dh[mi][i]=1; /* At least one step */
5105: bh[mi][i]=ju; /* At least one step */
5106: /* printf(" bh=%d ju=%d jl=%d dh=%d jk=%d stepm=%d %d\n",bh[mi][i],ju,jl,dh[mi][i],jk,stepm,i);*/
5107: }
5108: } /* end if mle */
1.126 brouard 5109: }
5110: } /* end wave */
5111: }
5112: jmean=sum/k;
5113: printf("Delay (in months) between two waves Min=%d (for indiviudal %ld) Max=%d (%ld) Mean=%f\n\n ",jmin, num[ijmin], jmax, num[ijmax], jmean);
1.141 brouard 5114: fprintf(ficlog,"Delay (in months) between two waves Min=%d (for indiviudal %d) Max=%d (%d) Mean=%f\n\n ",jmin, ijmin, jmax, ijmax, jmean);
1.227 brouard 5115: }
1.126 brouard 5116:
5117: /*********** Tricode ****************************/
1.220 brouard 5118: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.242 brouard 5119: {
5120: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
5121: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
5122: * Boring subroutine which should only output nbcode[Tvar[j]][k]
5123: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
5124: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
5125: */
1.130 brouard 5126:
1.242 brouard 5127: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
5128: int modmaxcovj=0; /* Modality max of covariates j */
5129: int cptcode=0; /* Modality max of covariates j */
5130: int modmincovj=0; /* Modality min of covariates j */
1.145 brouard 5131:
5132:
1.242 brouard 5133: /* cptcoveff=0; */
5134: /* *cptcov=0; */
1.126 brouard 5135:
1.242 brouard 5136: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 5137:
1.242 brouard 5138: /* Loop on covariates without age and products and no quantitative variable */
5139: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
5140: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
5141: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
5142: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
5143: switch(Fixed[k]) {
5144: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
5145: for (i=1; i<=imx; i++) { /* Loop on individuals: reads the data file to get the maximum value of the modality of this covariate Vj*/
5146: ij=(int)(covar[Tvar[k]][i]);
5147: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
5148: * If product of Vn*Vm, still boolean *:
5149: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
5150: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
5151: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
5152: modality of the nth covariate of individual i. */
5153: if (ij > modmaxcovj)
5154: modmaxcovj=ij;
5155: else if (ij < modmincovj)
5156: modmincovj=ij;
5157: if ((ij < -1) && (ij > NCOVMAX)){
5158: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
5159: exit(1);
5160: }else
5161: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
5162: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
5163: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
5164: /* getting the maximum value of the modality of the covariate
5165: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
5166: female ies 1, then modmaxcovj=1.
5167: */
5168: } /* end for loop on individuals i */
5169: printf(" Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5170: fprintf(ficlog," Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5171: cptcode=modmaxcovj;
5172: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
5173: /*for (i=0; i<=cptcode; i++) {*/
5174: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
5175: printf("Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5176: fprintf(ficlog, "Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5177: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
5178: if( j != -1){
5179: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
5180: covariate for which somebody answered excluding
5181: undefined. Usually 2: 0 and 1. */
5182: }
5183: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
5184: covariate for which somebody answered including
5185: undefined. Usually 3: -1, 0 and 1. */
5186: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
5187: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
5188: } /* Ndum[-1] number of undefined modalities */
1.231 brouard 5189:
1.242 brouard 5190: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
5191: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
5192: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
5193: /* modmincovj=3; modmaxcovj = 7; */
5194: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
5195: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
5196: /* defining two dummy variables: variables V1_1 and V1_2.*/
5197: /* nbcode[Tvar[j]][ij]=k; */
5198: /* nbcode[Tvar[j]][1]=0; */
5199: /* nbcode[Tvar[j]][2]=1; */
5200: /* nbcode[Tvar[j]][3]=2; */
5201: /* To be continued (not working yet). */
5202: ij=0; /* ij is similar to i but can jump over null modalities */
5203: for (i=modmincovj; i<=modmaxcovj; i++) { /* i= 1 to 2 for dichotomous, or from 1 to 3 or from -1 or 0 to 1 currently*/
5204: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
5205: break;
5206: }
5207: ij++;
5208: nbcode[Tvar[k]][ij]=i; /* stores the original value of modality i in an array nbcode, ij modality from 1 to last non-nul modality. nbcode[1][1]=0 nbcode[1][2]=1*/
5209: cptcode = ij; /* New max modality for covar j */
5210: } /* end of loop on modality i=-1 to 1 or more */
5211: break;
5212: case 1: /* Testing on varying covariate, could be simple and
5213: * should look at waves or product of fixed *
5214: * varying. No time to test -1, assuming 0 and 1 only */
5215: ij=0;
5216: for(i=0; i<=1;i++){
5217: nbcode[Tvar[k]][++ij]=i;
5218: }
5219: break;
5220: default:
5221: break;
5222: } /* end switch */
5223: } /* end dummy test */
5224:
5225: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
5226: /* /\*recode from 0 *\/ */
5227: /* k is a modality. If we have model=V1+V1*sex */
5228: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
5229: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
5230: /* } */
5231: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
5232: /* if (ij > ncodemax[j]) { */
5233: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5234: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5235: /* break; */
5236: /* } */
5237: /* } /\* end of loop on modality k *\/ */
5238: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
5239:
5240: for (k=-1; k< maxncov; k++) Ndum[k]=0;
5241: /* Look at fixed dummy (single or product) covariates to check empty modalities */
5242: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
5243: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
5244: ij=Tvar[i]; /* Tvar 5,4,3,6,5,7,1,4 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V4*age */
5245: Ndum[ij]++; /* Count the # of 1, 2 etc: {1,1,1,2,2,1,1} because V1 once, V2 once, two V4 and V5 in above */
5246: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
5247: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
5248:
5249: ij=0;
5250: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
5251: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
5252: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
5253: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
5254: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
5255: /* If product not in single variable we don't print results */
5256: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
5257: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
5258: Tvaraff[ij]=Tvar[k]; /* For printing combination *//* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, Tvar {5, 4, 3, 6, 5, 2, 7, 1, 1} Tvaraff={4, 3, 1} V4, V3, V1*/
5259: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
5260: TmodelInvind[ij]=Tvar[k]- ncovcol-nqv; /* Inverse TmodelInvind[2=V4]=2 second dummy varying cov (V4)4-1-1 {0, 2, 1, } TmodelInvind[3]=1 */
5261: if(Fixed[k]!=0)
5262: anyvaryingduminmodel=1;
5263: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
5264: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
5265: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
5266: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
5267: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
5268: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
5269: }
5270: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
5271: /* ij--; */
5272: /* cptcoveff=ij; /\*Number of total covariates*\/ */
5273: *cptcov=ij; /*Number of total real effective covariates: effective
5274: * because they can be excluded from the model and real
5275: * if in the model but excluded because missing values, but how to get k from ij?*/
5276: for(j=ij+1; j<= cptcovt; j++){
5277: Tvaraff[j]=0;
5278: Tmodelind[j]=0;
5279: }
5280: for(j=ntveff+1; j<= cptcovt; j++){
5281: TmodelInvind[j]=0;
5282: }
5283: /* To be sorted */
5284: ;
5285: }
1.126 brouard 5286:
1.145 brouard 5287:
1.126 brouard 5288: /*********** Health Expectancies ****************/
5289:
1.235 brouard 5290: void evsij(double ***eij, double x[], int nlstate, int stepm, int bage, int fage, double **oldm, double **savm, int cij, int estepm,char strstart[], int nres )
1.126 brouard 5291:
5292: {
5293: /* Health expectancies, no variances */
1.164 brouard 5294: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 5295: int nhstepma, nstepma; /* Decreasing with age */
5296: double age, agelim, hf;
5297: double ***p3mat;
5298: double eip;
5299:
1.238 brouard 5300: /* pstamp(ficreseij); */
1.126 brouard 5301: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
5302: fprintf(ficreseij,"# Age");
5303: for(i=1; i<=nlstate;i++){
5304: for(j=1; j<=nlstate;j++){
5305: fprintf(ficreseij," e%1d%1d ",i,j);
5306: }
5307: fprintf(ficreseij," e%1d. ",i);
5308: }
5309: fprintf(ficreseij,"\n");
5310:
5311:
5312: if(estepm < stepm){
5313: printf ("Problem %d lower than %d\n",estepm, stepm);
5314: }
5315: else hstepm=estepm;
5316: /* We compute the life expectancy from trapezoids spaced every estepm months
5317: * This is mainly to measure the difference between two models: for example
5318: * if stepm=24 months pijx are given only every 2 years and by summing them
5319: * we are calculating an estimate of the Life Expectancy assuming a linear
5320: * progression in between and thus overestimating or underestimating according
5321: * to the curvature of the survival function. If, for the same date, we
5322: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5323: * to compare the new estimate of Life expectancy with the same linear
5324: * hypothesis. A more precise result, taking into account a more precise
5325: * curvature will be obtained if estepm is as small as stepm. */
5326:
5327: /* For example we decided to compute the life expectancy with the smallest unit */
5328: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5329: nhstepm is the number of hstepm from age to agelim
5330: nstepm is the number of stepm from age to agelin.
5331: Look at hpijx to understand the reason of that which relies in memory size
5332: and note for a fixed period like estepm months */
5333: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5334: survival function given by stepm (the optimization length). Unfortunately it
5335: means that if the survival funtion is printed only each two years of age and if
5336: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5337: results. So we changed our mind and took the option of the best precision.
5338: */
5339: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5340:
5341: agelim=AGESUP;
5342: /* If stepm=6 months */
5343: /* Computed by stepm unit matrices, product of hstepm matrices, stored
5344: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
5345:
5346: /* nhstepm age range expressed in number of stepm */
5347: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5348: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5349: /* if (stepm >= YEARM) hstepm=1;*/
5350: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5351: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5352:
5353: for (age=bage; age<=fage; age ++){
5354: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5355: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5356: /* if (stepm >= YEARM) hstepm=1;*/
5357: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
5358:
5359: /* If stepm=6 months */
5360: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5361: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5362:
1.235 brouard 5363: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 5364:
5365: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5366:
5367: printf("%d|",(int)age);fflush(stdout);
5368: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5369:
5370: /* Computing expectancies */
5371: for(i=1; i<=nlstate;i++)
5372: for(j=1; j<=nlstate;j++)
5373: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5374: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
5375:
5376: /* if((int)age==70)printf("i=%2d,j=%2d,h=%2d,age=%3d,%9.4f,%9.4f,%9.4f\n",i,j,h,(int)age,p3mat[i][j][h],hf,eij[i][j][(int)age]);*/
5377:
5378: }
5379:
5380: fprintf(ficreseij,"%3.0f",age );
5381: for(i=1; i<=nlstate;i++){
5382: eip=0;
5383: for(j=1; j<=nlstate;j++){
5384: eip +=eij[i][j][(int)age];
5385: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
5386: }
5387: fprintf(ficreseij,"%9.4f", eip );
5388: }
5389: fprintf(ficreseij,"\n");
5390:
5391: }
5392: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5393: printf("\n");
5394: fprintf(ficlog,"\n");
5395:
5396: }
5397:
1.235 brouard 5398: void cvevsij(double ***eij, double x[], int nlstate, int stepm, int bage, int fage, double **oldm, double **savm, int cij, int estepm,double delti[],double **matcov,char strstart[], int nres )
1.126 brouard 5399:
5400: {
5401: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5402: to initial status i, ei. .
1.126 brouard 5403: */
5404: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5405: int nhstepma, nstepma; /* Decreasing with age */
5406: double age, agelim, hf;
5407: double ***p3matp, ***p3matm, ***varhe;
5408: double **dnewm,**doldm;
5409: double *xp, *xm;
5410: double **gp, **gm;
5411: double ***gradg, ***trgradg;
5412: int theta;
5413:
5414: double eip, vip;
5415:
5416: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5417: xp=vector(1,npar);
5418: xm=vector(1,npar);
5419: dnewm=matrix(1,nlstate*nlstate,1,npar);
5420: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5421:
5422: pstamp(ficresstdeij);
5423: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5424: fprintf(ficresstdeij,"# Age");
5425: for(i=1; i<=nlstate;i++){
5426: for(j=1; j<=nlstate;j++)
5427: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5428: fprintf(ficresstdeij," e%1d. ",i);
5429: }
5430: fprintf(ficresstdeij,"\n");
5431:
5432: pstamp(ficrescveij);
5433: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5434: fprintf(ficrescveij,"# Age");
5435: for(i=1; i<=nlstate;i++)
5436: for(j=1; j<=nlstate;j++){
5437: cptj= (j-1)*nlstate+i;
5438: for(i2=1; i2<=nlstate;i2++)
5439: for(j2=1; j2<=nlstate;j2++){
5440: cptj2= (j2-1)*nlstate+i2;
5441: if(cptj2 <= cptj)
5442: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5443: }
5444: }
5445: fprintf(ficrescveij,"\n");
5446:
5447: if(estepm < stepm){
5448: printf ("Problem %d lower than %d\n",estepm, stepm);
5449: }
5450: else hstepm=estepm;
5451: /* We compute the life expectancy from trapezoids spaced every estepm months
5452: * This is mainly to measure the difference between two models: for example
5453: * if stepm=24 months pijx are given only every 2 years and by summing them
5454: * we are calculating an estimate of the Life Expectancy assuming a linear
5455: * progression in between and thus overestimating or underestimating according
5456: * to the curvature of the survival function. If, for the same date, we
5457: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5458: * to compare the new estimate of Life expectancy with the same linear
5459: * hypothesis. A more precise result, taking into account a more precise
5460: * curvature will be obtained if estepm is as small as stepm. */
5461:
5462: /* For example we decided to compute the life expectancy with the smallest unit */
5463: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5464: nhstepm is the number of hstepm from age to agelim
5465: nstepm is the number of stepm from age to agelin.
5466: Look at hpijx to understand the reason of that which relies in memory size
5467: and note for a fixed period like estepm months */
5468: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5469: survival function given by stepm (the optimization length). Unfortunately it
5470: means that if the survival funtion is printed only each two years of age and if
5471: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5472: results. So we changed our mind and took the option of the best precision.
5473: */
5474: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5475:
5476: /* If stepm=6 months */
5477: /* nhstepm age range expressed in number of stepm */
5478: agelim=AGESUP;
5479: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5480: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5481: /* if (stepm >= YEARM) hstepm=1;*/
5482: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5483:
5484: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5485: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5486: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5487: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5488: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5489: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5490:
5491: for (age=bage; age<=fage; age ++){
5492: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5493: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5494: /* if (stepm >= YEARM) hstepm=1;*/
5495: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5496:
1.126 brouard 5497: /* If stepm=6 months */
5498: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5499: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5500:
5501: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5502:
1.126 brouard 5503: /* Computing Variances of health expectancies */
5504: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5505: decrease memory allocation */
5506: for(theta=1; theta <=npar; theta++){
5507: for(i=1; i<=npar; i++){
1.222 brouard 5508: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5509: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5510: }
1.235 brouard 5511: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5512: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5513:
1.126 brouard 5514: for(j=1; j<= nlstate; j++){
1.222 brouard 5515: for(i=1; i<=nlstate; i++){
5516: for(h=0; h<=nhstepm-1; h++){
5517: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5518: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5519: }
5520: }
1.126 brouard 5521: }
1.218 brouard 5522:
1.126 brouard 5523: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5524: for(h=0; h<=nhstepm-1; h++){
5525: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5526: }
1.126 brouard 5527: }/* End theta */
5528:
5529:
5530: for(h=0; h<=nhstepm-1; h++)
5531: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5532: for(theta=1; theta <=npar; theta++)
5533: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5534:
1.218 brouard 5535:
1.222 brouard 5536: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5537: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5538: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5539:
1.222 brouard 5540: printf("%d|",(int)age);fflush(stdout);
5541: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5542: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5543: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5544: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5545: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5546: for(ij=1;ij<=nlstate*nlstate;ij++)
5547: for(ji=1;ji<=nlstate*nlstate;ji++)
5548: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5549: }
5550: }
1.218 brouard 5551:
1.126 brouard 5552: /* Computing expectancies */
1.235 brouard 5553: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5554: for(i=1; i<=nlstate;i++)
5555: for(j=1; j<=nlstate;j++)
1.222 brouard 5556: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5557: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5558:
1.222 brouard 5559: /* if((int)age==70)printf("i=%2d,j=%2d,h=%2d,age=%3d,%9.4f,%9.4f,%9.4f\n",i,j,h,(int)age,p3mat[i][j][h],hf,eij[i][j][(int)age]);*/
1.218 brouard 5560:
1.222 brouard 5561: }
1.218 brouard 5562:
1.126 brouard 5563: fprintf(ficresstdeij,"%3.0f",age );
5564: for(i=1; i<=nlstate;i++){
5565: eip=0.;
5566: vip=0.;
5567: for(j=1; j<=nlstate;j++){
1.222 brouard 5568: eip += eij[i][j][(int)age];
5569: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5570: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5571: fprintf(ficresstdeij," %9.4f (%.4f)", eij[i][j][(int)age], sqrt(varhe[(j-1)*nlstate+i][(j-1)*nlstate+i][(int)age]) );
1.126 brouard 5572: }
5573: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5574: }
5575: fprintf(ficresstdeij,"\n");
1.218 brouard 5576:
1.126 brouard 5577: fprintf(ficrescveij,"%3.0f",age );
5578: for(i=1; i<=nlstate;i++)
5579: for(j=1; j<=nlstate;j++){
1.222 brouard 5580: cptj= (j-1)*nlstate+i;
5581: for(i2=1; i2<=nlstate;i2++)
5582: for(j2=1; j2<=nlstate;j2++){
5583: cptj2= (j2-1)*nlstate+i2;
5584: if(cptj2 <= cptj)
5585: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5586: }
1.126 brouard 5587: }
5588: fprintf(ficrescveij,"\n");
1.218 brouard 5589:
1.126 brouard 5590: }
5591: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5592: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5593: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5594: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5595: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5596: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5597: printf("\n");
5598: fprintf(ficlog,"\n");
1.218 brouard 5599:
1.126 brouard 5600: free_vector(xm,1,npar);
5601: free_vector(xp,1,npar);
5602: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5603: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5604: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5605: }
1.218 brouard 5606:
1.126 brouard 5607: /************ Variance ******************/
1.235 brouard 5608: void varevsij(char optionfilefiname[], double ***vareij, double **matcov, double x[], double delti[], int nlstate, int stepm, double bage, double fage, double **oldm, double **savm, double **prlim, double ftolpl, int *ncvyearp, int ij, int estepm, int cptcov, int cptcod, int popbased, int mobilav, char strstart[], int nres)
1.218 brouard 5609: {
5610: /* Variance of health expectancies */
5611: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5612: /* double **newm;*/
5613: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5614:
5615: /* int movingaverage(); */
5616: double **dnewm,**doldm;
5617: double **dnewmp,**doldmp;
5618: int i, j, nhstepm, hstepm, h, nstepm ;
5619: int k;
5620: double *xp;
5621: double **gp, **gm; /* for var eij */
5622: double ***gradg, ***trgradg; /*for var eij */
5623: double **gradgp, **trgradgp; /* for var p point j */
5624: double *gpp, *gmp; /* for var p point j */
5625: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5626: double ***p3mat;
5627: double age,agelim, hf;
5628: /* double ***mobaverage; */
5629: int theta;
5630: char digit[4];
5631: char digitp[25];
5632:
5633: char fileresprobmorprev[FILENAMELENGTH];
5634:
5635: if(popbased==1){
5636: if(mobilav!=0)
5637: strcpy(digitp,"-POPULBASED-MOBILAV_");
5638: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5639: }
5640: else
5641: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5642:
1.218 brouard 5643: /* if (mobilav!=0) { */
5644: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5645: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5646: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5647: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5648: /* } */
5649: /* } */
5650:
5651: strcpy(fileresprobmorprev,"PRMORPREV-");
5652: sprintf(digit,"%-d",ij);
5653: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5654: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5655: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5656: strcat(fileresprobmorprev,fileresu);
5657: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5658: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5659: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5660: }
5661: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5662: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5663: pstamp(ficresprobmorprev);
5664: fprintf(ficresprobmorprev,"# probabilities of dying before estepm=%d months for people of exact age and weighted probabilities w1*p1j+w2*p2j+... stand dev in()\n",estepm);
1.238 brouard 5665: fprintf(ficresprobmorprev,"# Selected quantitative variables and dummies");
5666: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
5667: fprintf(ficresprobmorprev," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
5668: }
5669: for(j=1;j<=cptcoveff;j++)
5670: fprintf(ficresprobmorprev,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(ij,j)]);
5671: fprintf(ficresprobmorprev,"\n");
5672:
1.218 brouard 5673: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5674: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5675: fprintf(ficresprobmorprev," p.%-d SE",j);
5676: for(i=1; i<=nlstate;i++)
5677: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5678: }
5679: fprintf(ficresprobmorprev,"\n");
5680:
5681: fprintf(ficgp,"\n# Routine varevsij");
5682: fprintf(ficgp,"\nunset title \n");
5683: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5684: fprintf(fichtm,"\n<li><h4> Computing probabilities of dying over estepm months as a weighted average (i.e global mortality independent of initial healh state)</h4></li>\n");
5685: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5686: /* } */
5687: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5688: pstamp(ficresvij);
5689: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5690: if(popbased==1)
5691: fprintf(ficresvij,"the age specific prevalence observed (cross-sectionally) in the population i.e cross-sectionally\n in each health state (popbased=1) (mobilav=%d\n",mobilav);
5692: else
5693: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5694: fprintf(ficresvij,"# Age");
5695: for(i=1; i<=nlstate;i++)
5696: for(j=1; j<=nlstate;j++)
5697: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5698: fprintf(ficresvij,"\n");
5699:
5700: xp=vector(1,npar);
5701: dnewm=matrix(1,nlstate,1,npar);
5702: doldm=matrix(1,nlstate,1,nlstate);
5703: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5704: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5705:
5706: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5707: gpp=vector(nlstate+1,nlstate+ndeath);
5708: gmp=vector(nlstate+1,nlstate+ndeath);
5709: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5710:
1.218 brouard 5711: if(estepm < stepm){
5712: printf ("Problem %d lower than %d\n",estepm, stepm);
5713: }
5714: else hstepm=estepm;
5715: /* For example we decided to compute the life expectancy with the smallest unit */
5716: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5717: nhstepm is the number of hstepm from age to agelim
5718: nstepm is the number of stepm from age to agelim.
5719: Look at function hpijx to understand why because of memory size limitations,
5720: we decided (b) to get a life expectancy respecting the most precise curvature of the
5721: survival function given by stepm (the optimization length). Unfortunately it
5722: means that if the survival funtion is printed every two years of age and if
5723: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5724: results. So we changed our mind and took the option of the best precision.
5725: */
5726: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5727: agelim = AGESUP;
5728: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5729: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5730: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5731: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5732: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5733: gp=matrix(0,nhstepm,1,nlstate);
5734: gm=matrix(0,nhstepm,1,nlstate);
5735:
5736:
5737: for(theta=1; theta <=npar; theta++){
5738: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5739: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5740: }
5741:
1.242 brouard 5742: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5743:
5744: if (popbased==1) {
5745: if(mobilav ==0){
5746: for(i=1; i<=nlstate;i++)
5747: prlim[i][i]=probs[(int)age][i][ij];
5748: }else{ /* mobilav */
5749: for(i=1; i<=nlstate;i++)
5750: prlim[i][i]=mobaverage[(int)age][i][ij];
5751: }
5752: }
5753:
1.235 brouard 5754: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres); /* Returns p3mat[i][j][h] for h=1 to nhstepm */
1.218 brouard 5755: for(j=1; j<= nlstate; j++){
5756: for(h=0; h<=nhstepm; h++){
5757: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5758: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5759: }
5760: }
5761: /* Next for computing probability of death (h=1 means
5762: computed over hstepm matrices product = hstepm*stepm months)
5763: as a weighted average of prlim.
5764: */
5765: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5766: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5767: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5768: }
5769: /* end probability of death */
5770:
5771: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5772: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5773:
1.242 brouard 5774: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nres);
1.218 brouard 5775:
5776: if (popbased==1) {
5777: if(mobilav ==0){
5778: for(i=1; i<=nlstate;i++)
5779: prlim[i][i]=probs[(int)age][i][ij];
5780: }else{ /* mobilav */
5781: for(i=1; i<=nlstate;i++)
5782: prlim[i][i]=mobaverage[(int)age][i][ij];
5783: }
5784: }
5785:
1.235 brouard 5786: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5787:
5788: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5789: for(h=0; h<=nhstepm; h++){
5790: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5791: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5792: }
5793: }
5794: /* This for computing probability of death (h=1 means
5795: computed over hstepm matrices product = hstepm*stepm months)
5796: as a weighted average of prlim.
5797: */
5798: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5799: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5800: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5801: }
5802: /* end probability of death */
5803:
5804: for(j=1; j<= nlstate; j++) /* vareij */
5805: for(h=0; h<=nhstepm; h++){
5806: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5807: }
5808:
5809: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5810: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5811: }
5812:
5813: } /* End theta */
5814:
5815: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5816:
5817: for(h=0; h<=nhstepm; h++) /* veij */
5818: for(j=1; j<=nlstate;j++)
5819: for(theta=1; theta <=npar; theta++)
5820: trgradg[h][j][theta]=gradg[h][theta][j];
5821:
5822: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5823: for(theta=1; theta <=npar; theta++)
5824: trgradgp[j][theta]=gradgp[theta][j];
5825:
5826:
5827: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5828: for(i=1;i<=nlstate;i++)
5829: for(j=1;j<=nlstate;j++)
5830: vareij[i][j][(int)age] =0.;
5831:
5832: for(h=0;h<=nhstepm;h++){
5833: for(k=0;k<=nhstepm;k++){
5834: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5835: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5836: for(i=1;i<=nlstate;i++)
5837: for(j=1;j<=nlstate;j++)
5838: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5839: }
5840: }
5841:
5842: /* pptj */
5843: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5844: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5845: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5846: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5847: varppt[j][i]=doldmp[j][i];
5848: /* end ppptj */
5849: /* x centered again */
5850:
1.242 brouard 5851: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5852:
5853: if (popbased==1) {
5854: if(mobilav ==0){
5855: for(i=1; i<=nlstate;i++)
5856: prlim[i][i]=probs[(int)age][i][ij];
5857: }else{ /* mobilav */
5858: for(i=1; i<=nlstate;i++)
5859: prlim[i][i]=mobaverage[(int)age][i][ij];
5860: }
5861: }
5862:
5863: /* This for computing probability of death (h=1 means
5864: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5865: as a weighted average of prlim.
5866: */
1.235 brouard 5867: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5868: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5869: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5870: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5871: }
5872: /* end probability of death */
5873:
5874: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5875: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5876: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5877: for(i=1; i<=nlstate;i++){
5878: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5879: }
5880: }
5881: fprintf(ficresprobmorprev,"\n");
5882:
5883: fprintf(ficresvij,"%.0f ",age );
5884: for(i=1; i<=nlstate;i++)
5885: for(j=1; j<=nlstate;j++){
5886: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5887: }
5888: fprintf(ficresvij,"\n");
5889: free_matrix(gp,0,nhstepm,1,nlstate);
5890: free_matrix(gm,0,nhstepm,1,nlstate);
5891: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5892: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5893: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5894: } /* End age */
5895: free_vector(gpp,nlstate+1,nlstate+ndeath);
5896: free_vector(gmp,nlstate+1,nlstate+ndeath);
5897: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5898: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5899: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5900: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5901: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5902: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5903: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5904: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5905: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5906: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5907: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5908: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5909: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5910: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5911: fprintf(fichtm,"\n<br> Probability is computed over estepm=%d months. <br> <img src=\"%s%s.svg\"> <br>\n", estepm,subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5912: /* fprintf(fichtm,"\n<br> Probability is computed over estepm=%d months and then divided by estepm and multiplied by %.0f in order to have the probability to die over a year <br> <img src=\"varmuptjgr%s%s.svg\"> <br>\n", stepm,YEARM,digitp,digit);
1.126 brouard 5913: */
1.218 brouard 5914: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5915: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5916:
1.218 brouard 5917: free_vector(xp,1,npar);
5918: free_matrix(doldm,1,nlstate,1,nlstate);
5919: free_matrix(dnewm,1,nlstate,1,npar);
5920: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5921: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5922: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5923: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5924: fclose(ficresprobmorprev);
5925: fflush(ficgp);
5926: fflush(fichtm);
5927: } /* end varevsij */
1.126 brouard 5928:
5929: /************ Variance of prevlim ******************/
1.235 brouard 5930: void varprevlim(char fileres[], double **varpl, double **matcov, double x[], double delti[], int nlstate, int stepm, double bage, double fage, double **oldm, double **savm, double **prlim, double ftolpl, int *ncvyearp, int ij, char strstart[], int nres)
1.126 brouard 5931: {
1.205 brouard 5932: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5933: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5934:
1.126 brouard 5935: double **dnewm,**doldm;
5936: int i, j, nhstepm, hstepm;
5937: double *xp;
5938: double *gp, *gm;
5939: double **gradg, **trgradg;
1.208 brouard 5940: double **mgm, **mgp;
1.126 brouard 5941: double age,agelim;
5942: int theta;
5943:
5944: pstamp(ficresvpl);
5945: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
1.241 brouard 5946: fprintf(ficresvpl,"# Age ");
5947: if(nresult >=1)
5948: fprintf(ficresvpl," Result# ");
1.126 brouard 5949: for(i=1; i<=nlstate;i++)
5950: fprintf(ficresvpl," %1d-%1d",i,i);
5951: fprintf(ficresvpl,"\n");
5952:
5953: xp=vector(1,npar);
5954: dnewm=matrix(1,nlstate,1,npar);
5955: doldm=matrix(1,nlstate,1,nlstate);
5956:
5957: hstepm=1*YEARM; /* Every year of age */
5958: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5959: agelim = AGESUP;
5960: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5961: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5962: if (stepm >= YEARM) hstepm=1;
5963: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5964: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5965: mgp=matrix(1,npar,1,nlstate);
5966: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5967: gp=vector(1,nlstate);
5968: gm=vector(1,nlstate);
5969:
5970: for(theta=1; theta <=npar; theta++){
5971: for(i=1; i<=npar; i++){ /* Computes gradient */
5972: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5973: }
1.209 brouard 5974: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5975: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5976: else
1.235 brouard 5977: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5978: for(i=1;i<=nlstate;i++){
1.126 brouard 5979: gp[i] = prlim[i][i];
1.208 brouard 5980: mgp[theta][i] = prlim[i][i];
5981: }
1.126 brouard 5982: for(i=1; i<=npar; i++) /* Computes gradient */
5983: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5984: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5985: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5986: else
1.235 brouard 5987: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5988: for(i=1;i<=nlstate;i++){
1.126 brouard 5989: gm[i] = prlim[i][i];
1.208 brouard 5990: mgm[theta][i] = prlim[i][i];
5991: }
1.126 brouard 5992: for(i=1;i<=nlstate;i++)
5993: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5994: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5995: } /* End theta */
5996:
5997: trgradg =matrix(1,nlstate,1,npar);
5998:
5999: for(j=1; j<=nlstate;j++)
6000: for(theta=1; theta <=npar; theta++)
6001: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 6002: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
6003: /* printf("\nmgm mgp %d ",(int)age); */
6004: /* for(j=1; j<=nlstate;j++){ */
6005: /* printf(" %d ",j); */
6006: /* for(theta=1; theta <=npar; theta++) */
6007: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
6008: /* printf("\n "); */
6009: /* } */
6010: /* } */
6011: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
6012: /* printf("\n gradg %d ",(int)age); */
6013: /* for(j=1; j<=nlstate;j++){ */
6014: /* printf("%d ",j); */
6015: /* for(theta=1; theta <=npar; theta++) */
6016: /* printf("%d %lf ",theta,gradg[theta][j]); */
6017: /* printf("\n "); */
6018: /* } */
6019: /* } */
1.126 brouard 6020:
6021: for(i=1;i<=nlstate;i++)
6022: varpl[i][(int)age] =0.;
1.209 brouard 6023: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 6024: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6025: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
6026: }else{
1.126 brouard 6027: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6028: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 6029: }
1.126 brouard 6030: for(i=1;i<=nlstate;i++)
6031: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
6032:
6033: fprintf(ficresvpl,"%.0f ",age );
1.241 brouard 6034: if(nresult >=1)
6035: fprintf(ficresvpl,"%d ",nres );
1.126 brouard 6036: for(i=1; i<=nlstate;i++)
6037: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
6038: fprintf(ficresvpl,"\n");
6039: free_vector(gp,1,nlstate);
6040: free_vector(gm,1,nlstate);
1.208 brouard 6041: free_matrix(mgm,1,npar,1,nlstate);
6042: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 6043: free_matrix(gradg,1,npar,1,nlstate);
6044: free_matrix(trgradg,1,nlstate,1,npar);
6045: } /* End age */
6046:
6047: free_vector(xp,1,npar);
6048: free_matrix(doldm,1,nlstate,1,npar);
6049: free_matrix(dnewm,1,nlstate,1,nlstate);
6050:
6051: }
6052:
6053: /************ Variance of one-step probabilities ******************/
6054: void varprob(char optionfilefiname[], double **matcov, double x[], double delti[], int nlstate, double bage, double fage, int ij, int *Tvar, int **nbcode, int *ncodemax, char strstart[])
1.222 brouard 6055: {
6056: int i, j=0, k1, l1, tj;
6057: int k2, l2, j1, z1;
6058: int k=0, l;
6059: int first=1, first1, first2;
6060: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
6061: double **dnewm,**doldm;
6062: double *xp;
6063: double *gp, *gm;
6064: double **gradg, **trgradg;
6065: double **mu;
6066: double age, cov[NCOVMAX+1];
6067: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
6068: int theta;
6069: char fileresprob[FILENAMELENGTH];
6070: char fileresprobcov[FILENAMELENGTH];
6071: char fileresprobcor[FILENAMELENGTH];
6072: double ***varpij;
6073:
6074: strcpy(fileresprob,"PROB_");
6075: strcat(fileresprob,fileres);
6076: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
6077: printf("Problem with resultfile: %s\n", fileresprob);
6078: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
6079: }
6080: strcpy(fileresprobcov,"PROBCOV_");
6081: strcat(fileresprobcov,fileresu);
6082: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
6083: printf("Problem with resultfile: %s\n", fileresprobcov);
6084: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
6085: }
6086: strcpy(fileresprobcor,"PROBCOR_");
6087: strcat(fileresprobcor,fileresu);
6088: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
6089: printf("Problem with resultfile: %s\n", fileresprobcor);
6090: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
6091: }
6092: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6093: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6094: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6095: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6096: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6097: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6098: pstamp(ficresprob);
6099: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
6100: fprintf(ficresprob,"# Age");
6101: pstamp(ficresprobcov);
6102: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
6103: fprintf(ficresprobcov,"# Age");
6104: pstamp(ficresprobcor);
6105: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
6106: fprintf(ficresprobcor,"# Age");
1.126 brouard 6107:
6108:
1.222 brouard 6109: for(i=1; i<=nlstate;i++)
6110: for(j=1; j<=(nlstate+ndeath);j++){
6111: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
6112: fprintf(ficresprobcov," p%1d-%1d ",i,j);
6113: fprintf(ficresprobcor," p%1d-%1d ",i,j);
6114: }
6115: /* fprintf(ficresprob,"\n");
6116: fprintf(ficresprobcov,"\n");
6117: fprintf(ficresprobcor,"\n");
6118: */
6119: xp=vector(1,npar);
6120: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6121: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6122: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
6123: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
6124: first=1;
6125: fprintf(ficgp,"\n# Routine varprob");
6126: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
6127: fprintf(fichtm,"\n");
6128:
6129: fprintf(fichtm,"\n<li><h4> <a href=\"%s\">Matrix of variance-covariance of one-step probabilities (drawings)</a></h4> this page is important in order to visualize confidence intervals and especially correlation between disability and recovery, or more generally, way in and way back.</li>\n",optionfilehtmcov);
6130: fprintf(fichtmcov,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Matrix of variance-covariance of pairs of step probabilities</h4>\n",optionfilehtmcov, optionfilehtmcov);
6131: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 6132: and drawn. It helps understanding how is the covariance between two incidences.\
6133: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 6134: fprintf(fichtmcov,"\n<br> Contour plot corresponding to x'cov<sup>-1</sup>x = 4 (where x is the column vector (pij,pkl)) are drawn. \
1.126 brouard 6135: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
6136: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
6137: standard deviations wide on each axis. <br>\
6138: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
6139: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
6140: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
6141:
1.222 brouard 6142: cov[1]=1;
6143: /* tj=cptcoveff; */
1.225 brouard 6144: tj = (int) pow(2,cptcoveff);
1.222 brouard 6145: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
6146: j1=0;
1.224 brouard 6147: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 6148: if (cptcovn>0) {
6149: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 6150: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6151: fprintf(ficresprob, "**********\n#\n");
6152: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 6153: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6154: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 6155:
1.222 brouard 6156: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 6157: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6158: fprintf(ficgp, "**********\n#\n");
1.220 brouard 6159:
6160:
1.222 brouard 6161: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 6162: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6163: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6164:
1.222 brouard 6165: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 6166: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6167: fprintf(ficresprobcor, "**********\n#");
6168: if(invalidvarcomb[j1]){
6169: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
6170: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
6171: continue;
6172: }
6173: }
6174: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
6175: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6176: gp=vector(1,(nlstate)*(nlstate+ndeath));
6177: gm=vector(1,(nlstate)*(nlstate+ndeath));
6178: for (age=bage; age<=fage; age ++){
6179: cov[2]=age;
6180: if(nagesqr==1)
6181: cov[3]= age*age;
6182: for (k=1; k<=cptcovn;k++) {
6183: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
6184: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
6185: * 1 1 1 1 1
6186: * 2 2 1 1 1
6187: * 3 1 2 1 1
6188: */
6189: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
6190: }
6191: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
6192: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
6193: for (k=1; k<=cptcovprod;k++)
6194: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 6195:
6196:
1.222 brouard 6197: for(theta=1; theta <=npar; theta++){
6198: for(i=1; i<=npar; i++)
6199: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 6200:
1.222 brouard 6201: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 6202:
1.222 brouard 6203: k=0;
6204: for(i=1; i<= (nlstate); i++){
6205: for(j=1; j<=(nlstate+ndeath);j++){
6206: k=k+1;
6207: gp[k]=pmmij[i][j];
6208: }
6209: }
1.220 brouard 6210:
1.222 brouard 6211: for(i=1; i<=npar; i++)
6212: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 6213:
1.222 brouard 6214: pmij(pmmij,cov,ncovmodel,xp,nlstate);
6215: k=0;
6216: for(i=1; i<=(nlstate); i++){
6217: for(j=1; j<=(nlstate+ndeath);j++){
6218: k=k+1;
6219: gm[k]=pmmij[i][j];
6220: }
6221: }
1.220 brouard 6222:
1.222 brouard 6223: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
6224: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
6225: }
1.126 brouard 6226:
1.222 brouard 6227: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
6228: for(theta=1; theta <=npar; theta++)
6229: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 6230:
1.222 brouard 6231: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
6232: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 6233:
1.222 brouard 6234: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 6235:
1.222 brouard 6236: k=0;
6237: for(i=1; i<=(nlstate); i++){
6238: for(j=1; j<=(nlstate+ndeath);j++){
6239: k=k+1;
6240: mu[k][(int) age]=pmmij[i][j];
6241: }
6242: }
6243: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
6244: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
6245: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 6246:
1.222 brouard 6247: /*printf("\n%d ",(int)age);
6248: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6249: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6250: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6251: }*/
1.220 brouard 6252:
1.222 brouard 6253: fprintf(ficresprob,"\n%d ",(int)age);
6254: fprintf(ficresprobcov,"\n%d ",(int)age);
6255: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 6256:
1.222 brouard 6257: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
6258: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
6259: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6260: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
6261: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
6262: }
6263: i=0;
6264: for (k=1; k<=(nlstate);k++){
6265: for (l=1; l<=(nlstate+ndeath);l++){
6266: i++;
6267: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
6268: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
6269: for (j=1; j<=i;j++){
6270: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
6271: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
6272: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
6273: }
6274: }
6275: }/* end of loop for state */
6276: } /* end of loop for age */
6277: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
6278: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
6279: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6280: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6281:
6282: /* Confidence intervalle of pij */
6283: /*
6284: fprintf(ficgp,"\nunset parametric;unset label");
6285: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
6286: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
6287: fprintf(fichtm,"\n<br>Probability with confidence intervals expressed in year<sup>-1</sup> :<a href=\"pijgr%s.png\">pijgr%s.png</A>, ",optionfilefiname,optionfilefiname);
6288: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
6289: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
6290: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
6291: */
6292:
6293: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
6294: first1=1;first2=2;
6295: for (k2=1; k2<=(nlstate);k2++){
6296: for (l2=1; l2<=(nlstate+ndeath);l2++){
6297: if(l2==k2) continue;
6298: j=(k2-1)*(nlstate+ndeath)+l2;
6299: for (k1=1; k1<=(nlstate);k1++){
6300: for (l1=1; l1<=(nlstate+ndeath);l1++){
6301: if(l1==k1) continue;
6302: i=(k1-1)*(nlstate+ndeath)+l1;
6303: if(i<=j) continue;
6304: for (age=bage; age<=fage; age ++){
6305: if ((int)age %5==0){
6306: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
6307: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
6308: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
6309: mu1=mu[i][(int) age]/stepm*YEARM ;
6310: mu2=mu[j][(int) age]/stepm*YEARM;
6311: c12=cv12/sqrt(v1*v2);
6312: /* Computing eigen value of matrix of covariance */
6313: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6314: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6315: if ((lc2 <0) || (lc1 <0) ){
6316: if(first2==1){
6317: first1=0;
6318: printf("Strange: j1=%d One eigen value of 2x2 matrix of covariance is negative, lc1=%11.3e, lc2=%11.3e, v1=%11.3e, v2=%11.3e, cv12=%11.3e.\n It means that the matrix was not well estimated (varpij), for i=%2d, j=%2d, age=%4d .\n See files %s and %s. Probably WRONG RESULTS. See log file for details...\n", j1, lc1, lc2, v1, v2, cv12, i, j, (int)age,fileresprobcov, fileresprobcor);
6319: }
6320: fprintf(ficlog,"Strange: j1=%d One eigen value of 2x2 matrix of covariance is negative, lc1=%11.3e, lc2=%11.3e, v1=%11.3e, v2=%11.3e, cv12=%11.3e.\n It means that the matrix was not well estimated (varpij), for i=%2d, j=%2d, age=%4d .\n See files %s and %s. Probably WRONG RESULTS.\n", j1, lc1, lc2, v1, v2, cv12, i, j, (int)age,fileresprobcov, fileresprobcor);fflush(ficlog);
6321: /* lc1=fabs(lc1); */ /* If we want to have them positive */
6322: /* lc2=fabs(lc2); */
6323: }
1.220 brouard 6324:
1.222 brouard 6325: /* Eigen vectors */
6326: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
6327: /*v21=sqrt(1.-v11*v11); *//* error */
6328: v21=(lc1-v1)/cv12*v11;
6329: v12=-v21;
6330: v22=v11;
6331: tnalp=v21/v11;
6332: if(first1==1){
6333: first1=0;
6334: printf("%d %d%d-%d%d mu %.4e %.4e Var %.4e %.4e cor %.3f cov %.4e Eig %.3e %.3e 1stv %.3f %.3f tang %.3f\nOthers in log...\n",(int) age,k1,l1,k2,l2,mu1,mu2,v1,v2,c12,cv12,lc1,lc2,v11,v21,tnalp);
6335: }
6336: fprintf(ficlog,"%d %d%d-%d%d mu %.4e %.4e Var %.4e %.4e cor %.3f cov %.4e Eig %.3e %.3e 1stv %.3f %.3f tan %.3f\n",(int) age,k1,l1,k2,l2,mu1,mu2,v1,v2,c12,cv12,lc1,lc2,v11,v21,tnalp);
6337: /*printf(fignu*/
6338: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
6339: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
6340: if(first==1){
6341: first=0;
6342: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
6343: fprintf(ficgp,"\nset parametric;unset label");
6344: fprintf(ficgp,"\nset log y;set log x; set xlabel \"p%1d%1d (year-1)\";set ylabel \"p%1d%1d (year-1)\"",k1,l1,k2,l2);
6345: fprintf(ficgp,"\nset ter svg size 640, 480");
6346: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 6347: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 6348: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 6349: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
6350: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6351: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6352: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
6353: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6354: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6355: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6356: fprintf(ficgp,"\nplot [-pi:pi] %11.3e+ %.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)), %11.3e +%.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)) not", \
6357: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6358: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6359: }else{
6360: first=0;
6361: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
6362: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6363: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6364: fprintf(ficgp,"\nreplot %11.3e+ %.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)), %11.3e +%.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)) not", \
6365: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6366: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6367: }/* if first */
6368: } /* age mod 5 */
6369: } /* end loop age */
6370: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6371: first=1;
6372: } /*l12 */
6373: } /* k12 */
6374: } /*l1 */
6375: }/* k1 */
6376: } /* loop on combination of covariates j1 */
6377: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
6378: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
6379: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6380: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
6381: free_vector(xp,1,npar);
6382: fclose(ficresprob);
6383: fclose(ficresprobcov);
6384: fclose(ficresprobcor);
6385: fflush(ficgp);
6386: fflush(fichtmcov);
6387: }
1.126 brouard 6388:
6389:
6390: /******************* Printing html file ***********/
1.201 brouard 6391: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 6392: int lastpass, int stepm, int weightopt, char model[],\
6393: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.258 brouard 6394: int popforecast, int mobilav, int prevfcast, int mobilavproj, int backcast, int estepm , \
1.213 brouard 6395: double jprev1, double mprev1,double anprev1, double dateprev1, \
6396: double jprev2, double mprev2,double anprev2, double dateprev2){
1.237 brouard 6397: int jj1, k1, i1, cpt, k4, nres;
1.126 brouard 6398:
6399: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
6400: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
6401: </ul>");
1.237 brouard 6402: fprintf(fichtm,"<ul><li> model=1+age+%s\n \
6403: </ul>", model);
1.214 brouard 6404: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
6405: fprintf(fichtm,"<li>- Observed frequency between two states (during the period defined between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf): <a href=\"%s\">%s</a> (html file)<br/>\n",
6406: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
6407: fprintf(fichtm,"<li> - Observed prevalence in each state (during the period defined between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf): <a href=\"%s\">%s</a> (html file) ",
1.213 brouard 6408: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
6409: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 6410: fprintf(fichtm,"\
6411: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 6412: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6413: fprintf(fichtm,"\
1.217 brouard 6414: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6415: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6416: fprintf(fichtm,"\
1.126 brouard 6417: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6418: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6419: fprintf(fichtm,"\
1.217 brouard 6420: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6421: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6422: fprintf(fichtm,"\
1.211 brouard 6423: - (a) Life expectancies by health status at initial age, e<sub>i.</sub> (b) health expectancies by health status at initial age, e<sub>ij</sub> . If one or more covariates are included, specific tables for each value of the covariate are output in sequences within the same file (estepm=%2d months): \
1.126 brouard 6424: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6425: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6426: if(prevfcast==1){
6427: fprintf(fichtm,"\
6428: - Prevalence projections by age and states: \
1.201 brouard 6429: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6430: }
1.126 brouard 6431:
1.222 brouard 6432: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 6433:
1.225 brouard 6434: m=pow(2,cptcoveff);
1.222 brouard 6435: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6436:
1.222 brouard 6437: jj1=0;
1.237 brouard 6438:
6439: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.241 brouard 6440: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
1.253 brouard 6441: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6442: continue;
1.220 brouard 6443:
1.222 brouard 6444: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6445: jj1++;
6446: if (cptcovn > 0) {
6447: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6448: for (cpt=1; cpt<=cptcoveff;cpt++){
1.237 brouard 6449: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6450: printf(" V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);fflush(stdout);
6451: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6452: /* printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout); */
1.222 brouard 6453: }
1.237 brouard 6454: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6455: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6456: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);fflush(stdout);
6457: }
6458:
1.230 brouard 6459: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6460: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6461: if(invalidvarcomb[k1]){
6462: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6463: printf("\nCombination (%d) ignored because no cases \n",k1);
6464: continue;
6465: }
6466: }
6467: /* aij, bij */
1.259 brouard 6468: fprintf(fichtm,"<br>- Logit model (yours is: logit(pij)=log(pij/pii)= aij+ bij age+%s) as a function of age: <a href=\"%s_%d-1-%d.svg\">%s_%d-1-%d.svg</a><br> \
1.241 brouard 6469: <img src=\"%s_%d-1-%d.svg\">",model,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6470: /* Pij */
1.241 brouard 6471: fprintf(fichtm,"<br>\n- P<sub>ij</sub> or conditional probabilities to be observed in state j being in state i, %d (stepm) months before: <a href=\"%s_%d-2-%d.svg\">%s_%d-2-%d.svg</a><br> \
6472: <img src=\"%s_%d-2-%d.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6473: /* Quasi-incidences */
6474: fprintf(fichtm,"<br>\n- I<sub>ij</sub> or Conditional probabilities to be observed in state j being in state i %d (stepm) months\
1.220 brouard 6475: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6476: incidence (rates) are the limit when h tends to zero of the ratio of the probability <sub>h</sub>P<sub>ij</sub> \
1.241 brouard 6477: divided by h: <sub>h</sub>P<sub>ij</sub>/h : <a href=\"%s_%d-3-%d.svg\">%s_%d-3-%d.svg</a><br> \
6478: <img src=\"%s_%d-3-%d.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6479: /* Survival functions (period) in state j */
6480: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6481: fprintf(fichtm,"<br>\n- Survival functions in state %d. Or probability to survive in state %d being in state (1 to %d) at different ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
6482: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.222 brouard 6483: }
6484: /* State specific survival functions (period) */
6485: for(cpt=1; cpt<=nlstate;cpt++){
6486: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6487: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.241 brouard 6488: <a href=\"%s_%d-%d-%d.svg\">%s_%d%d-%d.svg</a><br> <img src=\"%s_%d-%d-%d.svg\">", cpt, nlstate, cpt, subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.222 brouard 6489: }
6490: /* Period (stable) prevalence in each health state */
6491: for(cpt=1; cpt<=nlstate;cpt++){
1.255 brouard 6492: fprintf(fichtm,"<br>\n- Convergence to period (stable) prevalence in state %d. Or probability to be in state %d some years earlier, knowing that we will be in state (1 to %d) at different ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
1.241 brouard 6493: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.222 brouard 6494: }
6495: if(backcast==1){
6496: /* Period (stable) back prevalence in each health state */
6497: for(cpt=1; cpt<=nlstate;cpt++){
1.255 brouard 6498: fprintf(fichtm,"<br>\n- Convergence to mixed (stable) back prevalence in state %d. Or probability to be in state %d at a younger age, knowing that we will be in state (1 to %d) at different older ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
1.241 brouard 6499: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"PB_"),cpt,k1,nres,subdirf2(optionfilefiname,"PB_"),cpt,k1,nres,subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.222 brouard 6500: }
1.217 brouard 6501: }
1.222 brouard 6502: if(prevfcast==1){
6503: /* Projection of prevalence up to period (stable) prevalence in each health state */
6504: for(cpt=1; cpt<=nlstate;cpt++){
1.258 brouard 6505: fprintf(fichtm,"<br>\n- Projection of cross-sectional prevalence (estimated with cases observed from %.1f to %.1f and mobil_average=%d) up to period (stable) prevalence in state %d. Or probability to be in state %d being in state (1 to %d) at different ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
6506: <img src=\"%s_%d-%d-%d.svg\">", dateprev1, dateprev2, mobilavproj, cpt, cpt, nlstate, subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.222 brouard 6507: }
6508: }
1.220 brouard 6509:
1.222 brouard 6510: for(cpt=1; cpt<=nlstate;cpt++) {
1.241 brouard 6511: fprintf(fichtm,"\n<br>- Life expectancy by health state (%d) at initial age and its decomposition into health expectancies in each alive state (1 to %d) (or area under each survival functions): <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a> <br> \
6512: <img src=\"%s_%d-%d-%d.svg\">",cpt,nlstate,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.222 brouard 6513: }
6514: /* } /\* end i1 *\/ */
6515: }/* End k1 */
6516: fprintf(fichtm,"</ul>");
1.126 brouard 6517:
1.222 brouard 6518: fprintf(fichtm,"\
1.126 brouard 6519: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6520: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6521: - 95%% confidence intervals and Wald tests of the estimated parameters are in the log file if optimization has been done (mle != 0).<br> \
1.197 brouard 6522: But because parameters are usually highly correlated (a higher incidence of disability \
6523: and a higher incidence of recovery can give very close observed transition) it might \
6524: be very useful to look not only at linear confidence intervals estimated from the \
6525: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6526: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6527: covariance matrix of the one-step probabilities. \
6528: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6529:
1.222 brouard 6530: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6531: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6532: fprintf(fichtm,"\
1.126 brouard 6533: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6534: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6535:
1.222 brouard 6536: fprintf(fichtm,"\
1.126 brouard 6537: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6538: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6539: fprintf(fichtm,"\
1.126 brouard 6540: - Variances and covariances of health expectancies by age and <b>initial health status</b> (cov(e<sup>ij</sup>,e<sup>kl</sup>)(estepm=%2d months): \
6541: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6542: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6543: fprintf(fichtm,"\
1.126 brouard 6544: - (a) Health expectancies by health status at initial age (e<sup>ij</sup>) and standard errors (in parentheses) (b) life expectancies and standard errors (e<sup>i.</sup>=e<sup>i1</sup>+e<sup>i2</sup>+...)(estepm=%2d months): \
6545: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6546: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6547: fprintf(fichtm,"\
1.128 brouard 6548: - Variances and covariances of health expectancies by age. Status (i) based health expectancies (in state j), e<sup>ij</sup> are weighted by the period prevalences in each state i (if popbased=1, an additional computation is done using the cross-sectional prevalences, i.e population based) (estepm=%d months): <a href=\"%s\">%s</a><br>\n",
1.222 brouard 6549: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6550: fprintf(fichtm,"\
1.128 brouard 6551: - Total life expectancy and total health expectancies to be spent in each health state e<sup>.j</sup> with their standard errors (if popbased=1, an additional computation is done using the cross-sectional prevalences, i.e population based) (estepm=%d months): <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6552: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6553: fprintf(fichtm,"\
1.126 brouard 6554: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6555: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6556:
6557: /* if(popforecast==1) fprintf(fichtm,"\n */
6558: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6559: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6560: /* <br>",fileres,fileres,fileres,fileres); */
6561: /* else */
6562: /* fprintf(fichtm,"\n No population forecast: popforecast = %d (instead of 1) or stepm = %d (instead of 1) or model=%s (instead of .)<br><br></li>\n",popforecast, stepm, model); */
1.222 brouard 6563: fflush(fichtm);
6564: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6565:
1.225 brouard 6566: m=pow(2,cptcoveff);
1.222 brouard 6567: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6568:
1.222 brouard 6569: jj1=0;
1.237 brouard 6570:
1.241 brouard 6571: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.222 brouard 6572: for(k1=1; k1<=m;k1++){
1.253 brouard 6573: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6574: continue;
1.222 brouard 6575: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6576: jj1++;
1.126 brouard 6577: if (cptcovn > 0) {
6578: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6579: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.237 brouard 6580: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);
6581: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6582: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6583: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6584: }
6585:
1.126 brouard 6586: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6587:
1.222 brouard 6588: if(invalidvarcomb[k1]){
6589: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6590: continue;
6591: }
1.126 brouard 6592: }
6593: for(cpt=1; cpt<=nlstate;cpt++) {
1.258 brouard 6594: fprintf(fichtm,"\n<br>- Observed (cross-sectional with mov_average=%d) and period (incidence based) \
1.241 brouard 6595: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d-%d.svg\"> %s_%d-%d-%d.svg</a>\n <br>\
1.258 brouard 6596: <img src=\"%s_%d-%d-%d.svg\">",mobilav,cpt,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
1.126 brouard 6597: }
6598: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6599: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6600: true period expectancies (those weighted with period prevalences are also\
6601: drawn in addition to the population based expectancies computed using\
1.241 brouard 6602: observed and cahotic prevalences: <a href=\"%s_%d-%d.svg\">%s_%d-%d.svg</a>\n<br>\
6603: <img src=\"%s_%d-%d.svg\">",subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres);
1.222 brouard 6604: /* } /\* end i1 *\/ */
6605: }/* End k1 */
1.241 brouard 6606: }/* End nres */
1.222 brouard 6607: fprintf(fichtm,"</ul>");
6608: fflush(fichtm);
1.126 brouard 6609: }
6610:
6611: /******************* Gnuplot file **************/
1.223 brouard 6612: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6613:
6614: char dirfileres[132],optfileres[132];
1.223 brouard 6615: char gplotcondition[132];
1.237 brouard 6616: int cpt=0,k1=0,i=0,k=0,j=0,jk=0,k2=0,k3=0,k4=0,ij=0, ijp=0, l=0;
1.211 brouard 6617: int lv=0, vlv=0, kl=0;
1.130 brouard 6618: int ng=0;
1.201 brouard 6619: int vpopbased;
1.223 brouard 6620: int ioffset; /* variable offset for columns */
1.235 brouard 6621: int nres=0; /* Index of resultline */
1.219 brouard 6622:
1.126 brouard 6623: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6624: /* printf("Problem with file %s",optionfilegnuplot); */
6625: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6626: /* } */
6627:
6628: /*#ifdef windows */
6629: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6630: /*#endif */
1.225 brouard 6631: m=pow(2,cptcoveff);
1.126 brouard 6632:
1.202 brouard 6633: /* Contribution to likelihood */
6634: /* Plot the probability implied in the likelihood */
1.223 brouard 6635: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6636: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6637: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6638: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6639: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6640: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6641: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6642: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6643: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6644: fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$13):6 t \"All sample, transitions colored by destination\" with dots lc variable; set out;\n",subdirf(fileresilk));
6645: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6646: fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$13):5 t \"All sample, transitions colored by origin\" with dots lc variable; set out;\n\n",subdirf(fileresilk));
6647: for (i=1; i<= nlstate ; i ++) {
6648: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6649: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6650: fprintf(ficgp," u 2:($5 == %d && $6==%d ? $10 : 1/0):($12/4.):6 t \"p%d%d\" with points pointtype 7 ps variable lc variable \\\n",i,1,i,1);
6651: for (j=2; j<= nlstate+ndeath ; j ++) {
6652: fprintf(ficgp,",\\\n \"\" u 2:($5 == %d && $6==%d ? $10 : 1/0):($12/4.):6 t \"p%d%d\" with points pointtype 7 ps variable lc variable ",i,j,i,j);
6653: }
6654: fprintf(ficgp,";\nset out; unset ylabel;\n");
6655: }
6656: /* unset log; plot "rrtest1_sorted_4/ILK_rrtest1_sorted_4.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with points lc variable */
6657: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6658: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6659: fprintf(ficgp,"\nset out;unset log\n");
6660: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6661:
1.126 brouard 6662: strcpy(dirfileres,optionfilefiname);
6663: strcpy(optfileres,"vpl");
1.223 brouard 6664: /* 1eme*/
1.238 brouard 6665: for (cpt=1; cpt<= nlstate ; cpt ++){ /* For each live state */
6666: for (k1=1; k1<= m ; k1 ++){ /* For each valid combination of covariate */
1.236 brouard 6667: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.238 brouard 6668: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
1.253 brouard 6669: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6670: continue;
6671: /* We are interested in selected combination by the resultline */
1.246 brouard 6672: /* printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt); */
1.238 brouard 6673: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
6674: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6675: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
6676: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6677: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6678: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6679: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6680: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.246 brouard 6681: /* printf(" V%d=%d ",Tvaraff[k],vlv); */
1.238 brouard 6682: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6683: }
6684: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.246 brouard 6685: /* printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); */
1.238 brouard 6686: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6687: }
1.246 brouard 6688: /* printf("\n#\n"); */
1.238 brouard 6689: fprintf(ficgp,"\n#\n");
6690: if(invalidvarcomb[k1]){
1.260 brouard 6691: /*k1=k1-1;*/ /* To be checked */
1.238 brouard 6692: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6693: continue;
6694: }
1.235 brouard 6695:
1.241 brouard 6696: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
6697: fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres);
1.260 brouard 6698: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
6699: /* fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres); */
6700: /* k1-1 error should be nres-1*/
1.238 brouard 6701: for (i=1; i<= nlstate ; i ++) {
6702: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6703: else fprintf(ficgp," %%*lf (%%*lf)");
6704: }
1.260 brouard 6705: fprintf(ficgp,"\" t\"Period (stable) prevalence\" w l lt 0,\"%s\" every :::%d::%d u 1:($2==%d ? $3+1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
1.238 brouard 6706: for (i=1; i<= nlstate ; i ++) {
6707: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6708: else fprintf(ficgp," %%*lf (%%*lf)");
6709: }
1.260 brouard 6710: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2==%d ? $3-1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
1.238 brouard 6711: for (i=1; i<= nlstate ; i ++) {
6712: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6713: else fprintf(ficgp," %%*lf (%%*lf)");
6714: }
6715: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6716: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6717: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
1.242 brouard 6718: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1, nres in 2 to be fixed */
1.238 brouard 6719: if(cptcoveff ==0){
1.245 brouard 6720: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
1.238 brouard 6721: }else{
6722: kl=0;
6723: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6724: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6725: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6726: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6727: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6728: vlv= nbcode[Tvaraff[k]][lv];
1.223 brouard 6729: kl++;
1.238 brouard 6730: /* kl=6+(cpt-1)*(nlstate+1)+1+(i-1); /\* 6+(1-1)*(2+1)+1+(1-1)=7, 6+(2-1)(2+1)+1+(1-1)=10 *\/ */
6731: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6732: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6733: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
6734: if(k==cptcoveff){
1.245 brouard 6735: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' w l lt 3",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
1.242 brouard 6736: 2+cptcoveff*2+(cpt-1), cpt ); /* 4 or 6 ?*/
1.238 brouard 6737: }else{
6738: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6739: kl++;
6740: }
6741: } /* end covariate */
6742: } /* end if no covariate */
6743: } /* end if backcast */
6744: fprintf(ficgp,"\nset out \n");
6745: } /* nres */
1.201 brouard 6746: } /* k1 */
6747: } /* cpt */
1.235 brouard 6748:
6749:
1.126 brouard 6750: /*2 eme*/
1.238 brouard 6751: for (k1=1; k1<= m ; k1 ++){
6752: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6753: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6754: continue;
6755: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
6756: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.225 brouard 6757: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6758: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6759: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6760: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6761: vlv= nbcode[Tvaraff[k]][lv];
6762: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6763: }
1.237 brouard 6764: /* for(k=1; k <= ncovds; k++){ */
1.236 brouard 6765: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 6766: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.236 brouard 6767: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6768: }
1.211 brouard 6769: fprintf(ficgp,"\n#\n");
1.223 brouard 6770: if(invalidvarcomb[k1]){
6771: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6772: continue;
6773: }
1.219 brouard 6774:
1.241 brouard 6775: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1,nres);
1.238 brouard 6776: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6777: if(vpopbased==0)
6778: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6779: else
6780: fprintf(ficgp,"\nreplot ");
6781: for (i=1; i<= nlstate+1 ; i ++) {
6782: k=2*i;
1.261 brouard 6783: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ?$4 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1, vpopbased);
1.238 brouard 6784: for (j=1; j<= nlstate+1 ; j ++) {
6785: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6786: else fprintf(ficgp," %%*lf (%%*lf)");
6787: }
6788: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6789: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
1.261 brouard 6790: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4-$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1,vpopbased);
1.238 brouard 6791: for (j=1; j<= nlstate+1 ; j ++) {
6792: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6793: else fprintf(ficgp," %%*lf (%%*lf)");
6794: }
6795: fprintf(ficgp,"\" t\"\" w l lt 0,");
1.261 brouard 6796: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4+$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1,vpopbased);
1.238 brouard 6797: for (j=1; j<= nlstate+1 ; j ++) {
6798: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6799: else fprintf(ficgp," %%*lf (%%*lf)");
6800: }
6801: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6802: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6803: } /* state */
6804: } /* vpopbased */
1.244 brouard 6805: fprintf(ficgp,"\nset out;set out \"%s_%d-%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1,nres); /* Buggy gnuplot */
1.238 brouard 6806: } /* end nres */
6807: } /* k1 end 2 eme*/
6808:
6809:
6810: /*3eme*/
6811: for (k1=1; k1<= m ; k1 ++){
6812: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6813: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6814: continue;
6815:
6816: for (cpt=1; cpt<= nlstate ; cpt ++) {
1.261 brouard 6817: fprintf(ficgp,"\n\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
1.238 brouard 6818: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6819: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6820: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6821: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6822: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6823: vlv= nbcode[Tvaraff[k]][lv];
6824: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6825: }
6826: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6827: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6828: }
6829: fprintf(ficgp,"\n#\n");
6830: if(invalidvarcomb[k1]){
6831: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6832: continue;
6833: }
6834:
6835: /* k=2+nlstate*(2*cpt-2); */
6836: k=2+(nlstate+1)*(cpt-1);
1.241 brouard 6837: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.238 brouard 6838: fprintf(ficgp,"set ter svg size 640, 480\n\
1.261 brouard 6839: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:%d t \"e%d1\" w l",ageminpar,fage,subdirf2(fileresu,"E_"),nres-1,nres-1,k,cpt);
1.238 brouard 6840: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6841: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6842: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6843: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6844: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6845: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6846:
1.238 brouard 6847: */
6848: for (i=1; i< nlstate ; i ++) {
1.261 brouard 6849: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileresu,"E_"),nres-1,nres-1,k+i,cpt,i+1);
1.238 brouard 6850: /* fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileres,"e"),k1-1,k1-1,k+2*i,cpt,i+1);*/
1.219 brouard 6851:
1.238 brouard 6852: }
1.261 brouard 6853: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),nres-1,nres-1,k+nlstate,cpt);
1.238 brouard 6854: }
6855: } /* end nres */
6856: } /* end kl 3eme */
1.126 brouard 6857:
1.223 brouard 6858: /* 4eme */
1.201 brouard 6859: /* Survival functions (period) from state i in state j by initial state i */
1.238 brouard 6860: for (k1=1; k1<=m; k1++){ /* For each covariate and each value */
6861: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6862: if(m != 1 && TKresult[nres]!= k1)
1.223 brouard 6863: continue;
1.238 brouard 6864: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state cpt*/
6865: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
6866: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6867: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6868: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6869: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6870: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6871: vlv= nbcode[Tvaraff[k]][lv];
6872: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6873: }
6874: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6875: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6876: }
6877: fprintf(ficgp,"\n#\n");
6878: if(invalidvarcomb[k1]){
6879: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6880: continue;
1.223 brouard 6881: }
1.238 brouard 6882:
1.241 brouard 6883: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.238 brouard 6884: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6885: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6886: k=3;
6887: for (i=1; i<= nlstate ; i ++){
6888: if(i==1){
6889: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6890: }else{
6891: fprintf(ficgp,", '' ");
6892: }
6893: l=(nlstate+ndeath)*(i-1)+1;
6894: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6895: for (j=2; j<= nlstate+ndeath ; j ++)
6896: fprintf(ficgp,"+$%d",k+l+j-1);
6897: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
6898: } /* nlstate */
6899: fprintf(ficgp,"\nset out\n");
6900: } /* end cpt state*/
6901: } /* end nres */
6902: } /* end covariate k1 */
6903:
1.220 brouard 6904: /* 5eme */
1.201 brouard 6905: /* Survival functions (period) from state i in state j by final state j */
1.238 brouard 6906: for (k1=1; k1<= m ; k1++){ /* For each covariate combination if any */
6907: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6908: if(m != 1 && TKresult[nres]!= k1)
1.227 brouard 6909: continue;
1.238 brouard 6910: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
6911: fprintf(ficgp,"\n#\n#\n# Survival functions in state j and all livestates from state i by final state j: 'lij' files, cov=%d state=%d",k1, cpt);
6912: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6913: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6914: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6915: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6916: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6917: vlv= nbcode[Tvaraff[k]][lv];
6918: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6919: }
6920: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6921: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6922: }
6923: fprintf(ficgp,"\n#\n");
6924: if(invalidvarcomb[k1]){
6925: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6926: continue;
6927: }
1.227 brouard 6928:
1.241 brouard 6929: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.238 brouard 6930: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6931: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6932: k=3;
6933: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6934: if(j==1)
6935: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6936: else
6937: fprintf(ficgp,", '' ");
6938: l=(nlstate+ndeath)*(cpt-1) +j;
6939: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6940: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6941: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6942: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
6943: } /* nlstate */
6944: fprintf(ficgp,", '' ");
6945: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6946: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6947: l=(nlstate+ndeath)*(cpt-1) +j;
6948: if(j < nlstate)
6949: fprintf(ficgp,"$%d +",k+l);
6950: else
6951: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
6952: }
6953: fprintf(ficgp,"\nset out\n");
6954: } /* end cpt state*/
6955: } /* end covariate */
6956: } /* end nres */
1.227 brouard 6957:
1.220 brouard 6958: /* 6eme */
1.202 brouard 6959: /* CV preval stable (period) for each covariate */
1.237 brouard 6960: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6961: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6962: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6963: continue;
1.255 brouard 6964: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state of arrival */
1.227 brouard 6965:
1.211 brouard 6966: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6967: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6968: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6969: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6970: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6971: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6972: vlv= nbcode[Tvaraff[k]][lv];
6973: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6974: }
1.237 brouard 6975: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6976: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6977: }
1.211 brouard 6978: fprintf(ficgp,"\n#\n");
1.223 brouard 6979: if(invalidvarcomb[k1]){
1.227 brouard 6980: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6981: continue;
1.223 brouard 6982: }
1.227 brouard 6983:
1.241 brouard 6984: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.126 brouard 6985: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6986: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6987: k=3; /* Offset */
1.255 brouard 6988: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 6989: if(i==1)
6990: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6991: else
6992: fprintf(ficgp,", '' ");
1.255 brouard 6993: l=(nlstate+ndeath)*(i-1)+1; /* 1, 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 6994: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6995: for (j=2; j<= nlstate ; j ++)
6996: fprintf(ficgp,"+$%d",k+l+j-1);
6997: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6998: } /* nlstate */
1.201 brouard 6999: fprintf(ficgp,"\nset out\n");
1.153 brouard 7000: } /* end cpt state*/
7001: } /* end covariate */
1.227 brouard 7002:
7003:
1.220 brouard 7004: /* 7eme */
1.218 brouard 7005: if(backcast == 1){
1.217 brouard 7006: /* CV back preval stable (period) for each covariate */
1.237 brouard 7007: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7008: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7009: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7010: continue;
1.255 brouard 7011: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life ending state */
7012: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pijb' files, covariatecombination#=%d state=%d",k1, cpt);
1.227 brouard 7013: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
7014: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
7015: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7016: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 7017: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 7018: vlv= nbcode[Tvaraff[k]][lv];
7019: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
7020: }
1.237 brouard 7021: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7022: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7023: }
1.227 brouard 7024: fprintf(ficgp,"\n#\n");
7025: if(invalidvarcomb[k1]){
7026: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7027: continue;
7028: }
7029:
1.241 brouard 7030: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.227 brouard 7031: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 7032: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7033: k=3; /* Offset */
1.255 brouard 7034: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 7035: if(i==1)
7036: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
7037: else
7038: fprintf(ficgp,", '' ");
7039: /* l=(nlstate+ndeath)*(i-1)+1; */
1.255 brouard 7040: l=(nlstate+ndeath)*(cpt-1)+1; /* fixed for i; cpt=1 1, cpt=2 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 7041: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
7042: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
1.255 brouard 7043: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+i-1); /* To be verified */
1.227 brouard 7044: /* for (j=2; j<= nlstate ; j ++) */
7045: /* fprintf(ficgp,"+$%d",k+l+j-1); */
7046: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
7047: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
7048: } /* nlstate */
7049: fprintf(ficgp,"\nset out\n");
1.218 brouard 7050: } /* end cpt state*/
7051: } /* end covariate */
7052: } /* End if backcast */
7053:
1.223 brouard 7054: /* 8eme */
1.218 brouard 7055: if(prevfcast==1){
7056: /* Projection from cross-sectional to stable (period) for each covariate */
7057:
1.237 brouard 7058: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7059: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7060: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7061: continue;
1.211 brouard 7062: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 7063: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
7064: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
7065: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
7066: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7067: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7068: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7069: vlv= nbcode[Tvaraff[k]][lv];
7070: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
7071: }
1.237 brouard 7072: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7073: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7074: }
1.227 brouard 7075: fprintf(ficgp,"\n#\n");
7076: if(invalidvarcomb[k1]){
7077: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7078: continue;
7079: }
7080:
7081: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
1.241 brouard 7082: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.227 brouard 7083: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
1.238 brouard 7084: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7085: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
7086: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7087: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7088: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7089: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7090: if(i==1){
7091: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
7092: }else{
7093: fprintf(ficgp,",\\\n '' ");
7094: }
7095: if(cptcoveff ==0){ /* No covariate */
7096: ioffset=2; /* Age is in 2 */
7097: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7098: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7099: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7100: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7101: fprintf(ficgp," u %d:(", ioffset);
7102: if(i==nlstate+1)
7103: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
7104: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7105: else
7106: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
7107: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7108: }else{ /* more than 2 covariates */
7109: if(cptcoveff ==1){
7110: ioffset=4; /* Age is in 4 */
7111: }else{
7112: ioffset=6; /* Age is in 6 */
7113: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7114: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7115: }
7116: fprintf(ficgp," u %d:(",ioffset);
7117: kl=0;
7118: strcpy(gplotcondition,"(");
7119: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
7120: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
7121: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7122: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7123: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7124: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
7125: kl++;
7126: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
7127: kl++;
7128: if(k <cptcoveff && cptcoveff>1)
7129: sprintf(gplotcondition+strlen(gplotcondition)," && ");
7130: }
7131: strcpy(gplotcondition+strlen(gplotcondition),")");
7132: /* kl=6+(cpt-1)*(nlstate+1)+1+(i-1); /\* 6+(1-1)*(2+1)+1+(1-1)=7, 6+(2-1)(2+1)+1+(1-1)=10 *\/ */
7133: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
7134: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
7135: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
7136: if(i==nlstate+1){
7137: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
7138: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7139: }else{
7140: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
7141: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7142: }
7143: } /* end if covariate */
7144: } /* nlstate */
7145: fprintf(ficgp,"\nset out\n");
1.223 brouard 7146: } /* end cpt state*/
7147: } /* end covariate */
7148: } /* End if prevfcast */
1.227 brouard 7149:
7150:
1.238 brouard 7151: /* 9eme writing MLE parameters */
7152: fprintf(ficgp,"\n##############\n#9eme MLE estimated parameters\n#############\n");
1.126 brouard 7153: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 7154: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 7155: for(k=1; k <=(nlstate+ndeath); k++){
7156: if (k != i) {
1.227 brouard 7157: fprintf(ficgp,"# current state %d\n",k);
7158: for(j=1; j <=ncovmodel; j++){
7159: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
7160: jk++;
7161: }
7162: fprintf(ficgp,"\n");
1.126 brouard 7163: }
7164: }
1.223 brouard 7165: }
1.187 brouard 7166: fprintf(ficgp,"##############\n#\n");
1.227 brouard 7167:
1.145 brouard 7168: /*goto avoid;*/
1.238 brouard 7169: /* 10eme Graphics of probabilities or incidences using written MLE parameters */
7170: fprintf(ficgp,"\n##############\n#10eme Graphics of probabilities or incidences\n#############\n");
1.187 brouard 7171: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
7172: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
7173: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
7174: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
7175: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7176: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7177: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7178: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7179: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
7180: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7181: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
7182: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
7183: fprintf(ficgp,"#\n");
1.223 brouard 7184: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
1.238 brouard 7185: fprintf(ficgp,"#Number of graphics: first is logit, 2nd is probabilities, third is incidences per year\n");
1.237 brouard 7186: fprintf(ficgp,"#model=%s \n",model);
1.238 brouard 7187: fprintf(ficgp,"# Type of graphic ng=%d\n",ng);
1.237 brouard 7188: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);/* to be checked */
7189: for(jk=1; jk <=m; jk++) /* For each combination of covariate */
7190: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7191: if(m != 1 && TKresult[nres]!= jk)
1.237 brouard 7192: continue;
7193: fprintf(ficgp,"# Combination of dummy jk=%d and ",jk);
7194: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7195: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7196: }
7197: fprintf(ficgp,"\n#\n");
1.241 brouard 7198: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng,nres);
1.223 brouard 7199: fprintf(ficgp,"\nset ter svg size 640, 480 ");
7200: if (ng==1){
7201: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
7202: fprintf(ficgp,"\nunset log y");
7203: }else if (ng==2){
7204: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
7205: fprintf(ficgp,"\nset log y");
7206: }else if (ng==3){
7207: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
7208: fprintf(ficgp,"\nset log y");
7209: }else
7210: fprintf(ficgp,"\nunset title ");
7211: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
7212: i=1;
7213: for(k2=1; k2<=nlstate; k2++) {
7214: k3=i;
7215: for(k=1; k<=(nlstate+ndeath); k++) {
7216: if (k != k2){
7217: switch( ng) {
7218: case 1:
7219: if(nagesqr==0)
7220: fprintf(ficgp," p%d+p%d*x",i,i+1);
7221: else /* nagesqr =1 */
7222: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7223: break;
7224: case 2: /* ng=2 */
7225: if(nagesqr==0)
7226: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
7227: else /* nagesqr =1 */
7228: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7229: break;
7230: case 3:
7231: if(nagesqr==0)
7232: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
7233: else /* nagesqr =1 */
7234: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
7235: break;
7236: }
7237: ij=1;/* To be checked else nbcode[0][0] wrong */
1.237 brouard 7238: ijp=1; /* product no age */
7239: /* for(j=3; j <=ncovmodel-nagesqr; j++) { */
7240: for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */
1.223 brouard 7241: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
1.237 brouard 7242: if(j==Tage[ij]) { /* Product by age */
7243: if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */
1.238 brouard 7244: if(DummyV[j]==0){
1.237 brouard 7245: fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);;
7246: }else{ /* quantitative */
7247: fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */
7248: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7249: }
7250: ij++;
7251: }
7252: }else if(j==Tprod[ijp]) { /* */
7253: /* printf("Tprod[%d]=%d, j=%d\n", ij, Tprod[ijp], j); */
7254: if(ijp <=cptcovprod) { /* Product */
1.238 brouard 7255: if(DummyV[Tvard[ijp][1]]==0){/* Vn is dummy */
7256: if(DummyV[Tvard[ijp][2]]==0){/* Vn and Vm are dummy */
1.237 brouard 7257: /* fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],nbcode[Tvard[ijp][2]][codtabm(jk,j)]); */
7258: fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tinvresult[nres][Tvard[ijp][2]]);
7259: }else{ /* Vn is dummy and Vm is quanti */
7260: /* fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],Tqinvresult[nres][Tvard[ijp][2]]); */
7261: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7262: }
7263: }else{ /* Vn*Vm Vn is quanti */
1.238 brouard 7264: if(DummyV[Tvard[ijp][2]]==0){
1.237 brouard 7265: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][2]],Tqinvresult[nres][Tvard[ijp][1]]);
7266: }else{ /* Both quanti */
7267: fprintf(ficgp,"+p%d*%f*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7268: }
7269: }
1.238 brouard 7270: ijp++;
1.237 brouard 7271: }
7272: } else{ /* simple covariate */
7273: /* fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,nbcode[Tvar[j]][codtabm(jk,j)]); /\* Valgrind bug nbcode *\/ */
7274: if(Dummy[j]==0){
7275: fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]); /* */
7276: }else{ /* quantitative */
7277: fprintf(ficgp,"+p%d*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* */
1.223 brouard 7278: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7279: }
1.237 brouard 7280: } /* end simple */
7281: } /* end j */
1.223 brouard 7282: }else{
7283: i=i-ncovmodel;
7284: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
7285: fprintf(ficgp," (1.");
7286: }
1.227 brouard 7287:
1.223 brouard 7288: if(ng != 1){
7289: fprintf(ficgp,")/(1");
1.227 brouard 7290:
1.223 brouard 7291: for(k1=1; k1 <=nlstate; k1++){
7292: if(nagesqr==0)
7293: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
7294: else /* nagesqr =1 */
7295: fprintf(ficgp,"+exp(p%d+p%d*x+p%d*x*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1,k3+(k1-1)*ncovmodel+1+nagesqr);
1.217 brouard 7296:
1.223 brouard 7297: ij=1;
7298: for(j=3; j <=ncovmodel-nagesqr; j++){
1.237 brouard 7299: if((j-2)==Tage[ij]) { /* Bug valgrind */
7300: if(ij <=cptcovage) { /* Bug valgrind */
1.223 brouard 7301: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
7302: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7303: ij++;
7304: }
7305: }
7306: else
1.225 brouard 7307: fprintf(ficgp,"+p%d*%d",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);/* Valgrind bug nbcode */
1.223 brouard 7308: }
7309: fprintf(ficgp,")");
7310: }
7311: fprintf(ficgp,")");
7312: if(ng ==2)
7313: fprintf(ficgp," t \"p%d%d\" ", k2,k);
7314: else /* ng= 3 */
7315: fprintf(ficgp," t \"i%d%d\" ", k2,k);
7316: }else{ /* end ng <> 1 */
7317: if( k !=k2) /* logit p11 is hard to draw */
7318: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
7319: }
7320: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
7321: fprintf(ficgp,",");
7322: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
7323: fprintf(ficgp,",");
7324: i=i+ncovmodel;
7325: } /* end k */
7326: } /* end k2 */
7327: fprintf(ficgp,"\n set out\n");
7328: } /* end jk */
7329: } /* end ng */
7330: /* avoid: */
7331: fflush(ficgp);
1.126 brouard 7332: } /* end gnuplot */
7333:
7334:
7335: /*************** Moving average **************/
1.219 brouard 7336: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 7337: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 7338:
1.222 brouard 7339: int i, cpt, cptcod;
7340: int modcovmax =1;
7341: int mobilavrange, mob;
7342: int iage=0;
7343:
7344: double sum=0.;
7345: double age;
7346: double *sumnewp, *sumnewm;
7347: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
7348:
7349:
1.225 brouard 7350: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 7351: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
7352:
7353: sumnewp = vector(1,ncovcombmax);
7354: sumnewm = vector(1,ncovcombmax);
7355: agemingood = vector(1,ncovcombmax);
7356: agemaxgood = vector(1,ncovcombmax);
7357:
7358: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7359: sumnewm[cptcod]=0.;
7360: sumnewp[cptcod]=0.;
7361: agemingood[cptcod]=0;
7362: agemaxgood[cptcod]=0;
7363: }
7364: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
7365:
7366: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
7367: if(mobilav==1) mobilavrange=5; /* default */
7368: else mobilavrange=mobilav;
7369: for (age=bage; age<=fage; age++)
7370: for (i=1; i<=nlstate;i++)
7371: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
7372: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
7373: /* We keep the original values on the extreme ages bage, fage and for
7374: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
7375: we use a 5 terms etc. until the borders are no more concerned.
7376: */
7377: for (mob=3;mob <=mobilavrange;mob=mob+2){
7378: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
7379: for (i=1; i<=nlstate;i++){
7380: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7381: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
7382: for (cpt=1;cpt<=(mob-1)/2;cpt++){
7383: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
7384: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
7385: }
7386: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
7387: }
7388: }
7389: }/* end age */
7390: }/* end mob */
7391: }else
7392: return -1;
7393: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7394: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
7395: if(invalidvarcomb[cptcod]){
7396: printf("\nCombination (%d) ignored because no cases \n",cptcod);
7397: continue;
7398: }
1.219 brouard 7399:
1.222 brouard 7400: agemingood[cptcod]=fage-(mob-1)/2;
7401: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
7402: sumnewm[cptcod]=0.;
7403: for (i=1; i<=nlstate;i++){
7404: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7405: }
7406: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7407: agemingood[cptcod]=age;
7408: }else{ /* bad */
7409: for (i=1; i<=nlstate;i++){
7410: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7411: } /* i */
7412: } /* end bad */
7413: }/* age */
7414: sum=0.;
7415: for (i=1; i<=nlstate;i++){
7416: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7417: }
7418: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7419: printf("For this combination of covariate cptcod=%d, we can't get a smoothed prevalence which sums to one at any descending age!\n",cptcod);
7420: /* for (i=1; i<=nlstate;i++){ */
7421: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7422: /* } /\* i *\/ */
7423: } /* end bad */
7424: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
7425: /* From youngest, finding the oldest wrong */
7426: agemaxgood[cptcod]=bage+(mob-1)/2;
7427: for (age=bage+(mob-1)/2; age<=fage; age++){
7428: sumnewm[cptcod]=0.;
7429: for (i=1; i<=nlstate;i++){
7430: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7431: }
7432: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7433: agemaxgood[cptcod]=age;
7434: }else{ /* bad */
7435: for (i=1; i<=nlstate;i++){
7436: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7437: } /* i */
7438: } /* end bad */
7439: }/* age */
7440: sum=0.;
7441: for (i=1; i<=nlstate;i++){
7442: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7443: }
7444: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7445: printf("For this combination of covariate cptcod=%d, we can't get a smoothed prevalence which sums to one at any ascending age!\n",cptcod);
7446: /* for (i=1; i<=nlstate;i++){ */
7447: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7448: /* } /\* i *\/ */
7449: } /* end bad */
7450:
7451: for (age=bage; age<=fage; age++){
1.235 brouard 7452: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 7453: sumnewp[cptcod]=0.;
7454: sumnewm[cptcod]=0.;
7455: for (i=1; i<=nlstate;i++){
7456: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
7457: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7458: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
7459: }
7460: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
7461: }
7462: /* printf("\n"); */
7463: /* } */
7464: /* brutal averaging */
7465: for (i=1; i<=nlstate;i++){
7466: for (age=1; age<=bage; age++){
7467: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7468: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7469: }
7470: for (age=fage; age<=AGESUP; age++){
7471: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7472: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7473: }
7474: } /* end i status */
7475: for (i=nlstate+1; i<=nlstate+ndeath;i++){
7476: for (age=1; age<=AGESUP; age++){
7477: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
7478: mobaverage[(int)age][i][cptcod]=0.;
7479: }
7480: }
7481: }/* end cptcod */
7482: free_vector(sumnewm,1, ncovcombmax);
7483: free_vector(sumnewp,1, ncovcombmax);
7484: free_vector(agemaxgood,1, ncovcombmax);
7485: free_vector(agemingood,1, ncovcombmax);
7486: return 0;
7487: }/* End movingaverage */
1.218 brouard 7488:
1.126 brouard 7489:
7490: /************** Forecasting ******************/
1.235 brouard 7491: void prevforecast(char fileres[], double anproj1, double mproj1, double jproj1, double ageminpar, double agemax, double dateprev1, double dateprev2, int mobilav, double bage, double fage, int firstpass, int lastpass, double anproj2, double p[], int cptcoveff){
1.126 brouard 7492: /* proj1, year, month, day of starting projection
7493: agemin, agemax range of age
7494: dateprev1 dateprev2 range of dates during which prevalence is computed
7495: anproj2 year of en of projection (same day and month as proj1).
7496: */
1.235 brouard 7497: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7498: double agec; /* generic age */
7499: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7500: double *popeffectif,*popcount;
7501: double ***p3mat;
1.218 brouard 7502: /* double ***mobaverage; */
1.126 brouard 7503: char fileresf[FILENAMELENGTH];
7504:
7505: agelim=AGESUP;
1.211 brouard 7506: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7507: in each health status at the date of interview (if between dateprev1 and dateprev2).
7508: We still use firstpass and lastpass as another selection.
7509: */
1.214 brouard 7510: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7511: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7512:
1.201 brouard 7513: strcpy(fileresf,"F_");
7514: strcat(fileresf,fileresu);
1.126 brouard 7515: if((ficresf=fopen(fileresf,"w"))==NULL) {
7516: printf("Problem with forecast resultfile: %s\n", fileresf);
7517: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7518: }
1.235 brouard 7519: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7520: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7521:
1.225 brouard 7522: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7523:
7524:
7525: stepsize=(int) (stepm+YEARM-1)/YEARM;
7526: if (stepm<=12) stepsize=1;
7527: if(estepm < stepm){
7528: printf ("Problem %d lower than %d\n",estepm, stepm);
7529: }
7530: else hstepm=estepm;
7531:
7532: hstepm=hstepm/stepm;
7533: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7534: fractional in yp1 */
7535: anprojmean=yp;
7536: yp2=modf((yp1*12),&yp);
7537: mprojmean=yp;
7538: yp1=modf((yp2*30.5),&yp);
7539: jprojmean=yp;
7540: if(jprojmean==0) jprojmean=1;
7541: if(mprojmean==0) jprojmean=1;
7542:
1.227 brouard 7543: i1=pow(2,cptcoveff);
1.126 brouard 7544: if (cptcovn < 1){i1=1;}
7545:
7546: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7547:
7548: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7549:
1.126 brouard 7550: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7551: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7552: for(k=1; k<=i1;k++){
1.253 brouard 7553: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 7554: continue;
1.227 brouard 7555: if(invalidvarcomb[k]){
7556: printf("\nCombination (%d) projection ignored because no cases \n",k);
7557: continue;
7558: }
7559: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7560: for(j=1;j<=cptcoveff;j++) {
7561: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7562: }
1.235 brouard 7563: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 7564: fprintf(ficresf," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.235 brouard 7565: }
1.227 brouard 7566: fprintf(ficresf," yearproj age");
7567: for(j=1; j<=nlstate+ndeath;j++){
7568: for(i=1; i<=nlstate;i++)
7569: fprintf(ficresf," p%d%d",i,j);
7570: fprintf(ficresf," wp.%d",j);
7571: }
7572: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7573: fprintf(ficresf,"\n");
7574: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7575: for (agec=fage; agec>=(ageminpar-1); agec--){
7576: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7577: nhstepm = nhstepm/hstepm;
7578: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7579: oldm=oldms;savm=savms;
1.235 brouard 7580: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7581:
7582: for (h=0; h<=nhstepm; h++){
7583: if (h*hstepm/YEARM*stepm ==yearp) {
7584: fprintf(ficresf,"\n");
7585: for(j=1;j<=cptcoveff;j++)
7586: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7587: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7588: }
7589: for(j=1; j<=nlstate+ndeath;j++) {
7590: ppij=0.;
7591: for(i=1; i<=nlstate;i++) {
7592: if (mobilav==1)
7593: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7594: else {
7595: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7596: }
7597: if (h*hstepm/YEARM*stepm== yearp) {
7598: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7599: }
7600: } /* end i */
7601: if (h*hstepm/YEARM*stepm==yearp) {
7602: fprintf(ficresf," %.3f", ppij);
7603: }
7604: }/* end j */
7605: } /* end h */
7606: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7607: } /* end agec */
7608: } /* end yearp */
7609: } /* end k */
1.219 brouard 7610:
1.126 brouard 7611: fclose(ficresf);
1.215 brouard 7612: printf("End of Computing forecasting \n");
7613: fprintf(ficlog,"End of Computing forecasting\n");
7614:
1.126 brouard 7615: }
7616:
1.218 brouard 7617: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7618: /* void prevbackforecast(char fileres[], double anback1, double mback1, double jback1, double ageminpar, double agemax, double dateprev1, double dateprev2, int mobilav, double bage, double fage, int firstpass, int lastpass, double anback2, double p[], int cptcoveff){ */
1.218 brouard 7619: /* /\* back1, year, month, day of starting backection */
7620: /* agemin, agemax range of age */
7621: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7622: /* anback2 year of en of backection (same day and month as back1). */
7623: /* *\/ */
7624: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7625: /* double agec; /\* generic age *\/ */
7626: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7627: /* double *popeffectif,*popcount; */
7628: /* double ***p3mat; */
7629: /* /\* double ***mobaverage; *\/ */
7630: /* char fileresfb[FILENAMELENGTH]; */
7631:
7632: /* agelim=AGESUP; */
7633: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7634: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7635: /* We still use firstpass and lastpass as another selection. */
7636: /* *\/ */
7637: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7638: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7639: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7640:
7641: /* strcpy(fileresfb,"FB_"); */
7642: /* strcat(fileresfb,fileresu); */
7643: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7644: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7645: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7646: /* } */
7647: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7648: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7649:
1.225 brouard 7650: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7651:
7652: /* /\* if (mobilav!=0) { *\/ */
7653: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7654: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7655: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7656: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7657: /* /\* } *\/ */
7658: /* /\* } *\/ */
7659:
7660: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7661: /* if (stepm<=12) stepsize=1; */
7662: /* if(estepm < stepm){ */
7663: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7664: /* } */
7665: /* else hstepm=estepm; */
7666:
7667: /* hstepm=hstepm/stepm; */
7668: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7669: /* fractional in yp1 *\/ */
7670: /* anprojmean=yp; */
7671: /* yp2=modf((yp1*12),&yp); */
7672: /* mprojmean=yp; */
7673: /* yp1=modf((yp2*30.5),&yp); */
7674: /* jprojmean=yp; */
7675: /* if(jprojmean==0) jprojmean=1; */
7676: /* if(mprojmean==0) jprojmean=1; */
7677:
1.225 brouard 7678: /* i1=cptcoveff; */
1.218 brouard 7679: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7680:
1.218 brouard 7681: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7682:
1.218 brouard 7683: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7684:
7685: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7686: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7687: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7688: /* k=k+1; */
7689: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7690: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7691: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7692: /* } */
7693: /* fprintf(ficresfb," yearbproj age"); */
7694: /* for(j=1; j<=nlstate+ndeath;j++){ */
7695: /* for(i=1; i<=nlstate;i++) */
7696: /* fprintf(ficresfb," p%d%d",i,j); */
7697: /* fprintf(ficresfb," p.%d",j); */
7698: /* } */
7699: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7700: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7701: /* fprintf(ficresfb,"\n"); */
7702: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7703: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7704: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7705: /* nhstepm = nhstepm/hstepm; */
7706: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7707: /* oldm=oldms;savm=savms; */
7708: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7709: /* for (h=0; h<=nhstepm; h++){ */
7710: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7711: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7712: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7713: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7714: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7715: /* } */
7716: /* for(j=1; j<=nlstate+ndeath;j++) { */
7717: /* ppij=0.; */
7718: /* for(i=1; i<=nlstate;i++) { */
7719: /* if (mobilav==1) */
7720: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7721: /* else { */
7722: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7723: /* } */
7724: /* if (h*hstepm/YEARM*stepm== yearp) { */
7725: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7726: /* } */
7727: /* } /\* end i *\/ */
7728: /* if (h*hstepm/YEARM*stepm==yearp) { */
7729: /* fprintf(ficresfb," %.3f", ppij); */
7730: /* } */
7731: /* }/\* end j *\/ */
7732: /* } /\* end h *\/ */
7733: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7734: /* } /\* end agec *\/ */
7735: /* } /\* end yearp *\/ */
7736: /* } /\* end cptcod *\/ */
7737: /* } /\* end cptcov *\/ */
7738:
7739: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7740:
7741: /* fclose(ficresfb); */
7742: /* printf("End of Computing Back forecasting \n"); */
7743: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7744:
1.218 brouard 7745: /* } */
1.217 brouard 7746:
1.126 brouard 7747: /************** Forecasting *****not tested NB*************/
1.227 brouard 7748: /* void populforecast(char fileres[], double anpyram,double mpyram,double jpyram,double ageminpar, double agemax,double dateprev1, double dateprev2s, int mobilav, double agedeb, double fage, int popforecast, char popfile[], double anpyram1,double p[], int i2){ */
1.126 brouard 7749:
1.227 brouard 7750: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7751: /* int *popage; */
7752: /* double calagedatem, agelim, kk1, kk2; */
7753: /* double *popeffectif,*popcount; */
7754: /* double ***p3mat,***tabpop,***tabpopprev; */
7755: /* /\* double ***mobaverage; *\/ */
7756: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7757:
1.227 brouard 7758: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7759: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7760: /* agelim=AGESUP; */
7761: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7762:
1.227 brouard 7763: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7764:
7765:
1.227 brouard 7766: /* strcpy(filerespop,"POP_"); */
7767: /* strcat(filerespop,fileresu); */
7768: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7769: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7770: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7771: /* } */
7772: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7773: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7774:
1.227 brouard 7775: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7776:
1.227 brouard 7777: /* /\* if (mobilav!=0) { *\/ */
7778: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7779: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7780: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7781: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7782: /* /\* } *\/ */
7783: /* /\* } *\/ */
1.126 brouard 7784:
1.227 brouard 7785: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7786: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7787:
1.227 brouard 7788: /* agelim=AGESUP; */
1.126 brouard 7789:
1.227 brouard 7790: /* hstepm=1; */
7791: /* hstepm=hstepm/stepm; */
1.218 brouard 7792:
1.227 brouard 7793: /* if (popforecast==1) { */
7794: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7795: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7796: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7797: /* } */
7798: /* popage=ivector(0,AGESUP); */
7799: /* popeffectif=vector(0,AGESUP); */
7800: /* popcount=vector(0,AGESUP); */
1.126 brouard 7801:
1.227 brouard 7802: /* i=1; */
7803: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7804:
1.227 brouard 7805: /* imx=i; */
7806: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7807: /* } */
1.218 brouard 7808:
1.227 brouard 7809: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7810: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7811: /* k=k+1; */
7812: /* fprintf(ficrespop,"\n#******"); */
7813: /* for(j=1;j<=cptcoveff;j++) { */
7814: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7815: /* } */
7816: /* fprintf(ficrespop,"******\n"); */
7817: /* fprintf(ficrespop,"# Age"); */
7818: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7819: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7820:
1.227 brouard 7821: /* for (cpt=0; cpt<=0;cpt++) { */
7822: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7823:
1.227 brouard 7824: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7825: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7826: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7827:
1.227 brouard 7828: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7829: /* oldm=oldms;savm=savms; */
7830: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7831:
1.227 brouard 7832: /* for (h=0; h<=nhstepm; h++){ */
7833: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7834: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7835: /* } */
7836: /* for(j=1; j<=nlstate+ndeath;j++) { */
7837: /* kk1=0.;kk2=0; */
7838: /* for(i=1; i<=nlstate;i++) { */
7839: /* if (mobilav==1) */
7840: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7841: /* else { */
7842: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7843: /* } */
7844: /* } */
7845: /* if (h==(int)(calagedatem+12*cpt)){ */
7846: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7847: /* /\*fprintf(ficrespop," %.3f", kk1); */
7848: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7849: /* } */
7850: /* } */
7851: /* for(i=1; i<=nlstate;i++){ */
7852: /* kk1=0.; */
7853: /* for(j=1; j<=nlstate;j++){ */
7854: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7855: /* } */
7856: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7857: /* } */
1.218 brouard 7858:
1.227 brouard 7859: /* if (h==(int)(calagedatem+12*cpt)) */
7860: /* for(j=1; j<=nlstate;j++) */
7861: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7862: /* } */
7863: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7864: /* } */
7865: /* } */
1.218 brouard 7866:
1.227 brouard 7867: /* /\******\/ */
1.218 brouard 7868:
1.227 brouard 7869: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7870: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7871: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7872: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7873: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7874:
1.227 brouard 7875: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7876: /* oldm=oldms;savm=savms; */
7877: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7878: /* for (h=0; h<=nhstepm; h++){ */
7879: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7880: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7881: /* } */
7882: /* for(j=1; j<=nlstate+ndeath;j++) { */
7883: /* kk1=0.;kk2=0; */
7884: /* for(i=1; i<=nlstate;i++) { */
7885: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7886: /* } */
7887: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7888: /* } */
7889: /* } */
7890: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7891: /* } */
7892: /* } */
7893: /* } */
7894: /* } */
1.218 brouard 7895:
1.227 brouard 7896: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7897:
1.227 brouard 7898: /* if (popforecast==1) { */
7899: /* free_ivector(popage,0,AGESUP); */
7900: /* free_vector(popeffectif,0,AGESUP); */
7901: /* free_vector(popcount,0,AGESUP); */
7902: /* } */
7903: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7904: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7905: /* fclose(ficrespop); */
7906: /* } /\* End of popforecast *\/ */
1.218 brouard 7907:
1.126 brouard 7908: int fileappend(FILE *fichier, char *optionfich)
7909: {
7910: if((fichier=fopen(optionfich,"a"))==NULL) {
7911: printf("Problem with file: %s\n", optionfich);
7912: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7913: return (0);
7914: }
7915: fflush(fichier);
7916: return (1);
7917: }
7918:
7919:
7920: /**************** function prwizard **********************/
7921: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7922: {
7923:
7924: /* Wizard to print covariance matrix template */
7925:
1.164 brouard 7926: char ca[32], cb[32];
7927: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7928: int numlinepar;
7929:
7930: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7931: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7932: for(i=1; i <=nlstate; i++){
7933: jj=0;
7934: for(j=1; j <=nlstate+ndeath; j++){
7935: if(j==i) continue;
7936: jj++;
7937: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7938: printf("%1d%1d",i,j);
7939: fprintf(ficparo,"%1d%1d",i,j);
7940: for(k=1; k<=ncovmodel;k++){
7941: /* printf(" %lf",param[i][j][k]); */
7942: /* fprintf(ficparo," %lf",param[i][j][k]); */
7943: printf(" 0.");
7944: fprintf(ficparo," 0.");
7945: }
7946: printf("\n");
7947: fprintf(ficparo,"\n");
7948: }
7949: }
7950: printf("# Scales (for hessian or gradient estimation)\n");
7951: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7952: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
7953: for(i=1; i <=nlstate; i++){
7954: jj=0;
7955: for(j=1; j <=nlstate+ndeath; j++){
7956: if(j==i) continue;
7957: jj++;
7958: fprintf(ficparo,"%1d%1d",i,j);
7959: printf("%1d%1d",i,j);
7960: fflush(stdout);
7961: for(k=1; k<=ncovmodel;k++){
7962: /* printf(" %le",delti3[i][j][k]); */
7963: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7964: printf(" 0.");
7965: fprintf(ficparo," 0.");
7966: }
7967: numlinepar++;
7968: printf("\n");
7969: fprintf(ficparo,"\n");
7970: }
7971: }
7972: printf("# Covariance matrix\n");
7973: /* # 121 Var(a12)\n\ */
7974: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7975: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7976: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7977: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7978: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7979: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7980: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7981: fflush(stdout);
7982: fprintf(ficparo,"# Covariance matrix\n");
7983: /* # 121 Var(a12)\n\ */
7984: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7985: /* # ...\n\ */
7986: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7987:
7988: for(itimes=1;itimes<=2;itimes++){
7989: jj=0;
7990: for(i=1; i <=nlstate; i++){
7991: for(j=1; j <=nlstate+ndeath; j++){
7992: if(j==i) continue;
7993: for(k=1; k<=ncovmodel;k++){
7994: jj++;
7995: ca[0]= k+'a'-1;ca[1]='\0';
7996: if(itimes==1){
7997: printf("#%1d%1d%d",i,j,k);
7998: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7999: }else{
8000: printf("%1d%1d%d",i,j,k);
8001: fprintf(ficparo,"%1d%1d%d",i,j,k);
8002: /* printf(" %.5le",matcov[i][j]); */
8003: }
8004: ll=0;
8005: for(li=1;li <=nlstate; li++){
8006: for(lj=1;lj <=nlstate+ndeath; lj++){
8007: if(lj==li) continue;
8008: for(lk=1;lk<=ncovmodel;lk++){
8009: ll++;
8010: if(ll<=jj){
8011: cb[0]= lk +'a'-1;cb[1]='\0';
8012: if(ll<jj){
8013: if(itimes==1){
8014: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
8015: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
8016: }else{
8017: printf(" 0.");
8018: fprintf(ficparo," 0.");
8019: }
8020: }else{
8021: if(itimes==1){
8022: printf(" Var(%s%1d%1d)",ca,i,j);
8023: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
8024: }else{
8025: printf(" 0.");
8026: fprintf(ficparo," 0.");
8027: }
8028: }
8029: }
8030: } /* end lk */
8031: } /* end lj */
8032: } /* end li */
8033: printf("\n");
8034: fprintf(ficparo,"\n");
8035: numlinepar++;
8036: } /* end k*/
8037: } /*end j */
8038: } /* end i */
8039: } /* end itimes */
8040:
8041: } /* end of prwizard */
8042: /******************* Gompertz Likelihood ******************************/
8043: double gompertz(double x[])
8044: {
8045: double A,B,L=0.0,sump=0.,num=0.;
8046: int i,n=0; /* n is the size of the sample */
8047:
1.220 brouard 8048: for (i=1;i<=imx ; i++) {
1.126 brouard 8049: sump=sump+weight[i];
8050: /* sump=sump+1;*/
8051: num=num+1;
8052: }
8053:
8054:
8055: /* for (i=0; i<=imx; i++)
8056: if (wav[i]>0) printf("i=%d ageex=%lf agecens=%lf agedc=%lf cens=%d %d\n" ,i,ageexmed[i],agecens[i],agedc[i],cens[i],wav[i]);*/
8057:
8058: for (i=1;i<=imx ; i++)
8059: {
8060: if (cens[i] == 1 && wav[i]>1)
8061: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
8062:
8063: if (cens[i] == 0 && wav[i]>1)
8064: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
8065: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
8066:
8067: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8068: if (wav[i] > 1 ) { /* ??? */
8069: L=L+A*weight[i];
8070: /* printf("\ni=%d A=%f L=%lf x[1]=%lf x[2]=%lf ageex=%lf agecens=%lf cens=%d agedc=%lf weight=%lf\n",i,A,L,x[1],x[2],ageexmed[i]*12,agecens[i]*12,cens[i],agedc[i]*12,weight[i]);*/
8071: }
8072: }
8073:
8074: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8075:
8076: return -2*L*num/sump;
8077: }
8078:
1.136 brouard 8079: #ifdef GSL
8080: /******************* Gompertz_f Likelihood ******************************/
8081: double gompertz_f(const gsl_vector *v, void *params)
8082: {
8083: double A,B,LL=0.0,sump=0.,num=0.;
8084: double *x= (double *) v->data;
8085: int i,n=0; /* n is the size of the sample */
8086:
8087: for (i=0;i<=imx-1 ; i++) {
8088: sump=sump+weight[i];
8089: /* sump=sump+1;*/
8090: num=num+1;
8091: }
8092:
8093:
8094: /* for (i=0; i<=imx; i++)
8095: if (wav[i]>0) printf("i=%d ageex=%lf agecens=%lf agedc=%lf cens=%d %d\n" ,i,ageexmed[i],agecens[i],agedc[i],cens[i],wav[i]);*/
8096: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
8097: for (i=1;i<=imx ; i++)
8098: {
8099: if (cens[i] == 1 && wav[i]>1)
8100: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
8101:
8102: if (cens[i] == 0 && wav[i]>1)
8103: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
8104: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
8105:
8106: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8107: if (wav[i] > 1 ) { /* ??? */
8108: LL=LL+A*weight[i];
8109: /* printf("\ni=%d A=%f L=%lf x[1]=%lf x[2]=%lf ageex=%lf agecens=%lf cens=%d agedc=%lf weight=%lf\n",i,A,L,x[1],x[2],ageexmed[i]*12,agecens[i]*12,cens[i],agedc[i]*12,weight[i]);*/
8110: }
8111: }
8112:
8113: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8114: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
8115:
8116: return -2*LL*num/sump;
8117: }
8118: #endif
8119:
1.126 brouard 8120: /******************* Printing html file ***********/
1.201 brouard 8121: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 8122: int lastpass, int stepm, int weightopt, char model[],\
8123: int imx, double p[],double **matcov,double agemortsup){
8124: int i,k;
8125:
8126: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
8127: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
8128: for (i=1;i<=2;i++)
8129: fprintf(fichtm," p[%d] = %lf [%f ; %f]<br>\n",i,p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.199 brouard 8130: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 8131: fprintf(fichtm,"</ul>");
8132:
8133: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
8134:
8135: fprintf(fichtm,"\nAge l<inf>x</inf> q<inf>x</inf> d(x,x+1) L<inf>x</inf> T<inf>x</inf> e<infx</inf><br>");
8136:
8137: for (k=agegomp;k<(agemortsup-2);k++)
8138: fprintf(fichtm,"%d %.0lf %lf %.0lf %.0lf %.0lf %lf<br>\n",k,lsurv[k],p[1]*exp(p[2]*(k-agegomp)),(p[1]*exp(p[2]*(k-agegomp)))*lsurv[k],lpop[k],tpop[k],tpop[k]/lsurv[k]);
8139:
8140:
8141: fflush(fichtm);
8142: }
8143:
8144: /******************* Gnuplot file **************/
1.201 brouard 8145: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 8146:
8147: char dirfileres[132],optfileres[132];
1.164 brouard 8148:
1.126 brouard 8149: int ng;
8150:
8151:
8152: /*#ifdef windows */
8153: fprintf(ficgp,"cd \"%s\" \n",pathc);
8154: /*#endif */
8155:
8156:
8157: strcpy(dirfileres,optionfilefiname);
8158: strcpy(optfileres,"vpl");
1.199 brouard 8159: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 8160: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 8161: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 8162: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 8163: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
8164:
8165: }
8166:
1.136 brouard 8167: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
8168: {
1.126 brouard 8169:
1.136 brouard 8170: /*-------- data file ----------*/
8171: FILE *fic;
8172: char dummy[]=" ";
1.240 brouard 8173: int i=0, j=0, n=0, iv=0, v;
1.223 brouard 8174: int lstra;
1.136 brouard 8175: int linei, month, year,iout;
8176: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 8177: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 8178: char *stratrunc;
1.223 brouard 8179:
1.240 brouard 8180: DummyV=ivector(1,NCOVMAX); /* 1 to 3 */
8181: FixedV=ivector(1,NCOVMAX); /* 1 to 3 */
1.126 brouard 8182:
1.240 brouard 8183: for(v=1; v <=ncovcol;v++){
8184: DummyV[v]=0;
8185: FixedV[v]=0;
8186: }
8187: for(v=ncovcol+1; v <=ncovcol+nqv;v++){
8188: DummyV[v]=1;
8189: FixedV[v]=0;
8190: }
8191: for(v=ncovcol+nqv+1; v <=ncovcol+nqv+ntv;v++){
8192: DummyV[v]=0;
8193: FixedV[v]=1;
8194: }
8195: for(v=ncovcol+nqv+ntv+1; v <=ncovcol+nqv+ntv+nqtv;v++){
8196: DummyV[v]=1;
8197: FixedV[v]=1;
8198: }
8199: for(v=1; v <=ncovcol+nqv+ntv+nqtv;v++){
8200: printf("Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
8201: fprintf(ficlog,"Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
8202: }
1.126 brouard 8203:
1.136 brouard 8204: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 8205: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
8206: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 8207: }
1.126 brouard 8208:
1.136 brouard 8209: i=1;
8210: linei=0;
8211: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
8212: linei=linei+1;
8213: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
8214: if(line[j] == '\t')
8215: line[j] = ' ';
8216: }
8217: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
8218: ;
8219: };
8220: line[j+1]=0; /* Trims blanks at end of line */
8221: if(line[0]=='#'){
8222: fprintf(ficlog,"Comment line\n%s\n",line);
8223: printf("Comment line\n%s\n",line);
8224: continue;
8225: }
8226: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 8227: strcpy(line, linetmp);
1.223 brouard 8228:
8229: /* Loops on waves */
8230: for (j=maxwav;j>=1;j--){
8231: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.238 brouard 8232: cutv(stra, strb, line, ' ');
8233: if(strb[0]=='.') { /* Missing value */
8234: lval=-1;
8235: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
8236: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
8237: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
8238: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. If missing, you should remove this individual or impute a value. Exiting.\n", strb, linei,i,line,iv, nqtv, j);
8239: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. If missing, you should remove this individual or impute a value. Exiting.\n", strb, linei,i,line,iv, nqtv, j);fflush(ficlog);
8240: return 1;
8241: }
8242: }else{
8243: errno=0;
8244: /* what_kind_of_number(strb); */
8245: dval=strtod(strb,&endptr);
8246: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
8247: /* if(strb != endptr && *endptr == '\0') */
8248: /* dval=dlval; */
8249: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8250: if( strb[0]=='\0' || (*endptr != '\0')){
8251: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, nqtv, j,maxwav);
8252: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqtv, j,maxwav);fflush(ficlog);
8253: return 1;
8254: }
8255: cotqvar[j][iv][i]=dval;
8256: cotvar[j][ntv+iv][i]=dval;
8257: }
8258: strcpy(line,stra);
1.223 brouard 8259: }/* end loop ntqv */
1.225 brouard 8260:
1.223 brouard 8261: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.238 brouard 8262: cutv(stra, strb, line, ' ');
8263: if(strb[0]=='.') { /* Missing value */
8264: lval=-1;
8265: }else{
8266: errno=0;
8267: lval=strtol(strb,&endptr,10);
8268: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8269: if( strb[0]=='\0' || (*endptr != '\0')){
8270: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th dummy covariate out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, ntv, j,maxwav);
8271: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d dummy covariate out of %d measured wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, ntv,j,maxwav);fflush(ficlog);
8272: return 1;
8273: }
8274: }
8275: if(lval <-1 || lval >1){
8276: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8277: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8278: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8279: For example, for multinomial values like 1, 2 and 3,\n \
8280: build V1=0 V2=0 for the reference value (1),\n \
8281: V1=1 V2=0 for (2) \n \
1.223 brouard 8282: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8283: output of IMaCh is often meaningless.\n \
1.223 brouard 8284: Exiting.\n",lval,linei, i,line,j);
1.238 brouard 8285: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8286: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8287: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8288: For example, for multinomial values like 1, 2 and 3,\n \
8289: build V1=0 V2=0 for the reference value (1),\n \
8290: V1=1 V2=0 for (2) \n \
1.223 brouard 8291: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8292: output of IMaCh is often meaningless.\n \
1.223 brouard 8293: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.238 brouard 8294: return 1;
8295: }
8296: cotvar[j][iv][i]=(double)(lval);
8297: strcpy(line,stra);
1.223 brouard 8298: }/* end loop ntv */
1.225 brouard 8299:
1.223 brouard 8300: /* Statuses at wave */
1.137 brouard 8301: cutv(stra, strb, line, ' ');
1.223 brouard 8302: if(strb[0]=='.') { /* Missing value */
1.238 brouard 8303: lval=-1;
1.136 brouard 8304: }else{
1.238 brouard 8305: errno=0;
8306: lval=strtol(strb,&endptr,10);
8307: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8308: if( strb[0]=='\0' || (*endptr != '\0')){
8309: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a status of wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,j,maxwav);
8310: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a status of wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,j,maxwav);fflush(ficlog);
8311: return 1;
8312: }
1.136 brouard 8313: }
1.225 brouard 8314:
1.136 brouard 8315: s[j][i]=lval;
1.225 brouard 8316:
1.223 brouard 8317: /* Date of Interview */
1.136 brouard 8318: strcpy(line,stra);
8319: cutv(stra, strb,line,' ');
1.169 brouard 8320: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8321: }
1.169 brouard 8322: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 8323: month=99;
8324: year=9999;
1.136 brouard 8325: }else{
1.225 brouard 8326: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of interview (mm/yyyy or .) at wave %d. Exiting.\n",strb, linei,i, line,j);
8327: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of interview (mm/yyyy or .) at wave %d. Exiting.\n",strb, linei,i, line,j);fflush(ficlog);
8328: return 1;
1.136 brouard 8329: }
8330: anint[j][i]= (double) year;
8331: mint[j][i]= (double)month;
8332: strcpy(line,stra);
1.223 brouard 8333: } /* End loop on waves */
1.225 brouard 8334:
1.223 brouard 8335: /* Date of death */
1.136 brouard 8336: cutv(stra, strb,line,' ');
1.169 brouard 8337: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8338: }
1.169 brouard 8339: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 8340: month=99;
8341: year=9999;
8342: }else{
1.141 brouard 8343: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of death (mm/yyyy or .). Exiting.\n",strb, linei,i,line);
1.225 brouard 8344: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of death (mm/yyyy or .). Exiting.\n",strb, linei,i,line);fflush(ficlog);
8345: return 1;
1.136 brouard 8346: }
8347: andc[i]=(double) year;
8348: moisdc[i]=(double) month;
8349: strcpy(line,stra);
8350:
1.223 brouard 8351: /* Date of birth */
1.136 brouard 8352: cutv(stra, strb,line,' ');
1.169 brouard 8353: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8354: }
1.169 brouard 8355: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 8356: month=99;
8357: year=9999;
8358: }else{
1.141 brouard 8359: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy or .). Exiting.\n",strb, linei,i,line);
8360: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy or .). Exiting.\n",strb, linei,i,line);fflush(ficlog);
1.225 brouard 8361: return 1;
1.136 brouard 8362: }
8363: if (year==9999) {
1.141 brouard 8364: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy) but at least the year of birth should be given. Exiting.\n",strb, linei,i,line);
8365: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy) but at least the year of birth should be given. Exiting.\n",strb, linei,i,line);fflush(ficlog);
1.225 brouard 8366: return 1;
8367:
1.136 brouard 8368: }
8369: annais[i]=(double)(year);
8370: moisnais[i]=(double)(month);
8371: strcpy(line,stra);
1.225 brouard 8372:
1.223 brouard 8373: /* Sample weight */
1.136 brouard 8374: cutv(stra, strb,line,' ');
8375: errno=0;
8376: dval=strtod(strb,&endptr);
8377: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 8378: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
8379: fprintf(ficlog,"Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
1.136 brouard 8380: fflush(ficlog);
8381: return 1;
8382: }
8383: weight[i]=dval;
8384: strcpy(line,stra);
1.225 brouard 8385:
1.223 brouard 8386: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
8387: cutv(stra, strb, line, ' ');
8388: if(strb[0]=='.') { /* Missing value */
1.225 brouard 8389: lval=-1;
1.223 brouard 8390: }else{
1.225 brouard 8391: errno=0;
8392: /* what_kind_of_number(strb); */
8393: dval=strtod(strb,&endptr);
8394: /* if(strb != endptr && *endptr == '\0') */
8395: /* dval=dlval; */
8396: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8397: if( strb[0]=='\0' || (*endptr != '\0')){
8398: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value (out of %d) constant for all waves. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqv, maxwav);
8399: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value (out of %d) constant for all waves. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqv, maxwav);fflush(ficlog);
8400: return 1;
8401: }
8402: coqvar[iv][i]=dval;
1.226 brouard 8403: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 8404: }
8405: strcpy(line,stra);
8406: }/* end loop nqv */
1.136 brouard 8407:
1.223 brouard 8408: /* Covariate values */
1.136 brouard 8409: for (j=ncovcol;j>=1;j--){
8410: cutv(stra, strb,line,' ');
1.223 brouard 8411: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 8412: lval=-1;
1.136 brouard 8413: }else{
1.225 brouard 8414: errno=0;
8415: lval=strtol(strb,&endptr,10);
8416: if( strb[0]=='\0' || (*endptr != '\0')){
8417: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\nShould be a covariate value (=0 for the reference or 1 for alternative). Exiting.\n",lval, linei,i, line);
8418: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\nShould be a covariate value (=0 for the reference or 1 for alternative). Exiting.\n",lval, linei,i, line);fflush(ficlog);
8419: return 1;
8420: }
1.136 brouard 8421: }
8422: if(lval <-1 || lval >1){
1.225 brouard 8423: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8424: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8425: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8426: For example, for multinomial values like 1, 2 and 3,\n \
8427: build V1=0 V2=0 for the reference value (1),\n \
8428: V1=1 V2=0 for (2) \n \
1.136 brouard 8429: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8430: output of IMaCh is often meaningless.\n \
1.136 brouard 8431: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 8432: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8433: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8434: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8435: For example, for multinomial values like 1, 2 and 3,\n \
8436: build V1=0 V2=0 for the reference value (1),\n \
8437: V1=1 V2=0 for (2) \n \
1.136 brouard 8438: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8439: output of IMaCh is often meaningless.\n \
1.136 brouard 8440: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 8441: return 1;
1.136 brouard 8442: }
8443: covar[j][i]=(double)(lval);
8444: strcpy(line,stra);
8445: }
8446: lstra=strlen(stra);
1.225 brouard 8447:
1.136 brouard 8448: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
8449: stratrunc = &(stra[lstra-9]);
8450: num[i]=atol(stratrunc);
8451: }
8452: else
8453: num[i]=atol(stra);
8454: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
8455: printf("%ld %.lf %.lf %.lf %.lf/%.lf %.lf/%.lf %.lf/%.lf %d %.lf/%.lf %d %.lf/%.lf %d %.lf/%.lf %d\n",num[i],(covar[1][i]), (covar[2][i]),weight[i], (moisnais[i]), (annais[i]), (moisdc[i]), (andc[i]), (mint[1][i]), (anint[1][i]), (s[1][i]), (mint[2][i]), (anint[2][i]), (s[2][i]), (mint[3][i]), (anint[3][i]), (s[3][i]), (mint[4][i]), (anint[4][i]), (s[4][i])); ij=ij+1;}*/
8456:
8457: i=i+1;
8458: } /* End loop reading data */
1.225 brouard 8459:
1.136 brouard 8460: *imax=i-1; /* Number of individuals */
8461: fclose(fic);
1.225 brouard 8462:
1.136 brouard 8463: return (0);
1.164 brouard 8464: /* endread: */
1.225 brouard 8465: printf("Exiting readdata: ");
8466: fclose(fic);
8467: return (1);
1.223 brouard 8468: }
1.126 brouard 8469:
1.234 brouard 8470: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 8471: char *p1 = *stri, *p2 = *stri;
1.235 brouard 8472: while (*p2 == ' ')
1.234 brouard 8473: p2++;
8474: /* while ((*p1++ = *p2++) !=0) */
8475: /* ; */
8476: /* do */
8477: /* while (*p2 == ' ') */
8478: /* p2++; */
8479: /* while (*p1++ == *p2++); */
8480: *stri=p2;
1.145 brouard 8481: }
8482:
1.235 brouard 8483: int decoderesult ( char resultline[], int nres)
1.230 brouard 8484: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
8485: {
1.235 brouard 8486: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 8487: char resultsav[MAXLINE];
1.234 brouard 8488: int resultmodel[MAXLINE];
8489: int modelresult[MAXLINE];
1.230 brouard 8490: char stra[80], strb[80], strc[80], strd[80],stre[80];
8491:
1.234 brouard 8492: removefirstspace(&resultline);
1.233 brouard 8493: printf("decoderesult:%s\n",resultline);
1.230 brouard 8494:
8495: if (strstr(resultline,"v") !=0){
8496: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
8497: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
8498: return 1;
8499: }
8500: trimbb(resultsav, resultline);
8501: if (strlen(resultsav) >1){
8502: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
8503: }
1.253 brouard 8504: if(j == 0){ /* Resultline but no = */
8505: TKresult[nres]=0; /* Combination for the nresult and the model */
8506: return (0);
8507: }
8508:
1.234 brouard 8509: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
8510: printf("ERROR: the number of variable in the resultline, %d, differs from the number of variable used in the model line, %d.\n",j, cptcovs);
8511: fprintf(ficlog,"ERROR: the number of variable in the resultline, %d, differs from the number of variable used in the model line, %d.\n",j, cptcovs);
8512: }
8513: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8514: if(nbocc(resultsav,'=') >1){
8515: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8516: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8517: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8518: }else
8519: cutl(strc,strd,resultsav,'=');
1.230 brouard 8520: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8521:
1.230 brouard 8522: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8523: Tvarsel[k]=atoi(strc);
8524: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8525: /* cptcovsel++; */
8526: if (nbocc(stra,'=') >0)
8527: strcpy(resultsav,stra); /* and analyzes it */
8528: }
1.235 brouard 8529: /* Checking for missing or useless values in comparison of current model needs */
1.236 brouard 8530: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8531: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8532: match=0;
1.236 brouard 8533: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.237 brouard 8534: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */
1.236 brouard 8535: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8536: match=1;
8537: break;
8538: }
8539: }
8540: if(match == 0){
8541: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8542: }
8543: }
8544: }
1.235 brouard 8545: /* Checking for missing or useless values in comparison of current model needs */
8546: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8547: match=0;
1.235 brouard 8548: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8549: if(Typevar[k1]==0){ /* Single */
1.237 brouard 8550: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */
1.235 brouard 8551: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8552: ++match;
8553: }
8554: }
8555: }
8556: if(match == 0){
8557: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8558: }else if(match > 1){
8559: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8560: }
8561: }
1.235 brouard 8562:
1.234 brouard 8563: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8564: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8565: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8566: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8567: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8568: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8569: /* 1 0 0 0 */
8570: /* 2 1 0 0 */
8571: /* 3 0 1 0 */
8572: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8573: /* 5 0 0 1 */
8574: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8575: /* 7 0 1 1 */
8576: /* 8 1 1 1 */
1.237 brouard 8577: /* V(Tvresult)=Tresult V4=1 V3=0 V1=1 Tresult[nres=1][2]=0 */
8578: /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */
8579: /* V5*age V5 known which value for nres? */
8580: /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */
1.235 brouard 8581: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8582: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
1.237 brouard 8583: k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */
1.235 brouard 8584: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8585: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
1.237 brouard 8586: Tresult[nres][k4+1]=Tvalsel[k3];/* Tresult[nres][1]=1(V4=1) Tresult[nres][2]=0(V3=0) */
8587: Tvresult[nres][k4+1]=(int)Tvarsel[k3];/* Tvresult[nres][1]=4 Tvresult[nres][3]=1 */
8588: Tinvresult[nres][(int)Tvarsel[k3]]=Tvalsel[k3]; /* Tinvresult[nres][4]=1 */
1.235 brouard 8589: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8590: k4++;;
8591: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8592: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8593: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
1.237 brouard 8594: Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */
8595: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */
8596: Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */
1.235 brouard 8597: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8598: k4q++;;
8599: }
8600: }
1.234 brouard 8601:
1.235 brouard 8602: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8603: return (0);
8604: }
1.235 brouard 8605:
1.230 brouard 8606: int decodemodel( char model[], int lastobs)
8607: /**< This routine decodes the model and returns:
1.224 brouard 8608: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8609: * - nagesqr = 1 if age*age in the model, otherwise 0.
8610: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8611: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8612: * - cptcovage number of covariates with age*products =2
8613: * - cptcovs number of simple covariates
8614: * - Tvar[k] is the id of the kth covariate Tvar[1]@12 {1, 2, 3, 8, 10, 11, 8, 3, 7, 8, 5, 6}, thus Tvar[5=V7*V8]=10
8615: * which is a new column after the 9 (ncovcol) variables.
8616: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8617: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8618: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8619: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8620: */
1.136 brouard 8621: {
1.238 brouard 8622: int i, j, k, ks, v;
1.227 brouard 8623: int j1, k1, k2, k3, k4;
1.136 brouard 8624: char modelsav[80];
1.145 brouard 8625: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8626: char *strpt;
1.136 brouard 8627:
1.145 brouard 8628: /*removespace(model);*/
1.136 brouard 8629: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8630: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8631: if (strstr(model,"AGE") !=0){
1.192 brouard 8632: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8633: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8634: return 1;
8635: }
1.141 brouard 8636: if (strstr(model,"v") !=0){
8637: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8638: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8639: return 1;
8640: }
1.187 brouard 8641: strcpy(modelsav,model);
8642: if ((strpt=strstr(model,"age*age")) !=0){
8643: printf(" strpt=%s, model=%s\n",strpt, model);
8644: if(strpt != model){
1.234 brouard 8645: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8646: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8647: corresponding column of parameters.\n",model);
1.234 brouard 8648: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8649: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8650: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8651: return 1;
1.225 brouard 8652: }
1.187 brouard 8653: nagesqr=1;
8654: if (strstr(model,"+age*age") !=0)
1.234 brouard 8655: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8656: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8657: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8658: else
1.234 brouard 8659: substrchaine(modelsav, model, "age*age");
1.187 brouard 8660: }else
8661: nagesqr=0;
8662: if (strlen(modelsav) >1){
8663: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8664: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8665: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8666: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8667: * cst, age and age*age
8668: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8669: /* including age products which are counted in cptcovage.
8670: * but the covariates which are products must be treated
8671: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8672: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8673: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8674:
8675:
1.187 brouard 8676: /* Design
8677: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8678: * < ncovcol=8 >
8679: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8680: * k= 1 2 3 4 5 6 7 8
8681: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8682: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8683: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8684: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8685: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8686: * Tage[++cptcovage]=k
8687: * if products, new covar are created after ncovcol with k1
8688: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8689: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8690: * Tvard[k1][1]=m Tvard[k1][2]=m; Tvard[1][1]=5 (V5) Tvard[1][2]=6 Tvard[2][1]=7 (V7) Tvard[2][2]=8
8691: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8692: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8693: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8694: * < ncovcol=8 >
8695: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8696: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8697: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8698: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8699: * p Tprod[1]@2={ 6, 5}
8700: *p Tvard[1][1]@4= {7, 8, 5, 6}
8701: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8702: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8703: *How to reorganize?
8704: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8705: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8706: * {2, 1, 4, 8, 5, 6, 3, 7}
8707: * Struct []
8708: */
1.225 brouard 8709:
1.187 brouard 8710: /* This loop fills the array Tvar from the string 'model'.*/
8711: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8712: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8713: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8714: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8715: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8716: /* k=1 Tvar[1]=2 (from V2) */
8717: /* k=5 Tvar[5] */
8718: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8719: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8720: /* } */
1.198 brouard 8721: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8722: /*
8723: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8724: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8725: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8726: }
1.187 brouard 8727: cptcovage=0;
8728: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8729: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8730: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8731: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8732: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8733: /*scanf("%d",i);*/
8734: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8735: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8736: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8737: /* covar is not filled and then is empty */
8738: cptcovprod--;
8739: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8740: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8741: Typevar[k]=1; /* 1 for age product */
8742: cptcovage++; /* Sums the number of covariates which include age as a product */
8743: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8744: /*printf("stre=%s ", stre);*/
8745: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8746: cptcovprod--;
8747: cutl(stre,strb,strc,'V');
8748: Tvar[k]=atoi(stre);
8749: Typevar[k]=1; /* 1 for age product */
8750: cptcovage++;
8751: Tage[cptcovage]=k;
8752: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8753: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8754: cptcovn++;
8755: cptcovprodnoage++;k1++;
8756: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8757: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8758: because this model-covariate is a construction we invent a new column
8759: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8760: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8761: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8762: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8763: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8764: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8765: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8766: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8767: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8768: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8769: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8770: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8771: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8772: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8773: for (i=1; i<=lastobs;i++){
8774: /* Computes the new covariate which is a product of
8775: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8776: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8777: }
8778: } /* End age is not in the model */
8779: } /* End if model includes a product */
8780: else { /* no more sum */
8781: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8782: /* scanf("%d",i);*/
8783: cutl(strd,strc,strb,'V');
8784: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8785: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8786: Tvar[k]=atoi(strd);
8787: Typevar[k]=0; /* 0 for simple covariates */
8788: }
8789: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8790: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8791: scanf("%d",i);*/
1.187 brouard 8792: } /* end of loop + on total covariates */
8793: } /* end if strlen(modelsave == 0) age*age might exist */
8794: } /* end if strlen(model == 0) */
1.136 brouard 8795:
8796: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8797: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8798:
1.136 brouard 8799: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8800: printf("cptcovprod=%d ", cptcovprod);
8801: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8802: scanf("%d ",i);*/
8803:
8804:
1.230 brouard 8805: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8806: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8807: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8808: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8809: k = 1 2 3 4 5 6 7 8 9
8810: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8811: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8812: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8813: Dummy[k] 1 0 0 0 3 1 1 2 3
8814: Tmodelind[combination of covar]=k;
1.225 brouard 8815: */
8816: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8817: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8818: /* Tvar[k] is the value n of Vn with n varying for 1 to nvcol, or p Vp=Vn*Vm for product */
1.226 brouard 8819: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8820: printf("Model=%s\n\
8821: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8822: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8823: Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model);
8824: fprintf(ficlog,"Model=%s\n\
8825: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8826: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8827: Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model);
1.240 brouard 8828: for(k=1;k<=cptcovt; k++){ Fixed[k]=0; Dummy[k]=0;}
1.234 brouard 8829: for(k=1, ncovf=0, nsd=0, nsq=0, ncovv=0, ncova=0, ncoveff=0, nqfveff=0, ntveff=0, nqtveff=0;k<=cptcovt; k++){ /* or cptocvt */
8830: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 8831: Fixed[k]= 0;
8832: Dummy[k]= 0;
1.225 brouard 8833: ncoveff++;
1.232 brouard 8834: ncovf++;
1.234 brouard 8835: nsd++;
8836: modell[k].maintype= FTYPE;
8837: TvarsD[nsd]=Tvar[k];
8838: TvarsDind[nsd]=k;
8839: TvarF[ncovf]=Tvar[k];
8840: TvarFind[ncovf]=k;
8841: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8842: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8843: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
8844: Fixed[k]= 0;
8845: Dummy[k]= 0;
8846: ncoveff++;
8847: ncovf++;
8848: modell[k].maintype= FTYPE;
8849: TvarF[ncovf]=Tvar[k];
8850: TvarFind[ncovf]=k;
1.230 brouard 8851: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8852: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.240 brouard 8853: }else if( Tvar[k] <=ncovcol+nqv && Typevar[k]==0){/* Remind that product Vn*Vm are added in k Only simple fixed quantitative variable */
1.227 brouard 8854: Fixed[k]= 0;
8855: Dummy[k]= 1;
1.230 brouard 8856: nqfveff++;
1.234 brouard 8857: modell[k].maintype= FTYPE;
8858: modell[k].subtype= FQ;
8859: nsq++;
8860: TvarsQ[nsq]=Tvar[k];
8861: TvarsQind[nsq]=k;
1.232 brouard 8862: ncovf++;
1.234 brouard 8863: TvarF[ncovf]=Tvar[k];
8864: TvarFind[ncovf]=k;
1.231 brouard 8865: TvarFQ[nqfveff]=Tvar[k]-ncovcol; /* TvarFQ[1]=V2-1=1st in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1.230 brouard 8866: TvarFQind[nqfveff]=k; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1.242 brouard 8867: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying dummy variables */
1.227 brouard 8868: Fixed[k]= 1;
8869: Dummy[k]= 0;
1.225 brouard 8870: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 8871: modell[k].maintype= VTYPE;
8872: modell[k].subtype= VD;
8873: nsd++;
8874: TvarsD[nsd]=Tvar[k];
8875: TvarsDind[nsd]=k;
8876: ncovv++; /* Only simple time varying variables */
8877: TvarV[ncovv]=Tvar[k];
1.242 brouard 8878: TvarVind[ncovv]=k; /* TvarVind[2]=2 TvarVind[3]=3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Any time varying singele */
1.231 brouard 8879: TvarVD[ntveff]=Tvar[k]; /* TvarVD[1]=V4 TvarVD[2]=V3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying dummy variable */
8880: TvarVDind[ntveff]=k; /* TvarVDind[1]=2 TvarVDind[2]=3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying dummy variable */
1.228 brouard 8881: printf("Quasi Tmodelind[%d]=%d,Tvar[Tmodelind[%d]]=V%d, ncovcol=%d, nqv=%d,Tvar[k]- ncovcol-nqv=%d\n",ntveff,k,ntveff,Tvar[k], ncovcol, nqv,Tvar[k]- ncovcol-nqv);
8882: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8883: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 8884: Fixed[k]= 1;
8885: Dummy[k]= 1;
8886: nqtveff++;
8887: modell[k].maintype= VTYPE;
8888: modell[k].subtype= VQ;
8889: ncovv++; /* Only simple time varying variables */
8890: nsq++;
8891: TvarsQ[nsq]=Tvar[k];
8892: TvarsQind[nsq]=k;
8893: TvarV[ncovv]=Tvar[k];
1.242 brouard 8894: TvarVind[ncovv]=k; /* TvarVind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Any time varying singele */
1.231 brouard 8895: TvarVQ[nqtveff]=Tvar[k]; /* TvarVQ[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
8896: TvarVQind[nqtveff]=k; /* TvarVQind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1.234 brouard 8897: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8898: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8899: printf("Quasi TmodelQind[%d]=%d,Tvar[TmodelQind[%d]]=V%d, ncovcol=%d, nqv=%d, ntv=%d,Tvar[k]- ncovcol-nqv-ntv=%d\n",nqtveff,k,nqtveff,Tvar[k], ncovcol, nqv, ntv, Tvar[k]- ncovcol-nqv-ntv);
1.228 brouard 8900: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8901: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 8902: ncova++;
8903: TvarA[ncova]=Tvar[k];
8904: TvarAind[ncova]=k;
1.231 brouard 8905: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.240 brouard 8906: Fixed[k]= 2;
8907: Dummy[k]= 2;
8908: modell[k].maintype= ATYPE;
8909: modell[k].subtype= APFD;
8910: /* ncoveff++; */
1.227 brouard 8911: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.240 brouard 8912: Fixed[k]= 2;
8913: Dummy[k]= 3;
8914: modell[k].maintype= ATYPE;
8915: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8916: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8917: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.240 brouard 8918: Fixed[k]= 3;
8919: Dummy[k]= 2;
8920: modell[k].maintype= ATYPE;
8921: modell[k].subtype= APVD; /* Product age * varying dummy */
8922: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8923: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8924: Fixed[k]= 3;
8925: Dummy[k]= 3;
8926: modell[k].maintype= ATYPE;
8927: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8928: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8929: }
8930: }else if (Typevar[k] == 2) { /* product without age */
8931: k1=Tposprod[k];
8932: if(Tvard[k1][1] <=ncovcol){
1.240 brouard 8933: if(Tvard[k1][2] <=ncovcol){
8934: Fixed[k]= 1;
8935: Dummy[k]= 0;
8936: modell[k].maintype= FTYPE;
8937: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
8938: ncovf++; /* Fixed variables without age */
8939: TvarF[ncovf]=Tvar[k];
8940: TvarFind[ncovf]=k;
8941: }else if(Tvard[k1][2] <=ncovcol+nqv){
8942: Fixed[k]= 0; /* or 2 ?*/
8943: Dummy[k]= 1;
8944: modell[k].maintype= FTYPE;
8945: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
8946: ncovf++; /* Varying variables without age */
8947: TvarF[ncovf]=Tvar[k];
8948: TvarFind[ncovf]=k;
8949: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8950: Fixed[k]= 1;
8951: Dummy[k]= 0;
8952: modell[k].maintype= VTYPE;
8953: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
8954: ncovv++; /* Varying variables without age */
8955: TvarV[ncovv]=Tvar[k];
8956: TvarVind[ncovv]=k;
8957: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8958: Fixed[k]= 1;
8959: Dummy[k]= 1;
8960: modell[k].maintype= VTYPE;
8961: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
8962: ncovv++; /* Varying variables without age */
8963: TvarV[ncovv]=Tvar[k];
8964: TvarVind[ncovv]=k;
8965: }
1.227 brouard 8966: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.240 brouard 8967: if(Tvard[k1][2] <=ncovcol){
8968: Fixed[k]= 0; /* or 2 ?*/
8969: Dummy[k]= 1;
8970: modell[k].maintype= FTYPE;
8971: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
8972: ncovf++; /* Fixed variables without age */
8973: TvarF[ncovf]=Tvar[k];
8974: TvarFind[ncovf]=k;
8975: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8976: Fixed[k]= 1;
8977: Dummy[k]= 1;
8978: modell[k].maintype= VTYPE;
8979: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
8980: ncovv++; /* Varying variables without age */
8981: TvarV[ncovv]=Tvar[k];
8982: TvarVind[ncovv]=k;
8983: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8984: Fixed[k]= 1;
8985: Dummy[k]= 1;
8986: modell[k].maintype= VTYPE;
8987: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
8988: ncovv++; /* Varying variables without age */
8989: TvarV[ncovv]=Tvar[k];
8990: TvarVind[ncovv]=k;
8991: ncovv++; /* Varying variables without age */
8992: TvarV[ncovv]=Tvar[k];
8993: TvarVind[ncovv]=k;
8994: }
1.227 brouard 8995: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.240 brouard 8996: if(Tvard[k1][2] <=ncovcol){
8997: Fixed[k]= 1;
8998: Dummy[k]= 1;
8999: modell[k].maintype= VTYPE;
9000: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
9001: ncovv++; /* Varying variables without age */
9002: TvarV[ncovv]=Tvar[k];
9003: TvarVind[ncovv]=k;
9004: }else if(Tvard[k1][2] <=ncovcol+nqv){
9005: Fixed[k]= 1;
9006: Dummy[k]= 1;
9007: modell[k].maintype= VTYPE;
9008: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
9009: ncovv++; /* Varying variables without age */
9010: TvarV[ncovv]=Tvar[k];
9011: TvarVind[ncovv]=k;
9012: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9013: Fixed[k]= 1;
9014: Dummy[k]= 0;
9015: modell[k].maintype= VTYPE;
9016: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
9017: ncovv++; /* Varying variables without age */
9018: TvarV[ncovv]=Tvar[k];
9019: TvarVind[ncovv]=k;
9020: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9021: Fixed[k]= 1;
9022: Dummy[k]= 1;
9023: modell[k].maintype= VTYPE;
9024: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
9025: ncovv++; /* Varying variables without age */
9026: TvarV[ncovv]=Tvar[k];
9027: TvarVind[ncovv]=k;
9028: }
1.227 brouard 9029: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 9030: if(Tvard[k1][2] <=ncovcol){
9031: Fixed[k]= 1;
9032: Dummy[k]= 1;
9033: modell[k].maintype= VTYPE;
9034: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
9035: ncovv++; /* Varying variables without age */
9036: TvarV[ncovv]=Tvar[k];
9037: TvarVind[ncovv]=k;
9038: }else if(Tvard[k1][2] <=ncovcol+nqv){
9039: Fixed[k]= 1;
9040: Dummy[k]= 1;
9041: modell[k].maintype= VTYPE;
9042: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
9043: ncovv++; /* Varying variables without age */
9044: TvarV[ncovv]=Tvar[k];
9045: TvarVind[ncovv]=k;
9046: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9047: Fixed[k]= 1;
9048: Dummy[k]= 1;
9049: modell[k].maintype= VTYPE;
9050: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
9051: ncovv++; /* Varying variables without age */
9052: TvarV[ncovv]=Tvar[k];
9053: TvarVind[ncovv]=k;
9054: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9055: Fixed[k]= 1;
9056: Dummy[k]= 1;
9057: modell[k].maintype= VTYPE;
9058: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
9059: ncovv++; /* Varying variables without age */
9060: TvarV[ncovv]=Tvar[k];
9061: TvarVind[ncovv]=k;
9062: }
1.227 brouard 9063: }else{
1.240 brouard 9064: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9065: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9066: } /*end k1*/
1.225 brouard 9067: }else{
1.226 brouard 9068: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
9069: fprintf(ficlog,"Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
1.225 brouard 9070: }
1.227 brouard 9071: printf("Decodemodel, k=%d, Tvar[%d]=V%d,Typevar=%d, Fixed=%d, Dummy=%d\n",k, k,Tvar[k],Typevar[k],Fixed[k],Dummy[k]);
1.231 brouard 9072: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 9073: fprintf(ficlog,"Decodemodel, k=%d, Tvar[%d]=V%d,Typevar=%d, Fixed=%d, Dummy=%d\n",k, k,Tvar[k],Typevar[k],Fixed[k],Dummy[k]);
9074: }
9075: /* Searching for doublons in the model */
9076: for(k1=1; k1<= cptcovt;k1++){
9077: for(k2=1; k2 <k1;k2++){
9078: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 9079: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
9080: if(Tvar[k1]==Tvar[k2]){
9081: printf("Error duplication in the model=%s at positions (+) %d and %d, Tvar[%d]=V%d, Tvar[%d]=V%d, Typevar=%d, Fixed=%d, Dummy=%d\n", model, k1,k2, k1, Tvar[k1], k2, Tvar[k2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]);
9082: fprintf(ficlog,"Error duplication in the model=%s at positions (+) %d and %d, Tvar[%d]=V%d, Tvar[%d]=V%d, Typevar=%d, Fixed=%d, Dummy=%d\n", model, k1,k2, k1, Tvar[k1], k2, Tvar[k2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]); fflush(ficlog);
9083: return(1);
9084: }
9085: }else if (Typevar[k1] ==2){
9086: k3=Tposprod[k1];
9087: k4=Tposprod[k2];
9088: if( ((Tvard[k3][1]== Tvard[k4][1])&&(Tvard[k3][2]== Tvard[k4][2])) || ((Tvard[k3][1]== Tvard[k4][2])&&(Tvard[k3][2]== Tvard[k4][1])) ){
9089: printf("Error duplication in the model=%s at positions (+) %d and %d, V%d*V%d, Typevar=%d, Fixed=%d, Dummy=%d\n",model, k1,k2, Tvard[k3][1], Tvard[k3][2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]);
9090: fprintf(ficlog,"Error duplication in the model=%s at positions (+) %d and %d, V%d*V%d, Typevar=%d, Fixed=%d, Dummy=%d\n",model, k1,k2, Tvard[k3][1], Tvard[k3][2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]); fflush(ficlog);
9091: return(1);
9092: }
9093: }
1.227 brouard 9094: }
9095: }
1.225 brouard 9096: }
9097: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
9098: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 9099: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
9100: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 9101: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 9102: /*endread:*/
1.225 brouard 9103: printf("Exiting decodemodel: ");
9104: return (1);
1.136 brouard 9105: }
9106:
1.169 brouard 9107: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.248 brouard 9108: {/* Check ages at death */
1.136 brouard 9109: int i, m;
1.218 brouard 9110: int firstone=0;
9111:
1.136 brouard 9112: for (i=1; i<=imx; i++) {
9113: for(m=2; (m<= maxwav); m++) {
9114: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
9115: anint[m][i]=9999;
1.216 brouard 9116: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
9117: s[m][i]=-1;
1.136 brouard 9118: }
9119: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.260 brouard 9120: *nberr = *nberr + 1;
1.218 brouard 9121: if(firstone == 0){
9122: firstone=1;
1.260 brouard 9123: printf("Warning (#%d)! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown but status is a death state %d at wave %d. If you don't know the vital status, please enter -2. If he/she is still alive but don't know the state, please code with '-1 or '.'. Here, we do not believe in a death, skipped.\nOther similar cases in log file\n", *nberr,(int)moisdc[i],(int)andc[i],num[i],i,s[m][i],m);
1.218 brouard 9124: }
1.262 ! brouard 9125: fprintf(ficlog,"Warning (#%d)! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown but status is a death state %d at wave %d. If you don't know the vital status, please enter -2. If he/she is still alive but don't know the state, please code with '-1 or '.'. Here, we do not believe in a death, skipped.\n", *nberr,(int)moisdc[i],(int)andc[i],num[i],i,s[m][i],m);
1.260 brouard 9126: s[m][i]=-1; /* Droping the death status */
1.136 brouard 9127: }
9128: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 9129: (*nberr)++;
1.259 brouard 9130: printf("Error (#%d)! Month of death of individual %ld on line %d was unknown (%2d) (year of death is %4d) and status is a death state %d at wave %d. Please impute an arbitrary (or not) month and rerun. Currently this transition to death will be skipped (status is set to -2).\nOther similar cases in log file\n", *nberr, num[i],i,(int)moisdc[i],(int)andc[i],s[m][i],m);
1.262 ! brouard 9131: fprintf(ficlog,"Error (#%d)! Month of death of individual %ld on line %d was unknown (%2d) (year of death is %4d) and status is a death state %d at wave %d. Please impute an arbitrary (or not) month and rerun. Currently this transition to death will be skipped (status is set to -2).\n", *nberr, num[i],i,(int)moisdc[i],(int)andc[i],s[m][i],m);
1.259 brouard 9132: s[m][i]=-2; /* We prefer to skip it (and to skip it in version 0.8a1 too */
1.136 brouard 9133: }
9134: }
9135: }
9136:
9137: for (i=1; i<=imx; i++) {
9138: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
9139: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 9140: if(s[m][i] >0 || s[m][i]==-1 || s[m][i]==-2 || s[m][i]==-4 || s[m][i]==-5){ /* What if s[m][i]=-1 */
1.136 brouard 9141: if (s[m][i] >= nlstate+1) {
1.169 brouard 9142: if(agedc[i]>0){
9143: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 9144: agev[m][i]=agedc[i];
1.214 brouard 9145: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 9146: }else {
1.136 brouard 9147: if ((int)andc[i]!=9999){
9148: nbwarn++;
9149: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
9150: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
9151: agev[m][i]=-1;
9152: }
9153: }
1.169 brouard 9154: } /* agedc > 0 */
1.214 brouard 9155: } /* end if */
1.136 brouard 9156: else if(s[m][i] !=9){ /* Standard case, age in fractional
9157: years but with the precision of a month */
9158: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
9159: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
9160: agev[m][i]=1;
9161: else if(agev[m][i] < *agemin){
9162: *agemin=agev[m][i];
9163: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
9164: }
9165: else if(agev[m][i] >*agemax){
9166: *agemax=agev[m][i];
1.156 brouard 9167: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 9168: }
9169: /*agev[m][i]=anint[m][i]-annais[i];*/
9170: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 9171: } /* en if 9*/
1.136 brouard 9172: else { /* =9 */
1.214 brouard 9173: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 9174: agev[m][i]=1;
9175: s[m][i]=-1;
9176: }
9177: }
1.214 brouard 9178: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 9179: agev[m][i]=1;
1.214 brouard 9180: else{
9181: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9182: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9183: agev[m][i]=0;
9184: }
9185: } /* End for lastpass */
9186: }
1.136 brouard 9187:
9188: for (i=1; i<=imx; i++) {
9189: for(m=firstpass; (m<=lastpass); m++){
9190: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 9191: (*nberr)++;
1.136 brouard 9192: printf("Error: on wave %d of individual %d status %d > (nlstate+ndeath)=(%d+%d)=%d\n",m,i,s[m][i],nlstate, ndeath, nlstate+ndeath);
9193: fprintf(ficlog,"Error: on wave %d of individual %d status %d > (nlstate+ndeath)=(%d+%d)=%d\n",m,i,s[m][i],nlstate, ndeath, nlstate+ndeath);
9194: return 1;
9195: }
9196: }
9197: }
9198:
9199: /*for (i=1; i<=imx; i++){
9200: for (m=firstpass; (m<lastpass); m++){
9201: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
9202: }
9203:
9204: }*/
9205:
9206:
1.139 brouard 9207: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
9208: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 9209:
9210: return (0);
1.164 brouard 9211: /* endread:*/
1.136 brouard 9212: printf("Exiting calandcheckages: ");
9213: return (1);
9214: }
9215:
1.172 brouard 9216: #if defined(_MSC_VER)
9217: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9218: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9219: //#include "stdafx.h"
9220: //#include <stdio.h>
9221: //#include <tchar.h>
9222: //#include <windows.h>
9223: //#include <iostream>
9224: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
9225:
9226: LPFN_ISWOW64PROCESS fnIsWow64Process;
9227:
9228: BOOL IsWow64()
9229: {
9230: BOOL bIsWow64 = FALSE;
9231:
9232: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
9233: // (HANDLE, PBOOL);
9234:
9235: //LPFN_ISWOW64PROCESS fnIsWow64Process;
9236:
9237: HMODULE module = GetModuleHandle(_T("kernel32"));
9238: const char funcName[] = "IsWow64Process";
9239: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
9240: GetProcAddress(module, funcName);
9241:
9242: if (NULL != fnIsWow64Process)
9243: {
9244: if (!fnIsWow64Process(GetCurrentProcess(),
9245: &bIsWow64))
9246: //throw std::exception("Unknown error");
9247: printf("Unknown error\n");
9248: }
9249: return bIsWow64 != FALSE;
9250: }
9251: #endif
1.177 brouard 9252:
1.191 brouard 9253: void syscompilerinfo(int logged)
1.167 brouard 9254: {
9255: /* #include "syscompilerinfo.h"*/
1.185 brouard 9256: /* command line Intel compiler 32bit windows, XP compatible:*/
9257: /* /GS /W3 /Gy
9258: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
9259: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
9260: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 9261: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
9262: */
9263: /* 64 bits */
1.185 brouard 9264: /*
9265: /GS /W3 /Gy
9266: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
9267: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
9268: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
9269: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
9270: /* Optimization are useless and O3 is slower than O2 */
9271: /*
9272: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
9273: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
9274: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
9275: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
9276: */
1.186 brouard 9277: /* Link is */ /* /OUT:"visual studio
1.185 brouard 9278: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
9279: /PDB:"visual studio
9280: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
9281: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
9282: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
9283: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
9284: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
9285: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
9286: uiAccess='false'"
9287: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
9288: /NOLOGO /TLBID:1
9289: */
1.177 brouard 9290: #if defined __INTEL_COMPILER
1.178 brouard 9291: #if defined(__GNUC__)
9292: struct utsname sysInfo; /* For Intel on Linux and OS/X */
9293: #endif
1.177 brouard 9294: #elif defined(__GNUC__)
1.179 brouard 9295: #ifndef __APPLE__
1.174 brouard 9296: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 9297: #endif
1.177 brouard 9298: struct utsname sysInfo;
1.178 brouard 9299: int cross = CROSS;
9300: if (cross){
9301: printf("Cross-");
1.191 brouard 9302: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 9303: }
1.174 brouard 9304: #endif
9305:
1.171 brouard 9306: #include <stdint.h>
1.178 brouard 9307:
1.191 brouard 9308: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 9309: #if defined(__clang__)
1.191 brouard 9310: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 9311: #endif
9312: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 9313: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 9314: #endif
9315: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 9316: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 9317: #endif
9318: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 9319: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 9320: #endif
9321: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 9322: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 9323: #endif
9324: #if defined(_MSC_VER)
1.191 brouard 9325: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 9326: #endif
9327: #if defined(__PGI)
1.191 brouard 9328: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 9329: #endif
9330: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 9331: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 9332: #endif
1.191 brouard 9333: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 9334:
1.167 brouard 9335: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
9336: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
9337: // Windows (x64 and x86)
1.191 brouard 9338: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 9339: #elif __unix__ // all unices, not all compilers
9340: // Unix
1.191 brouard 9341: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 9342: #elif __linux__
9343: // linux
1.191 brouard 9344: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 9345: #elif __APPLE__
1.174 brouard 9346: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 9347: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 9348: #endif
9349:
9350: /* __MINGW32__ */
9351: /* __CYGWIN__ */
9352: /* __MINGW64__ */
9353: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
9354: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
9355: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
9356: /* _WIN64 // Defined for applications for Win64. */
9357: /* _M_X64 // Defined for compilations that target x64 processors. */
9358: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 9359:
1.167 brouard 9360: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 9361: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 9362: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 9363: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 9364: #else
1.191 brouard 9365: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 9366: #endif
9367:
1.169 brouard 9368: #if defined(__GNUC__)
9369: # if defined(__GNUC_PATCHLEVEL__)
9370: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9371: + __GNUC_MINOR__ * 100 \
9372: + __GNUC_PATCHLEVEL__)
9373: # else
9374: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9375: + __GNUC_MINOR__ * 100)
9376: # endif
1.174 brouard 9377: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 9378: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 9379:
9380: if (uname(&sysInfo) != -1) {
9381: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 9382: if(logged) fprintf(ficlog,"Running on: %s %s %s %s %s\n ",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.176 brouard 9383: }
9384: else
9385: perror("uname() error");
1.179 brouard 9386: //#ifndef __INTEL_COMPILER
9387: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 9388: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 9389: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 9390: #endif
1.169 brouard 9391: #endif
1.172 brouard 9392:
9393: // void main()
9394: // {
1.169 brouard 9395: #if defined(_MSC_VER)
1.174 brouard 9396: if (IsWow64()){
1.191 brouard 9397: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
9398: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 9399: }
9400: else{
1.191 brouard 9401: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
9402: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 9403: }
1.172 brouard 9404: // printf("\nPress Enter to continue...");
9405: // getchar();
9406: // }
9407:
1.169 brouard 9408: #endif
9409:
1.167 brouard 9410:
1.219 brouard 9411: }
1.136 brouard 9412:
1.219 brouard 9413: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 9414: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 9415: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 9416: /* double ftolpl = 1.e-10; */
1.180 brouard 9417: double age, agebase, agelim;
1.203 brouard 9418: double tot;
1.180 brouard 9419:
1.202 brouard 9420: strcpy(filerespl,"PL_");
9421: strcat(filerespl,fileresu);
9422: if((ficrespl=fopen(filerespl,"w"))==NULL) {
9423: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9424: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9425: }
1.227 brouard 9426: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
9427: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 9428: pstamp(ficrespl);
1.203 brouard 9429: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 9430: fprintf(ficrespl,"#Age ");
9431: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
9432: fprintf(ficrespl,"\n");
1.180 brouard 9433:
1.219 brouard 9434: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 9435:
1.219 brouard 9436: agebase=ageminpar;
9437: agelim=agemaxpar;
1.180 brouard 9438:
1.227 brouard 9439: /* i1=pow(2,ncoveff); */
1.234 brouard 9440: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 9441: if (cptcovn < 1){i1=1;}
1.180 brouard 9442:
1.238 brouard 9443: for(k=1; k<=i1;k++){ /* For each combination k of dummy covariates in the model */
9444: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 9445: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9446: continue;
1.235 brouard 9447:
1.238 brouard 9448: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9449: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
9450: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
9451: /* k=k+1; */
9452: /* to clean */
9453: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9454: fprintf(ficrespl,"#******");
9455: printf("#******");
9456: fprintf(ficlog,"#******");
9457: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9458: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
9459: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9460: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9461: }
9462: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9463: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9464: fprintf(ficrespl," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9465: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9466: }
9467: fprintf(ficrespl,"******\n");
9468: printf("******\n");
9469: fprintf(ficlog,"******\n");
9470: if(invalidvarcomb[k]){
9471: printf("\nCombination (%d) ignored because no case \n",k);
9472: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
9473: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
9474: continue;
9475: }
1.219 brouard 9476:
1.238 brouard 9477: fprintf(ficrespl,"#Age ");
9478: for(j=1;j<=cptcoveff;j++) {
9479: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9480: }
9481: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
9482: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 9483:
1.238 brouard 9484: for (age=agebase; age<=agelim; age++){
9485: /* for (age=agebase; age<=agebase; age++){ */
9486: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
9487: fprintf(ficrespl,"%.0f ",age );
9488: for(j=1;j<=cptcoveff;j++)
9489: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9490: tot=0.;
9491: for(i=1; i<=nlstate;i++){
9492: tot += prlim[i][i];
9493: fprintf(ficrespl," %.5f", prlim[i][i]);
9494: }
9495: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
9496: } /* Age */
9497: /* was end of cptcod */
9498: } /* cptcov */
9499: } /* nres */
1.219 brouard 9500: return 0;
1.180 brouard 9501: }
9502:
1.218 brouard 9503: int back_prevalence_limit(double *p, double **bprlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp, double dateprev1,double dateprev2, int firstpass, int lastpass, int mobilavproj){
9504: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
9505:
9506: /* Computes the back prevalence limit for any combination of covariate values
9507: * at any age between ageminpar and agemaxpar
9508: */
1.235 brouard 9509: int i, j, k, i1, nres=0 ;
1.217 brouard 9510: /* double ftolpl = 1.e-10; */
9511: double age, agebase, agelim;
9512: double tot;
1.218 brouard 9513: /* double ***mobaverage; */
9514: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 9515:
9516: strcpy(fileresplb,"PLB_");
9517: strcat(fileresplb,fileresu);
9518: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
9519: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9520: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9521: }
9522: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9523: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9524: pstamp(ficresplb);
9525: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9526: fprintf(ficresplb,"#Age ");
9527: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9528: fprintf(ficresplb,"\n");
9529:
1.218 brouard 9530:
9531: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9532:
9533: agebase=ageminpar;
9534: agelim=agemaxpar;
9535:
9536:
1.227 brouard 9537: i1=pow(2,cptcoveff);
1.218 brouard 9538: if (cptcovn < 1){i1=1;}
1.227 brouard 9539:
1.238 brouard 9540: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9541: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9542: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9543: continue;
9544: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9545: fprintf(ficresplb,"#******");
9546: printf("#******");
9547: fprintf(ficlog,"#******");
9548: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9549: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9550: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9551: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9552: }
9553: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9554: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9555: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9556: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9557: }
9558: fprintf(ficresplb,"******\n");
9559: printf("******\n");
9560: fprintf(ficlog,"******\n");
9561: if(invalidvarcomb[k]){
9562: printf("\nCombination (%d) ignored because no cases \n",k);
9563: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9564: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9565: continue;
9566: }
1.218 brouard 9567:
1.238 brouard 9568: fprintf(ficresplb,"#Age ");
9569: for(j=1;j<=cptcoveff;j++) {
9570: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9571: }
9572: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9573: fprintf(ficresplb,"Total Years_to_converge\n");
1.218 brouard 9574:
9575:
1.238 brouard 9576: for (age=agebase; age<=agelim; age++){
9577: /* for (age=agebase; age<=agebase; age++){ */
9578: if(mobilavproj > 0){
9579: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9580: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9581: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k, nres);
1.238 brouard 9582: }else if (mobilavproj == 0){
9583: printf("There is no chance to get back prevalence limit if data aren't non zero and summing to 1, please try a non null mobil_average(=%d) parameter or mobil_average=-1 if you want to try at your own risk.\n",mobilavproj);
9584: fprintf(ficlog,"There is no chance to get back prevalence limit if data aren't non zero and summing to 1, please try a non null mobil_average(=%d) parameter or mobil_average=-1 if you want to try at your own risk.\n",mobilavproj);
9585: exit(1);
9586: }else{
9587: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9588: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k,nres);
1.238 brouard 9589: }
9590: fprintf(ficresplb,"%.0f ",age );
9591: for(j=1;j<=cptcoveff;j++)
9592: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9593: tot=0.;
9594: for(i=1; i<=nlstate;i++){
9595: tot += bprlim[i][i];
9596: fprintf(ficresplb," %.5f", bprlim[i][i]);
9597: }
9598: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9599: } /* Age */
9600: /* was end of cptcod */
1.255 brouard 9601: /*fprintf(ficresplb,"\n");*/ /* Seems to be necessary for gnuplot only if two result lines and no covariate. */
1.238 brouard 9602: } /* end of any combination */
9603: } /* end of nres */
1.218 brouard 9604: /* hBijx(p, bage, fage); */
9605: /* fclose(ficrespijb); */
9606:
9607: return 0;
1.217 brouard 9608: }
1.218 brouard 9609:
1.180 brouard 9610: int hPijx(double *p, int bage, int fage){
9611: /*------------- h Pij x at various ages ------------*/
9612:
9613: int stepsize;
9614: int agelim;
9615: int hstepm;
9616: int nhstepm;
1.235 brouard 9617: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9618:
9619: double agedeb;
9620: double ***p3mat;
9621:
1.201 brouard 9622: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9623: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9624: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9625: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9626: }
9627: printf("Computing pij: result on file '%s' \n", filerespij);
9628: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9629:
9630: stepsize=(int) (stepm+YEARM-1)/YEARM;
9631: /*if (stepm<=24) stepsize=2;*/
9632:
9633: agelim=AGESUP;
9634: hstepm=stepsize*YEARM; /* Every year of age */
9635: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9636:
1.180 brouard 9637: /* hstepm=1; aff par mois*/
9638: pstamp(ficrespij);
9639: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9640: i1= pow(2,cptcoveff);
1.218 brouard 9641: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9642: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9643: /* k=k+1; */
1.235 brouard 9644: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9645: for(k=1; k<=i1;k++){
1.253 brouard 9646: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 9647: continue;
1.183 brouard 9648: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9649: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9650: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9651: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9652: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9653: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9654: }
1.183 brouard 9655: fprintf(ficrespij,"******\n");
9656:
9657: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9658: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9659: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9660:
9661: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9662:
1.183 brouard 9663: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9664: oldm=oldms;savm=savms;
1.235 brouard 9665: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9666: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9667: for(i=1; i<=nlstate;i++)
9668: for(j=1; j<=nlstate+ndeath;j++)
9669: fprintf(ficrespij," %1d-%1d",i,j);
9670: fprintf(ficrespij,"\n");
9671: for (h=0; h<=nhstepm; h++){
9672: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9673: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9674: for(i=1; i<=nlstate;i++)
9675: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9676: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9677: fprintf(ficrespij,"\n");
9678: }
1.183 brouard 9679: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9680: fprintf(ficrespij,"\n");
9681: }
1.180 brouard 9682: /*}*/
9683: }
1.218 brouard 9684: return 0;
1.180 brouard 9685: }
1.218 brouard 9686:
9687: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9688: /*------------- h Bij x at various ages ------------*/
9689:
9690: int stepsize;
1.218 brouard 9691: /* int agelim; */
9692: int ageminl;
1.217 brouard 9693: int hstepm;
9694: int nhstepm;
1.238 brouard 9695: int h, i, i1, j, k, nres;
1.218 brouard 9696:
1.217 brouard 9697: double agedeb;
9698: double ***p3mat;
1.218 brouard 9699:
9700: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9701: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9702: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9703: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9704: }
9705: printf("Computing pij back: result on file '%s' \n", filerespijb);
9706: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9707:
9708: stepsize=(int) (stepm+YEARM-1)/YEARM;
9709: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9710:
1.218 brouard 9711: /* agelim=AGESUP; */
9712: ageminl=30;
9713: hstepm=stepsize*YEARM; /* Every year of age */
9714: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9715:
9716: /* hstepm=1; aff par mois*/
9717: pstamp(ficrespijb);
1.255 brouard 9718: fprintf(ficrespijb,"#****** h Bij x Back probability to be in state i at age x-h being in j at x: B1j+B2j+...=1 ");
1.227 brouard 9719: i1= pow(2,cptcoveff);
1.218 brouard 9720: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9721: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9722: /* k=k+1; */
1.238 brouard 9723: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9724: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9725: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9726: continue;
9727: fprintf(ficrespijb,"\n#****** ");
9728: for(j=1;j<=cptcoveff;j++)
9729: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9730: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9731: fprintf(ficrespijb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9732: }
9733: fprintf(ficrespijb,"******\n");
9734: if(invalidvarcomb[k]){
9735: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9736: continue;
9737: }
9738:
9739: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9740: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9741: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9742: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9743: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9744:
9745: /* nhstepm=nhstepm*YEARM; aff par mois*/
9746:
9747: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9748: /* oldm=oldms;savm=savms; */
9749: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9750: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9751: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
1.255 brouard 9752: fprintf(ficrespijb,"# Cov Agex agex-h hbijx with i,j=");
1.217 brouard 9753: for(i=1; i<=nlstate;i++)
9754: for(j=1; j<=nlstate+ndeath;j++)
1.238 brouard 9755: fprintf(ficrespijb," %1d-%1d",i,j);
1.217 brouard 9756: fprintf(ficrespijb,"\n");
1.238 brouard 9757: for (h=0; h<=nhstepm; h++){
9758: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9759: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9760: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
9761: for(i=1; i<=nlstate;i++)
9762: for(j=1; j<=nlstate+ndeath;j++)
9763: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
9764: fprintf(ficrespijb,"\n");
9765: }
9766: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9767: fprintf(ficrespijb,"\n");
9768: } /* end age deb */
9769: } /* end combination */
9770: } /* end nres */
1.218 brouard 9771: return 0;
9772: } /* hBijx */
1.217 brouard 9773:
1.180 brouard 9774:
1.136 brouard 9775: /***********************************************/
9776: /**************** Main Program *****************/
9777: /***********************************************/
9778:
9779: int main(int argc, char *argv[])
9780: {
9781: #ifdef GSL
9782: const gsl_multimin_fminimizer_type *T;
9783: size_t iteri = 0, it;
9784: int rval = GSL_CONTINUE;
9785: int status = GSL_SUCCESS;
9786: double ssval;
9787: #endif
9788: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9789: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9790: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9791: int jj, ll, li, lj, lk;
1.136 brouard 9792: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9793: int num_filled;
1.136 brouard 9794: int itimes;
9795: int NDIM=2;
9796: int vpopbased=0;
1.235 brouard 9797: int nres=0;
1.258 brouard 9798: int endishere=0;
1.136 brouard 9799:
1.164 brouard 9800: char ca[32], cb[32];
1.136 brouard 9801: /* FILE *fichtm; *//* Html File */
9802: /* FILE *ficgp;*/ /*Gnuplot File */
9803: struct stat info;
1.191 brouard 9804: double agedeb=0.;
1.194 brouard 9805:
9806: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9807: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9808:
1.165 brouard 9809: double fret;
1.191 brouard 9810: double dum=0.; /* Dummy variable */
1.136 brouard 9811: double ***p3mat;
1.218 brouard 9812: /* double ***mobaverage; */
1.164 brouard 9813:
9814: char line[MAXLINE];
1.197 brouard 9815: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9816:
1.234 brouard 9817: char modeltemp[MAXLINE];
1.230 brouard 9818: char resultline[MAXLINE];
9819:
1.136 brouard 9820: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9821: char *tok, *val; /* pathtot */
1.136 brouard 9822: int firstobs=1, lastobs=10;
1.195 brouard 9823: int c, h , cpt, c2;
1.191 brouard 9824: int jl=0;
9825: int i1, j1, jk, stepsize=0;
1.194 brouard 9826: int count=0;
9827:
1.164 brouard 9828: int *tab;
1.136 brouard 9829: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9830: int backcast=0;
1.136 brouard 9831: int mobilav=0,popforecast=0;
1.191 brouard 9832: int hstepm=0, nhstepm=0;
1.136 brouard 9833: int agemortsup;
9834: float sumlpop=0.;
9835: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9836: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9837:
1.191 brouard 9838: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9839: double ftolpl=FTOL;
9840: double **prlim;
1.217 brouard 9841: double **bprlim;
1.136 brouard 9842: double ***param; /* Matrix of parameters */
1.251 brouard 9843: double ***paramstart; /* Matrix of starting parameter values */
9844: double *p, *pstart; /* p=param[1][1] pstart is for starting values guessed by freqsummary */
1.136 brouard 9845: double **matcov; /* Matrix of covariance */
1.203 brouard 9846: double **hess; /* Hessian matrix */
1.136 brouard 9847: double ***delti3; /* Scale */
9848: double *delti; /* Scale */
9849: double ***eij, ***vareij;
9850: double **varpl; /* Variances of prevalence limits by age */
9851: double *epj, vepp;
1.164 brouard 9852:
1.136 brouard 9853: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9854: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9855:
1.136 brouard 9856: double **ximort;
1.145 brouard 9857: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9858: int *dcwave;
9859:
1.164 brouard 9860: char z[1]="c";
1.136 brouard 9861:
9862: /*char *strt;*/
9863: char strtend[80];
1.126 brouard 9864:
1.164 brouard 9865:
1.126 brouard 9866: /* setlocale (LC_ALL, ""); */
9867: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9868: /* textdomain (PACKAGE); */
9869: /* setlocale (LC_CTYPE, ""); */
9870: /* setlocale (LC_MESSAGES, ""); */
9871:
9872: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9873: rstart_time = time(NULL);
9874: /* (void) gettimeofday(&start_time,&tzp);*/
9875: start_time = *localtime(&rstart_time);
1.126 brouard 9876: curr_time=start_time;
1.157 brouard 9877: /*tml = *localtime(&start_time.tm_sec);*/
9878: /* strcpy(strstart,asctime(&tml)); */
9879: strcpy(strstart,asctime(&start_time));
1.126 brouard 9880:
9881: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9882: /* tp.tm_sec = tp.tm_sec +86400; */
9883: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9884: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9885: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9886: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9887: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9888: /* strt=asctime(&tmg); */
9889: /* printf("Time(after) =%s",strstart); */
9890: /* (void) time (&time_value);
9891: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9892: * tm = *localtime(&time_value);
9893: * strstart=asctime(&tm);
9894: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9895: */
9896:
9897: nberr=0; /* Number of errors and warnings */
9898: nbwarn=0;
1.184 brouard 9899: #ifdef WIN32
9900: _getcwd(pathcd, size);
9901: #else
1.126 brouard 9902: getcwd(pathcd, size);
1.184 brouard 9903: #endif
1.191 brouard 9904: syscompilerinfo(0);
1.196 brouard 9905: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9906: if(argc <=1){
9907: printf("\nEnter the parameter file name: ");
1.205 brouard 9908: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9909: printf("ERROR Empty parameter file name\n");
9910: goto end;
9911: }
1.126 brouard 9912: i=strlen(pathr);
9913: if(pathr[i-1]=='\n')
9914: pathr[i-1]='\0';
1.156 brouard 9915: i=strlen(pathr);
1.205 brouard 9916: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9917: pathr[i-1]='\0';
1.205 brouard 9918: }
9919: i=strlen(pathr);
9920: if( i==0 ){
9921: printf("ERROR Empty parameter file name\n");
9922: goto end;
9923: }
9924: for (tok = pathr; tok != NULL; ){
1.126 brouard 9925: printf("Pathr |%s|\n",pathr);
9926: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9927: printf("val= |%s| pathr=%s\n",val,pathr);
9928: strcpy (pathtot, val);
9929: if(pathr[0] == '\0') break; /* Dirty */
9930: }
9931: }
9932: else{
9933: strcpy(pathtot,argv[1]);
9934: }
9935: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9936: /*cygwin_split_path(pathtot,path,optionfile);
9937: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9938: /* cutv(path,optionfile,pathtot,'\\');*/
9939:
9940: /* Split argv[0], imach program to get pathimach */
9941: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9942: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9943: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9944: /* strcpy(pathimach,argv[0]); */
9945: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9946: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9947: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9948: #ifdef WIN32
9949: _chdir(path); /* Can be a relative path */
9950: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9951: #else
1.126 brouard 9952: chdir(path); /* Can be a relative path */
1.184 brouard 9953: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9954: #endif
9955: printf("Current directory %s!\n",pathcd);
1.126 brouard 9956: strcpy(command,"mkdir ");
9957: strcat(command,optionfilefiname);
9958: if((outcmd=system(command)) != 0){
1.169 brouard 9959: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9960: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9961: /* fclose(ficlog); */
9962: /* exit(1); */
9963: }
9964: /* if((imk=mkdir(optionfilefiname))<0){ */
9965: /* perror("mkdir"); */
9966: /* } */
9967:
9968: /*-------- arguments in the command line --------*/
9969:
1.186 brouard 9970: /* Main Log file */
1.126 brouard 9971: strcat(filelog, optionfilefiname);
9972: strcat(filelog,".log"); /* */
9973: if((ficlog=fopen(filelog,"w"))==NULL) {
9974: printf("Problem with logfile %s\n",filelog);
9975: goto end;
9976: }
9977: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9978: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9979: fprintf(ficlog,"\nEnter the parameter file name: \n");
9980: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9981: path=%s \n\
9982: optionfile=%s\n\
9983: optionfilext=%s\n\
1.156 brouard 9984: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9985:
1.197 brouard 9986: syscompilerinfo(1);
1.167 brouard 9987:
1.126 brouard 9988: printf("Local time (at start):%s",strstart);
9989: fprintf(ficlog,"Local time (at start): %s",strstart);
9990: fflush(ficlog);
9991: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9992: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9993:
9994: /* */
9995: strcpy(fileres,"r");
9996: strcat(fileres, optionfilefiname);
1.201 brouard 9997: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9998: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9999: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 10000:
1.186 brouard 10001: /* Main ---------arguments file --------*/
1.126 brouard 10002:
10003: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 10004: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
10005: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 10006: fflush(ficlog);
1.149 brouard 10007: /* goto end; */
10008: exit(70);
1.126 brouard 10009: }
10010:
10011:
10012:
10013: strcpy(filereso,"o");
1.201 brouard 10014: strcat(filereso,fileresu);
1.126 brouard 10015: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
10016: printf("Problem with Output resultfile: %s\n", filereso);
10017: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
10018: fflush(ficlog);
10019: goto end;
10020: }
10021:
10022: /* Reads comments: lines beginning with '#' */
10023: numlinepar=0;
1.197 brouard 10024:
10025: /* First parameter line */
10026: while(fgets(line, MAXLINE, ficpar)) {
10027: /* If line starts with a # it is a comment */
10028: if (line[0] == '#') {
10029: numlinepar++;
10030: fputs(line,stdout);
10031: fputs(line,ficparo);
10032: fputs(line,ficlog);
10033: continue;
10034: }else
10035: break;
10036: }
10037: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
10038: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
10039: if (num_filled != 5) {
10040: printf("Should be 5 parameters\n");
10041: }
1.126 brouard 10042: numlinepar++;
1.197 brouard 10043: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
10044: }
10045: /* Second parameter line */
10046: while(fgets(line, MAXLINE, ficpar)) {
10047: /* If line starts with a # it is a comment */
10048: if (line[0] == '#') {
10049: numlinepar++;
10050: fputs(line,stdout);
10051: fputs(line,ficparo);
10052: fputs(line,ficlog);
10053: continue;
10054: }else
10055: break;
10056: }
1.223 brouard 10057: if((num_filled=sscanf(line,"ftol=%lf stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\n", \
10058: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
10059: if (num_filled != 11) {
10060: printf("Not 11 parameters, for example:ftol=1.e-8 stepm=12 ncovcol=2 nqv=1 ntv=2 nqtv=1 nlstate=2 ndeath=1 maxwav=3 mle=1 weight=1\n");
1.209 brouard 10061: printf("but line=%s\n",line);
1.197 brouard 10062: }
1.223 brouard 10063: printf("ftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\n",ftol, stepm, ncovcol, nqv, ntv, nqtv, nlstate, ndeath, maxwav, mle, weightopt);
1.126 brouard 10064: }
1.203 brouard 10065: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 10066: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 10067: /* Third parameter line */
10068: while(fgets(line, MAXLINE, ficpar)) {
10069: /* If line starts with a # it is a comment */
10070: if (line[0] == '#') {
10071: numlinepar++;
10072: fputs(line,stdout);
10073: fputs(line,ficparo);
10074: fputs(line,ficlog);
10075: continue;
10076: }else
10077: break;
10078: }
1.201 brouard 10079: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
10080: if (num_filled == 0)
10081: model[0]='\0';
10082: else if (num_filled != 1){
1.197 brouard 10083: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10084: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10085: model[0]='\0';
10086: goto end;
10087: }
10088: else{
10089: if (model[0]=='+'){
10090: for(i=1; i<=strlen(model);i++)
10091: modeltemp[i-1]=model[i];
1.201 brouard 10092: strcpy(model,modeltemp);
1.197 brouard 10093: }
10094: }
1.199 brouard 10095: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 10096: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 10097: }
10098: /* fscanf(ficpar,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%lf stepm=%d ncovcol=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d model=1+age+%s\n",title, datafile, &lastobs, &firstpass,&lastpass,&ftol, &stepm, &ncovcol, &nlstate,&ndeath, &maxwav, &mle, &weightopt,model); */
10099: /* numlinepar=numlinepar+3; /\* In general *\/ */
10100: /* printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s\n", title, datafile, lastobs, firstpass,lastpass,ftol, stepm, ncovcol, nlstate,ndeath, maxwav, mle, weightopt,model); */
1.223 brouard 10101: fprintf(ficparo,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol,stepm,ncovcol, nqv, ntv, nqtv, nlstate,ndeath,maxwav, mle, weightopt,model);
10102: fprintf(ficlog,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol,stepm,ncovcol, nqv, ntv, nqtv, nlstate,ndeath,maxwav, mle, weightopt,model);
1.126 brouard 10103: fflush(ficlog);
1.190 brouard 10104: /* if(model[0]=='#'|| model[0]== '\0'){ */
10105: if(model[0]=='#'){
1.187 brouard 10106: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
10107: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
10108: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
10109: if(mle != -1){
10110: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
10111: exit(1);
10112: }
10113: }
1.126 brouard 10114: while((c=getc(ficpar))=='#' && c!= EOF){
10115: ungetc(c,ficpar);
10116: fgets(line, MAXLINE, ficpar);
10117: numlinepar++;
1.195 brouard 10118: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
10119: z[0]=line[1];
10120: }
10121: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 10122: fputs(line, stdout);
10123: //puts(line);
1.126 brouard 10124: fputs(line,ficparo);
10125: fputs(line,ficlog);
10126: }
10127: ungetc(c,ficpar);
10128:
10129:
1.145 brouard 10130: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 10131: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 10132: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 10133: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 10134: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
10135: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
10136: v1+v2*age+v2*v3 makes cptcovn = 3
10137: */
10138: if (strlen(model)>1)
1.187 brouard 10139: ncovmodel=2+nbocc(model,'+')+1; /*Number of variables including intercept and age = cptcovn + intercept + age : v1+v2+v3+v2*v4+v5*age makes 5+2=7,age*age makes 3*/
1.145 brouard 10140: else
1.187 brouard 10141: ncovmodel=2; /* Constant and age */
1.133 brouard 10142: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
10143: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 10144: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
10145: printf("Too complex model for current IMaCh: npar=(nlstate+ndeath-1)*nlstate*ncovmodel=%d >= %d(MAXPARM) or nlstate=%d >= %d(NLSTATEMAX) or ndeath=%d >= %d(NDEATHMAX) or ncovmodel=(k+age+#of+signs)=%d(NCOVMAX) >= %d\n",npar, MAXPARM, nlstate, NLSTATEMAX, ndeath, NDEATHMAX, ncovmodel, NCOVMAX);
10146: fprintf(ficlog,"Too complex model for current IMaCh: %d >=%d(MAXPARM) or %d >=%d(NLSTATEMAX) or %d >=%d(NDEATHMAX) or %d(NCOVMAX) >=%d\n",npar, MAXPARM, nlstate, NLSTATEMAX, ndeath, NDEATHMAX, ncovmodel, NCOVMAX);
10147: fflush(stdout);
10148: fclose (ficlog);
10149: goto end;
10150: }
1.126 brouard 10151: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10152: delti=delti3[1][1];
10153: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
10154: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
1.247 brouard 10155: /* We could also provide initial parameters values giving by simple logistic regression
10156: * only one way, that is without matrix product. We will have nlstate maximizations */
10157: /* for(i=1;i<nlstate;i++){ */
10158: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10159: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10160: /* } */
1.126 brouard 10161: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 10162: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
10163: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10164: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
10165: fclose (ficparo);
10166: fclose (ficlog);
10167: goto end;
10168: exit(0);
1.220 brouard 10169: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 10170: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 10171: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
10172: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10173: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10174: matcov=matrix(1,npar,1,npar);
1.203 brouard 10175: hess=matrix(1,npar,1,npar);
1.220 brouard 10176: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 10177: /* Read guessed parameters */
1.126 brouard 10178: /* Reads comments: lines beginning with '#' */
10179: while((c=getc(ficpar))=='#' && c!= EOF){
10180: ungetc(c,ficpar);
10181: fgets(line, MAXLINE, ficpar);
10182: numlinepar++;
1.141 brouard 10183: fputs(line,stdout);
1.126 brouard 10184: fputs(line,ficparo);
10185: fputs(line,ficlog);
10186: }
10187: ungetc(c,ficpar);
10188:
10189: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.251 brouard 10190: paramstart= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.126 brouard 10191: for(i=1; i <=nlstate; i++){
1.234 brouard 10192: j=0;
1.126 brouard 10193: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 10194: if(jj==i) continue;
10195: j++;
10196: fscanf(ficpar,"%1d%1d",&i1,&j1);
10197: if ((i1 != i) || (j1 != jj)){
10198: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 10199: It might be a problem of design; if ncovcol and the model are correct\n \
10200: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 10201: exit(1);
10202: }
10203: fprintf(ficparo,"%1d%1d",i1,j1);
10204: if(mle==1)
10205: printf("%1d%1d",i,jj);
10206: fprintf(ficlog,"%1d%1d",i,jj);
10207: for(k=1; k<=ncovmodel;k++){
10208: fscanf(ficpar," %lf",¶m[i][j][k]);
10209: if(mle==1){
10210: printf(" %lf",param[i][j][k]);
10211: fprintf(ficlog," %lf",param[i][j][k]);
10212: }
10213: else
10214: fprintf(ficlog," %lf",param[i][j][k]);
10215: fprintf(ficparo," %lf",param[i][j][k]);
10216: }
10217: fscanf(ficpar,"\n");
10218: numlinepar++;
10219: if(mle==1)
10220: printf("\n");
10221: fprintf(ficlog,"\n");
10222: fprintf(ficparo,"\n");
1.126 brouard 10223: }
10224: }
10225: fflush(ficlog);
1.234 brouard 10226:
1.251 brouard 10227: /* Reads parameters values */
1.126 brouard 10228: p=param[1][1];
1.251 brouard 10229: pstart=paramstart[1][1];
1.126 brouard 10230:
10231: /* Reads comments: lines beginning with '#' */
10232: while((c=getc(ficpar))=='#' && c!= EOF){
10233: ungetc(c,ficpar);
10234: fgets(line, MAXLINE, ficpar);
10235: numlinepar++;
1.141 brouard 10236: fputs(line,stdout);
1.126 brouard 10237: fputs(line,ficparo);
10238: fputs(line,ficlog);
10239: }
10240: ungetc(c,ficpar);
10241:
10242: for(i=1; i <=nlstate; i++){
10243: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 10244: fscanf(ficpar,"%1d%1d",&i1,&j1);
10245: if ( (i1-i) * (j1-j) != 0){
10246: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
10247: exit(1);
10248: }
10249: printf("%1d%1d",i,j);
10250: fprintf(ficparo,"%1d%1d",i1,j1);
10251: fprintf(ficlog,"%1d%1d",i1,j1);
10252: for(k=1; k<=ncovmodel;k++){
10253: fscanf(ficpar,"%le",&delti3[i][j][k]);
10254: printf(" %le",delti3[i][j][k]);
10255: fprintf(ficparo," %le",delti3[i][j][k]);
10256: fprintf(ficlog," %le",delti3[i][j][k]);
10257: }
10258: fscanf(ficpar,"\n");
10259: numlinepar++;
10260: printf("\n");
10261: fprintf(ficparo,"\n");
10262: fprintf(ficlog,"\n");
1.126 brouard 10263: }
10264: }
10265: fflush(ficlog);
1.234 brouard 10266:
1.145 brouard 10267: /* Reads covariance matrix */
1.126 brouard 10268: delti=delti3[1][1];
1.220 brouard 10269:
10270:
1.126 brouard 10271: /* free_ma3x(delti3,1,nlstate,1,nlstate+ndeath-1,1,ncovmodel); */ /* Hasn't to to freed here otherwise delti is no more allocated */
1.220 brouard 10272:
1.126 brouard 10273: /* Reads comments: lines beginning with '#' */
10274: while((c=getc(ficpar))=='#' && c!= EOF){
10275: ungetc(c,ficpar);
10276: fgets(line, MAXLINE, ficpar);
10277: numlinepar++;
1.141 brouard 10278: fputs(line,stdout);
1.126 brouard 10279: fputs(line,ficparo);
10280: fputs(line,ficlog);
10281: }
10282: ungetc(c,ficpar);
1.220 brouard 10283:
1.126 brouard 10284: matcov=matrix(1,npar,1,npar);
1.203 brouard 10285: hess=matrix(1,npar,1,npar);
1.131 brouard 10286: for(i=1; i <=npar; i++)
10287: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 10288:
1.194 brouard 10289: /* Scans npar lines */
1.126 brouard 10290: for(i=1; i <=npar; i++){
1.226 brouard 10291: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 10292: if(count != 3){
1.226 brouard 10293: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10294: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10295: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10296: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10297: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10298: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10299: exit(1);
1.220 brouard 10300: }else{
1.226 brouard 10301: if(mle==1)
10302: printf("%1d%1d%d",i1,j1,jk);
10303: }
10304: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
10305: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 10306: for(j=1; j <=i; j++){
1.226 brouard 10307: fscanf(ficpar," %le",&matcov[i][j]);
10308: if(mle==1){
10309: printf(" %.5le",matcov[i][j]);
10310: }
10311: fprintf(ficlog," %.5le",matcov[i][j]);
10312: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 10313: }
10314: fscanf(ficpar,"\n");
10315: numlinepar++;
10316: if(mle==1)
1.220 brouard 10317: printf("\n");
1.126 brouard 10318: fprintf(ficlog,"\n");
10319: fprintf(ficparo,"\n");
10320: }
1.194 brouard 10321: /* End of read covariance matrix npar lines */
1.126 brouard 10322: for(i=1; i <=npar; i++)
10323: for(j=i+1;j<=npar;j++)
1.226 brouard 10324: matcov[i][j]=matcov[j][i];
1.126 brouard 10325:
10326: if(mle==1)
10327: printf("\n");
10328: fprintf(ficlog,"\n");
10329:
10330: fflush(ficlog);
10331:
10332: /*-------- Rewriting parameter file ----------*/
10333: strcpy(rfileres,"r"); /* "Rparameterfile */
10334: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
10335: strcat(rfileres,"."); /* */
10336: strcat(rfileres,optionfilext); /* Other files have txt extension */
10337: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 10338: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
10339: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 10340: }
10341: fprintf(ficres,"#%s\n",version);
10342: } /* End of mle != -3 */
1.218 brouard 10343:
1.186 brouard 10344: /* Main data
10345: */
1.126 brouard 10346: n= lastobs;
10347: num=lvector(1,n);
10348: moisnais=vector(1,n);
10349: annais=vector(1,n);
10350: moisdc=vector(1,n);
10351: andc=vector(1,n);
1.220 brouard 10352: weight=vector(1,n);
1.126 brouard 10353: agedc=vector(1,n);
10354: cod=ivector(1,n);
1.220 brouard 10355: for(i=1;i<=n;i++){
1.234 brouard 10356: num[i]=0;
10357: moisnais[i]=0;
10358: annais[i]=0;
10359: moisdc[i]=0;
10360: andc[i]=0;
10361: agedc[i]=0;
10362: cod[i]=0;
10363: weight[i]=1.0; /* Equal weights, 1 by default */
10364: }
1.126 brouard 10365: mint=matrix(1,maxwav,1,n);
10366: anint=matrix(1,maxwav,1,n);
1.131 brouard 10367: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 10368: tab=ivector(1,NCOVMAX);
1.144 brouard 10369: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 10370: ncodemaxwundef=ivector(1,NCOVMAX); /* Number of code per covariate; if - 1 O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.126 brouard 10371:
1.136 brouard 10372: /* Reads data from file datafile */
10373: if (readdata(datafile, firstobs, lastobs, &imx)==1)
10374: goto end;
10375:
10376: /* Calculation of the number of parameters from char model */
1.234 brouard 10377: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 10378: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
10379: k=3 V4 Tvar[k=3]= 4 (from V4)
10380: k=2 V1 Tvar[k=2]= 1 (from V1)
10381: k=1 Tvar[1]=2 (from V2)
1.234 brouard 10382: */
10383:
10384: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
10385: TvarsDind=ivector(1,NCOVMAX); /* */
10386: TvarsD=ivector(1,NCOVMAX); /* */
10387: TvarsQind=ivector(1,NCOVMAX); /* */
10388: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 10389: TvarF=ivector(1,NCOVMAX); /* */
10390: TvarFind=ivector(1,NCOVMAX); /* */
10391: TvarV=ivector(1,NCOVMAX); /* */
10392: TvarVind=ivector(1,NCOVMAX); /* */
10393: TvarA=ivector(1,NCOVMAX); /* */
10394: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 10395: TvarFD=ivector(1,NCOVMAX); /* */
10396: TvarFDind=ivector(1,NCOVMAX); /* */
10397: TvarFQ=ivector(1,NCOVMAX); /* */
10398: TvarFQind=ivector(1,NCOVMAX); /* */
10399: TvarVD=ivector(1,NCOVMAX); /* */
10400: TvarVDind=ivector(1,NCOVMAX); /* */
10401: TvarVQ=ivector(1,NCOVMAX); /* */
10402: TvarVQind=ivector(1,NCOVMAX); /* */
10403:
1.230 brouard 10404: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 10405: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 10406: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
10407: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
10408: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 10409: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
10410: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
10411: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
10412: */
10413: /* For model-covariate k tells which data-covariate to use but
10414: because this model-covariate is a construction we invent a new column
10415: ncovcol + k1
10416: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
10417: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 10418: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
10419: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 10420: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
10421: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 10422: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 10423: */
1.145 brouard 10424: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
10425: Tvard=imatrix(1,NCOVMAX,1,2); /* n=Tvard[k1][1] and m=Tvard[k1][2] gives the couple n,m of the k1 th product Vn*Vm
1.141 brouard 10426: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
10427: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 10428: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 10429: 4 covariates (3 plus signs)
10430: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
10431: */
1.230 brouard 10432: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 10433: * individual dummy, fixed or varying:
10434: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
10435: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 10436: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
10437: * V1 df, V2 qf, V3 & V4 dv, V5 qv
10438: * Tmodelind[1]@9={9,0,3,2,}*/
10439: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
10440: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 10441: * individual quantitative, fixed or varying:
10442: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
10443: * 3, 1, 0, 0, 0, 0, 0, 0},
10444: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 10445: /* Main decodemodel */
10446:
1.187 brouard 10447:
1.223 brouard 10448: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 10449: goto end;
10450:
1.137 brouard 10451: if((double)(lastobs-imx)/(double)imx > 1.10){
10452: nbwarn++;
10453: printf("Warning: The value of parameter lastobs=%d is big compared to the \n effective number of cases imx=%d, please adjust, \n otherwise you are allocating more memory than necessary.\n",lastobs, imx);
10454: fprintf(ficlog,"Warning: The value of parameter lastobs=%d is big compared to the \n effective number of cases imx=%d, please adjust, \n otherwise you are allocating more memory than necessary.\n",lastobs, imx);
10455: }
1.136 brouard 10456: /* if(mle==1){*/
1.137 brouard 10457: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
10458: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 10459: }
10460:
10461: /*-calculation of age at interview from date of interview and age at death -*/
10462: agev=matrix(1,maxwav,1,imx);
10463:
10464: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
10465: goto end;
10466:
1.126 brouard 10467:
1.136 brouard 10468: agegomp=(int)agemin;
10469: free_vector(moisnais,1,n);
10470: free_vector(annais,1,n);
1.126 brouard 10471: /* free_matrix(mint,1,maxwav,1,n);
10472: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 10473: /* free_vector(moisdc,1,n); */
10474: /* free_vector(andc,1,n); */
1.145 brouard 10475: /* */
10476:
1.126 brouard 10477: wav=ivector(1,imx);
1.214 brouard 10478: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
10479: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
10480: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
10481: dh=imatrix(1,lastpass-firstpass+2,1,imx); /* We are adding a wave if status is unknown at last wave but death occurs after last wave.*/
10482: bh=imatrix(1,lastpass-firstpass+2,1,imx);
10483: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 10484:
10485: /* Concatenates waves */
1.214 brouard 10486: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
10487: Death is a valid wave (if date is known).
10488: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
10489: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
10490: and mw[mi+1][i]. dh depends on stepm.
10491: */
10492:
1.126 brouard 10493: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.248 brouard 10494: /* Concatenates waves */
1.145 brouard 10495:
1.215 brouard 10496: free_vector(moisdc,1,n);
10497: free_vector(andc,1,n);
10498:
1.126 brouard 10499: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
10500: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
10501: ncodemax[1]=1;
1.145 brouard 10502: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 10503: cptcoveff=0;
1.220 brouard 10504: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
10505: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 10506: }
10507:
10508: ncovcombmax=pow(2,cptcoveff);
10509: invalidvarcomb=ivector(1, ncovcombmax);
10510: for(i=1;i<ncovcombmax;i++)
10511: invalidvarcomb[i]=0;
10512:
1.211 brouard 10513: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 10514: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 10515: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 10516:
1.200 brouard 10517: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 10518: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 10519: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 10520: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
10521: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
10522: * (currently 0 or 1) in the data.
10523: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
10524: * corresponding modality (h,j).
10525: */
10526:
1.145 brouard 10527: h=0;
10528: /*if (cptcovn > 0) */
1.126 brouard 10529: m=pow(2,cptcoveff);
10530:
1.144 brouard 10531: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 10532: * For k=4 covariates, h goes from 1 to m=2**k
10533: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
10534: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 10535: * h\k 1 2 3 4
1.143 brouard 10536: *______________________________
10537: * 1 i=1 1 i=1 1 i=1 1 i=1 1
10538: * 2 2 1 1 1
10539: * 3 i=2 1 2 1 1
10540: * 4 2 2 1 1
10541: * 5 i=3 1 i=2 1 2 1
10542: * 6 2 1 2 1
10543: * 7 i=4 1 2 2 1
10544: * 8 2 2 2 1
1.197 brouard 10545: * 9 i=5 1 i=3 1 i=2 1 2
10546: * 10 2 1 1 2
10547: * 11 i=6 1 2 1 2
10548: * 12 2 2 1 2
10549: * 13 i=7 1 i=4 1 2 2
10550: * 14 2 1 2 2
10551: * 15 i=8 1 2 2 2
10552: * 16 2 2 2 2
1.143 brouard 10553: */
1.212 brouard 10554: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10555: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10556: * and the value of each covariate?
10557: * V1=1, V2=1, V3=2, V4=1 ?
10558: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10559: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10560: * In order to get the real value in the data, we use nbcode
10561: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10562: * We are keeping this crazy system in order to be able (in the future?)
10563: * to have more than 2 values (0 or 1) for a covariate.
10564: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10565: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10566: * bbbbbbbb
10567: * 76543210
10568: * h-1 00000101 (6-1=5)
1.219 brouard 10569: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10570: * &
10571: * 1 00000001 (1)
1.219 brouard 10572: * 00000000 = 1 & ((h-1) >> (k-1))
10573: * +1= 00000001 =1
1.211 brouard 10574: *
10575: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10576: * h' 1101 =2^3+2^2+0x2^1+2^0
10577: * >>k' 11
10578: * & 00000001
10579: * = 00000001
10580: * +1 = 00000010=2 = codtabm(14,3)
10581: * Reverse h=6 and m=16?
10582: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10583: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10584: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10585: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10586: * V3=decodtabm(14,3,2**4)=2
10587: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10588: *(h-1) >> (j-1) 0011 =13 >> 2
10589: * &1 000000001
10590: * = 000000001
10591: * +1= 000000010 =2
10592: * 2211
10593: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10594: * V3=2
1.220 brouard 10595: * codtabm and decodtabm are identical
1.211 brouard 10596: */
10597:
1.145 brouard 10598:
10599: free_ivector(Ndum,-1,NCOVMAX);
10600:
10601:
1.126 brouard 10602:
1.186 brouard 10603: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10604: strcpy(optionfilegnuplot,optionfilefiname);
10605: if(mle==-3)
1.201 brouard 10606: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10607: strcat(optionfilegnuplot,".gp");
10608:
10609: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10610: printf("Problem with file %s",optionfilegnuplot);
10611: }
10612: else{
1.204 brouard 10613: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10614: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10615: //fprintf(ficgp,"set missing 'NaNq'\n");
10616: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10617: }
10618: /* fclose(ficgp);*/
1.186 brouard 10619:
10620:
10621: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10622:
10623: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10624: if(mle==-3)
1.201 brouard 10625: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10626: strcat(optionfilehtm,".htm");
10627: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10628: printf("Problem with %s \n",optionfilehtm);
10629: exit(0);
1.126 brouard 10630: }
10631:
10632: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10633: strcat(optionfilehtmcov,"-cov.htm");
10634: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10635: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10636: }
10637: else{
10638: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10639: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10640: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10641: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10642: }
10643:
1.213 brouard 10644: fprintf(fichtm,"<html><head>\n<head>\n<meta charset=\"utf-8\"/><meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<title>IMaCh %s</title></head>\n <body><font size=\"7\"><a href=http:/euroreves.ined.fr/imach>IMaCh for Interpolated Markov Chain</a> </font><br>\n<font size=\"3\">Sponsored by Copyright (C) 2002-2015 <a href=http://www.ined.fr>INED</a>-EUROREVES-Institut de longévité-2013-2016-Japan Society for the Promotion of Sciences 日本学術振興会 (<a href=https://www.jsps.go.jp/english/e-grants/>Grant-in-Aid for Scientific Research 25293121</a>) - <a href=https://software.intel.com/en-us>Intel Software 2015-2018</a></font><br> \
1.204 brouard 10645: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10646: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10647: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10648: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10649: \n\
10650: <hr size=\"2\" color=\"#EC5E5E\">\
10651: <ul><li><h4>Parameter files</h4>\n\
10652: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10653: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10654: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10655: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10656: - Date and time at start: %s</ul>\n",\
10657: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10658: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10659: fileres,fileres,\
10660: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10661: fflush(fichtm);
10662:
10663: strcpy(pathr,path);
10664: strcat(pathr,optionfilefiname);
1.184 brouard 10665: #ifdef WIN32
10666: _chdir(optionfilefiname); /* Move to directory named optionfile */
10667: #else
1.126 brouard 10668: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10669: #endif
10670:
1.126 brouard 10671:
1.220 brouard 10672: /* Calculates basic frequencies. Computes observed prevalence at single age
10673: and for any valid combination of covariates
1.126 brouard 10674: and prints on file fileres'p'. */
1.251 brouard 10675: freqsummary(fileres, p, pstart, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
1.227 brouard 10676: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10677:
10678: fprintf(fichtm,"\n");
10679: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10680: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10681: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10682: imx,agemin,agemax,jmin,jmax,jmean);
10683: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10684: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10685: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10686: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10687: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10688:
1.126 brouard 10689: /* For Powell, parameters are in a vector p[] starting at p[1]
10690: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10691: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10692:
10693: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10694: /* For mortality only */
1.126 brouard 10695: if (mle==-3){
1.136 brouard 10696: ximort=matrix(1,NDIM,1,NDIM);
1.248 brouard 10697: for(i=1;i<=NDIM;i++)
10698: for(j=1;j<=NDIM;j++)
10699: ximort[i][j]=0.;
1.186 brouard 10700: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10701: cens=ivector(1,n);
10702: ageexmed=vector(1,n);
10703: agecens=vector(1,n);
10704: dcwave=ivector(1,n);
1.223 brouard 10705:
1.126 brouard 10706: for (i=1; i<=imx; i++){
10707: dcwave[i]=-1;
10708: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10709: if (s[m][i]>nlstate) {
10710: dcwave[i]=m;
10711: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10712: break;
10713: }
1.126 brouard 10714: }
1.226 brouard 10715:
1.126 brouard 10716: for (i=1; i<=imx; i++) {
10717: if (wav[i]>0){
1.226 brouard 10718: ageexmed[i]=agev[mw[1][i]][i];
10719: j=wav[i];
10720: agecens[i]=1.;
10721:
10722: if (ageexmed[i]> 1 && wav[i] > 0){
10723: agecens[i]=agev[mw[j][i]][i];
10724: cens[i]= 1;
10725: }else if (ageexmed[i]< 1)
10726: cens[i]= -1;
10727: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10728: cens[i]=0 ;
1.126 brouard 10729: }
10730: else cens[i]=-1;
10731: }
10732:
10733: for (i=1;i<=NDIM;i++) {
10734: for (j=1;j<=NDIM;j++)
1.226 brouard 10735: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10736: }
10737:
1.145 brouard 10738: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10739: /*printf("%lf %lf", p[1], p[2]);*/
10740:
10741:
1.136 brouard 10742: #ifdef GSL
10743: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10744: #else
1.126 brouard 10745: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10746: #endif
1.201 brouard 10747: strcpy(filerespow,"POW-MORT_");
10748: strcat(filerespow,fileresu);
1.126 brouard 10749: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10750: printf("Problem with resultfile: %s\n", filerespow);
10751: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10752: }
1.136 brouard 10753: #ifdef GSL
10754: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10755: #else
1.126 brouard 10756: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10757: #endif
1.126 brouard 10758: /* for (i=1;i<=nlstate;i++)
10759: for(j=1;j<=nlstate+ndeath;j++)
10760: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10761: */
10762: fprintf(ficrespow,"\n");
1.136 brouard 10763: #ifdef GSL
10764: /* gsl starts here */
10765: T = gsl_multimin_fminimizer_nmsimplex;
10766: gsl_multimin_fminimizer *sfm = NULL;
10767: gsl_vector *ss, *x;
10768: gsl_multimin_function minex_func;
10769:
10770: /* Initial vertex size vector */
10771: ss = gsl_vector_alloc (NDIM);
10772:
10773: if (ss == NULL){
10774: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10775: }
10776: /* Set all step sizes to 1 */
10777: gsl_vector_set_all (ss, 0.001);
10778:
10779: /* Starting point */
1.126 brouard 10780:
1.136 brouard 10781: x = gsl_vector_alloc (NDIM);
10782:
10783: if (x == NULL){
10784: gsl_vector_free(ss);
10785: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10786: }
10787:
10788: /* Initialize method and iterate */
10789: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10790: /* gsl_vector_set(x, 0, 0.0268); */
10791: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10792: gsl_vector_set(x, 0, p[1]);
10793: gsl_vector_set(x, 1, p[2]);
10794:
10795: minex_func.f = &gompertz_f;
10796: minex_func.n = NDIM;
10797: minex_func.params = (void *)&p; /* ??? */
10798:
10799: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10800: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10801:
10802: printf("Iterations beginning .....\n\n");
10803: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10804:
10805: iteri=0;
10806: while (rval == GSL_CONTINUE){
10807: iteri++;
10808: status = gsl_multimin_fminimizer_iterate(sfm);
10809:
10810: if (status) printf("error: %s\n", gsl_strerror (status));
10811: fflush(0);
10812:
10813: if (status)
10814: break;
10815:
10816: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10817: ssval = gsl_multimin_fminimizer_size (sfm);
10818:
10819: if (rval == GSL_SUCCESS)
10820: printf ("converged to a local maximum at\n");
10821:
10822: printf("%5d ", iteri);
10823: for (it = 0; it < NDIM; it++){
10824: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10825: }
10826: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10827: }
10828:
10829: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10830:
10831: gsl_vector_free(x); /* initial values */
10832: gsl_vector_free(ss); /* inital step size */
10833: for (it=0; it<NDIM; it++){
10834: p[it+1]=gsl_vector_get(sfm->x,it);
10835: fprintf(ficrespow," %.12lf", p[it]);
10836: }
10837: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10838: #endif
10839: #ifdef POWELL
10840: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10841: #endif
1.126 brouard 10842: fclose(ficrespow);
10843:
1.203 brouard 10844: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10845:
10846: for(i=1; i <=NDIM; i++)
10847: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10848: matcov[i][j]=matcov[j][i];
1.126 brouard 10849:
10850: printf("\nCovariance matrix\n ");
1.203 brouard 10851: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10852: for(i=1; i <=NDIM; i++) {
10853: for(j=1;j<=NDIM;j++){
1.220 brouard 10854: printf("%f ",matcov[i][j]);
10855: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10856: }
1.203 brouard 10857: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10858: }
10859:
10860: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10861: for (i=1;i<=NDIM;i++) {
1.126 brouard 10862: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10863: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10864: }
1.126 brouard 10865: lsurv=vector(1,AGESUP);
10866: lpop=vector(1,AGESUP);
10867: tpop=vector(1,AGESUP);
10868: lsurv[agegomp]=100000;
10869:
10870: for (k=agegomp;k<=AGESUP;k++) {
10871: agemortsup=k;
10872: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10873: }
10874:
10875: for (k=agegomp;k<agemortsup;k++)
10876: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10877:
10878: for (k=agegomp;k<agemortsup;k++){
10879: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10880: sumlpop=sumlpop+lpop[k];
10881: }
10882:
10883: tpop[agegomp]=sumlpop;
10884: for (k=agegomp;k<(agemortsup-3);k++){
10885: /* tpop[k+1]=2;*/
10886: tpop[k+1]=tpop[k]-lpop[k];
10887: }
10888:
10889:
10890: printf("\nAge lx qx dx Lx Tx e(x)\n");
10891: for (k=agegomp;k<(agemortsup-2);k++)
10892: printf("%d %.0lf %lf %.0lf %.0lf %.0lf %lf\n",k,lsurv[k],p[1]*exp(p[2]*(k-agegomp)),(p[1]*exp(p[2]*(k-agegomp)))*lsurv[k],lpop[k],tpop[k],tpop[k]/lsurv[k]);
10893:
10894:
10895: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10896: ageminpar=50;
10897: agemaxpar=100;
1.194 brouard 10898: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10899: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10900: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10901: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10902: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10903: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10904: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10905: }else{
10906: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10907: fprintf(ficlog,"Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
1.201 brouard 10908: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10909: }
1.201 brouard 10910: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10911: stepm, weightopt,\
10912: model,imx,p,matcov,agemortsup);
10913:
10914: free_vector(lsurv,1,AGESUP);
10915: free_vector(lpop,1,AGESUP);
10916: free_vector(tpop,1,AGESUP);
1.220 brouard 10917: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10918: free_ivector(cens,1,n);
10919: free_vector(agecens,1,n);
10920: free_ivector(dcwave,1,n);
1.220 brouard 10921: #ifdef GSL
1.136 brouard 10922: #endif
1.186 brouard 10923: } /* Endof if mle==-3 mortality only */
1.205 brouard 10924: /* Standard */
10925: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10926: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10927: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10928: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10929: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10930: for (k=1; k<=npar;k++)
10931: printf(" %d %8.5f",k,p[k]);
10932: printf("\n");
1.205 brouard 10933: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10934: /* mlikeli uses func not funcone */
1.247 brouard 10935: /* for(i=1;i<nlstate;i++){ */
10936: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10937: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10938: /* } */
1.205 brouard 10939: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10940: }
10941: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10942: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10943: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10944: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10945: }
10946: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10947: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10948: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10949: for (k=1; k<=npar;k++)
10950: printf(" %d %8.5f",k,p[k]);
10951: printf("\n");
10952:
10953: /*--------- results files --------------*/
1.224 brouard 10954: fprintf(ficres,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle= 0 weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol, stepm, ncovcol, nqv, ntv, nqtv, nlstate, ndeath, maxwav, weightopt,model);
1.126 brouard 10955:
10956:
10957: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10958: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10959: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10960: for(i=1,jk=1; i <=nlstate; i++){
10961: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10962: if (k != i) {
10963: printf("%d%d ",i,k);
10964: fprintf(ficlog,"%d%d ",i,k);
10965: fprintf(ficres,"%1d%1d ",i,k);
10966: for(j=1; j <=ncovmodel; j++){
10967: printf("%12.7f ",p[jk]);
10968: fprintf(ficlog,"%12.7f ",p[jk]);
10969: fprintf(ficres,"%12.7f ",p[jk]);
10970: jk++;
10971: }
10972: printf("\n");
10973: fprintf(ficlog,"\n");
10974: fprintf(ficres,"\n");
10975: }
1.126 brouard 10976: }
10977: }
1.203 brouard 10978: if(mle != 0){
10979: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10980: ftolhess=ftol; /* Usually correct */
1.203 brouard 10981: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10982: printf("Parameters and 95%% confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W .\n But be careful that parameters are highly correlated because incidence of disability is highly correlated to incidence of recovery.\n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
10983: fprintf(ficlog, "Parameters, Wald tests and Wald-based confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W \n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
10984: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10985: for(k=1; k <=(nlstate+ndeath); k++){
10986: if (k != i) {
10987: printf("%d%d ",i,k);
10988: fprintf(ficlog,"%d%d ",i,k);
10989: for(j=1; j <=ncovmodel; j++){
10990: printf("%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk]));
10991: fprintf(ficlog,"%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk]));
10992: jk++;
10993: }
10994: printf("\n");
10995: fprintf(ficlog,"\n");
10996: }
10997: }
1.193 brouard 10998: }
1.203 brouard 10999: } /* end of hesscov and Wald tests */
1.225 brouard 11000:
1.203 brouard 11001: /* */
1.126 brouard 11002: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
11003: printf("# Scales (for hessian or gradient estimation)\n");
11004: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
11005: for(i=1,jk=1; i <=nlstate; i++){
11006: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 11007: if (j!=i) {
11008: fprintf(ficres,"%1d%1d",i,j);
11009: printf("%1d%1d",i,j);
11010: fprintf(ficlog,"%1d%1d",i,j);
11011: for(k=1; k<=ncovmodel;k++){
11012: printf(" %.5e",delti[jk]);
11013: fprintf(ficlog," %.5e",delti[jk]);
11014: fprintf(ficres," %.5e",delti[jk]);
11015: jk++;
11016: }
11017: printf("\n");
11018: fprintf(ficlog,"\n");
11019: fprintf(ficres,"\n");
11020: }
1.126 brouard 11021: }
11022: }
11023:
11024: fprintf(ficres,"# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
1.203 brouard 11025: if(mle >= 1) /* To big for the screen */
1.126 brouard 11026: printf("# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
11027: fprintf(ficlog,"# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
11028: /* # 121 Var(a12)\n\ */
11029: /* # 122 Cov(b12,a12) Var(b12)\n\ */
11030: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
11031: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
11032: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
11033: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
11034: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
11035: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
11036:
11037:
11038: /* Just to have a covariance matrix which will be more understandable
11039: even is we still don't want to manage dictionary of variables
11040: */
11041: for(itimes=1;itimes<=2;itimes++){
11042: jj=0;
11043: for(i=1; i <=nlstate; i++){
1.225 brouard 11044: for(j=1; j <=nlstate+ndeath; j++){
11045: if(j==i) continue;
11046: for(k=1; k<=ncovmodel;k++){
11047: jj++;
11048: ca[0]= k+'a'-1;ca[1]='\0';
11049: if(itimes==1){
11050: if(mle>=1)
11051: printf("#%1d%1d%d",i,j,k);
11052: fprintf(ficlog,"#%1d%1d%d",i,j,k);
11053: fprintf(ficres,"#%1d%1d%d",i,j,k);
11054: }else{
11055: if(mle>=1)
11056: printf("%1d%1d%d",i,j,k);
11057: fprintf(ficlog,"%1d%1d%d",i,j,k);
11058: fprintf(ficres,"%1d%1d%d",i,j,k);
11059: }
11060: ll=0;
11061: for(li=1;li <=nlstate; li++){
11062: for(lj=1;lj <=nlstate+ndeath; lj++){
11063: if(lj==li) continue;
11064: for(lk=1;lk<=ncovmodel;lk++){
11065: ll++;
11066: if(ll<=jj){
11067: cb[0]= lk +'a'-1;cb[1]='\0';
11068: if(ll<jj){
11069: if(itimes==1){
11070: if(mle>=1)
11071: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11072: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11073: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11074: }else{
11075: if(mle>=1)
11076: printf(" %.5e",matcov[jj][ll]);
11077: fprintf(ficlog," %.5e",matcov[jj][ll]);
11078: fprintf(ficres," %.5e",matcov[jj][ll]);
11079: }
11080: }else{
11081: if(itimes==1){
11082: if(mle>=1)
11083: printf(" Var(%s%1d%1d)",ca,i,j);
11084: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
11085: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
11086: }else{
11087: if(mle>=1)
11088: printf(" %.7e",matcov[jj][ll]);
11089: fprintf(ficlog," %.7e",matcov[jj][ll]);
11090: fprintf(ficres," %.7e",matcov[jj][ll]);
11091: }
11092: }
11093: }
11094: } /* end lk */
11095: } /* end lj */
11096: } /* end li */
11097: if(mle>=1)
11098: printf("\n");
11099: fprintf(ficlog,"\n");
11100: fprintf(ficres,"\n");
11101: numlinepar++;
11102: } /* end k*/
11103: } /*end j */
1.126 brouard 11104: } /* end i */
11105: } /* end itimes */
11106:
11107: fflush(ficlog);
11108: fflush(ficres);
1.225 brouard 11109: while(fgets(line, MAXLINE, ficpar)) {
11110: /* If line starts with a # it is a comment */
11111: if (line[0] == '#') {
11112: numlinepar++;
11113: fputs(line,stdout);
11114: fputs(line,ficparo);
11115: fputs(line,ficlog);
11116: continue;
11117: }else
11118: break;
11119: }
11120:
1.209 brouard 11121: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
11122: /* ungetc(c,ficpar); */
11123: /* fgets(line, MAXLINE, ficpar); */
11124: /* fputs(line,stdout); */
11125: /* fputs(line,ficparo); */
11126: /* } */
11127: /* ungetc(c,ficpar); */
1.126 brouard 11128:
11129: estepm=0;
1.209 brouard 11130: if((num_filled=sscanf(line,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm, &ftolpl)) !=EOF){
1.225 brouard 11131:
11132: if (num_filled != 6) {
11133: printf("Error: Not 6 parameters in line, for example:agemin=60 agemax=95 bage=55 fage=95 estepm=24 ftolpl=6e-4\n, your line=%s . Probably you are running an older format.\n",line);
11134: fprintf(ficlog,"Error: Not 6 parameters in line, for example:agemin=60 agemax=95 bage=55 fage=95 estepm=24 ftolpl=6e-4\n, your line=%s . Probably you are running an older format.\n",line);
11135: goto end;
11136: }
11137: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
11138: }
11139: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
11140: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
11141:
1.209 brouard 11142: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 11143: if (estepm==0 || estepm < stepm) estepm=stepm;
11144: if (fage <= 2) {
11145: bage = ageminpar;
11146: fage = agemaxpar;
11147: }
11148:
11149: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 11150: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
11151: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 11152:
1.186 brouard 11153: /* Other stuffs, more or less useful */
1.254 brouard 11154: while(fgets(line, MAXLINE, ficpar)) {
11155: /* If line starts with a # it is a comment */
11156: if (line[0] == '#') {
11157: numlinepar++;
11158: fputs(line,stdout);
11159: fputs(line,ficparo);
11160: fputs(line,ficlog);
11161: continue;
11162: }else
11163: break;
11164: }
11165:
11166: if((num_filled=sscanf(line,"begin-prev-date=%lf/%lf/%lf end-prev-date=%lf/%lf/%lf mov_average=%d\n",&jprev1, &mprev1,&anprev1,&jprev2, &mprev2,&anprev2,&mobilav)) !=EOF){
11167:
11168: if (num_filled != 7) {
11169: printf("Error: Not 7 (data)parameters in line but %d, for example:begin-prev-date=1/1/1990 end-prev-date=1/6/2004 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11170: fprintf(ficlog,"Error: Not 7 (data)parameters in line but %d, for example:begin-prev-date=1/1/1990 end-prev-date=1/6/2004 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11171: goto end;
11172: }
11173: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11174: fprintf(ficparo,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11175: fprintf(ficres,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11176: fprintf(ficlog,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
1.126 brouard 11177: }
1.254 brouard 11178:
11179: while(fgets(line, MAXLINE, ficpar)) {
11180: /* If line starts with a # it is a comment */
11181: if (line[0] == '#') {
11182: numlinepar++;
11183: fputs(line,stdout);
11184: fputs(line,ficparo);
11185: fputs(line,ficlog);
11186: continue;
11187: }else
11188: break;
1.126 brouard 11189: }
11190:
11191:
11192: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
11193: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
11194:
1.254 brouard 11195: if((num_filled=sscanf(line,"pop_based=%d\n",&popbased)) !=EOF){
11196: if (num_filled != 1) {
11197: printf("Error: Not 1 (data)parameters in line but %d, for example:pop_based=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11198: fprintf(ficlog,"Error: Not 1 (data)parameters in line but %d, for example: pop_based=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11199: goto end;
11200: }
11201: printf("pop_based=%d\n",popbased);
11202: fprintf(ficlog,"pop_based=%d\n",popbased);
11203: fprintf(ficparo,"pop_based=%d\n",popbased);
11204: fprintf(ficres,"pop_based=%d\n",popbased);
11205: }
11206:
1.258 brouard 11207: /* Results */
11208: nresult=0;
11209: do{
11210: if(!fgets(line, MAXLINE, ficpar)){
11211: endishere=1;
11212: parameterline=14;
11213: }else if (line[0] == '#') {
11214: /* If line starts with a # it is a comment */
1.254 brouard 11215: numlinepar++;
11216: fputs(line,stdout);
11217: fputs(line,ficparo);
11218: fputs(line,ficlog);
11219: continue;
1.258 brouard 11220: }else if(sscanf(line,"prevforecast=%[^\n]\n",modeltemp))
11221: parameterline=11;
11222: else if(sscanf(line,"backcast=%[^\n]\n",modeltemp))
11223: parameterline=12;
11224: else if(sscanf(line,"result:%[^\n]\n",modeltemp))
11225: parameterline=13;
11226: else{
11227: parameterline=14;
1.254 brouard 11228: }
1.258 brouard 11229: switch (parameterline){
11230: case 11:
11231: if((num_filled=sscanf(line,"prevforecast=%d starting-proj-date=%lf/%lf/%lf final-proj-date=%lf/%lf/%lf mobil_average=%d\n",&prevfcast,&jproj1,&mproj1,&anproj1,&jproj2,&mproj2,&anproj2,&mobilavproj)) !=EOF){
11232: if (num_filled != 8) {
11233: printf("Error: Not 8 (data)parameters in line but %d, for example:prevforecast=1 starting-proj-date=1/1/1990 final-proj-date=1/1/2000 mobil_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11234: fprintf(ficlog,"Error: Not 8 (data)parameters in line but %d, for example:prevforecast=1 starting-proj-date=1/1/1990 final-proj-date=1/1/2000 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11235: goto end;
11236: }
11237: fprintf(ficparo,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11238: printf("prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11239: fprintf(ficlog,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11240: fprintf(ficres,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11241: /* day and month of proj2 are not used but only year anproj2.*/
11242: }
1.254 brouard 11243: break;
1.258 brouard 11244: case 12:
11245: /*fscanf(ficpar,"backcast=%d starting-back-date=%lf/%lf/%lf final-back-date=%lf/%lf/%lf mobil_average=%d\n",&backcast,&jback1,&mback1,&anback1,&jback2,&mback2,&anback2,&mobilavproj);*/
11246: if((num_filled=sscanf(line,"backcast=%d starting-back-date=%lf/%lf/%lf final-back-date=%lf/%lf/%lf mobil_average=%d\n",&backcast,&jback1,&mback1,&anback1,&jback2,&mback2,&anback2,&mobilavproj)) !=EOF){
11247: if (num_filled != 8) {
1.262 ! brouard 11248: printf("Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 final-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
! 11249: fprintf(ficlog,"Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 final-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
1.258 brouard 11250: goto end;
11251: }
11252: printf("backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11253: fprintf(ficparo,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11254: fprintf(ficlog,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11255: fprintf(ficres,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11256: /* day and month of proj2 are not used but only year anproj2.*/
11257: }
1.230 brouard 11258: break;
1.258 brouard 11259: case 13:
11260: if((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
11261: if (num_filled == 0){
11262: resultline[0]='\0';
11263: printf("Warning %d: no result line! It should be at minimum 'result: V2=0 V1=1 or result:.\n%s\n", num_filled, line);
11264: fprintf(ficlog,"Warning %d: no result line! It should be at minimum 'result: V2=0 V1=1 or result:.\n%s\n", num_filled, line);
11265: break;
11266: } else if (num_filled != 1){
11267: printf("ERROR %d: result line! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",num_filled, line);
11268: fprintf(ficlog,"ERROR %d: result line! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",num_filled, line);
11269: }
11270: nresult++; /* Sum of resultlines */
11271: printf("Result %d: result=%s\n",nresult, resultline);
11272: if(nresult > MAXRESULTLINES){
11273: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11274: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11275: goto end;
11276: }
11277: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
11278: fprintf(ficparo,"result: %s\n",resultline);
11279: fprintf(ficres,"result: %s\n",resultline);
11280: fprintf(ficlog,"result: %s\n",resultline);
1.230 brouard 11281: break;
1.258 brouard 11282: case 14:
1.259 brouard 11283: if(ncovmodel >2 && nresult==0 ){
11284: printf("ERROR: no result lines! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",line);
1.258 brouard 11285: goto end;
11286: }
1.259 brouard 11287: break;
1.258 brouard 11288: default:
11289: nresult=1;
11290: decoderesult(".",nresult ); /* No covariate */
11291: }
11292: } /* End switch parameterline */
11293: }while(endishere==0); /* End do */
1.126 brouard 11294:
1.230 brouard 11295: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 11296: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 11297:
11298: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 11299: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 11300: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11301: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11302: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 11303: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11304: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11305: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11306: }else{
1.218 brouard 11307: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 11308: }
11309: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.258 brouard 11310: model,imx,jmin,jmax,jmean,rfileres,popforecast,mobilav,prevfcast,mobilavproj,backcast, estepm, \
1.225 brouard 11311: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 11312:
1.225 brouard 11313: /*------------ free_vector -------------*/
11314: /* chdir(path); */
1.220 brouard 11315:
1.215 brouard 11316: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
11317: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
11318: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
11319: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 11320: free_lvector(num,1,n);
11321: free_vector(agedc,1,n);
11322: /*free_matrix(covar,0,NCOVMAX,1,n);*/
11323: /*free_matrix(covar,1,NCOVMAX,1,n);*/
11324: fclose(ficparo);
11325: fclose(ficres);
1.220 brouard 11326:
11327:
1.186 brouard 11328: /* Other results (useful)*/
1.220 brouard 11329:
11330:
1.126 brouard 11331: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 11332: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
11333: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 11334: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 11335: fclose(ficrespl);
11336:
11337: /*------------- h Pij x at various ages ------------*/
1.180 brouard 11338: /*#include "hpijx.h"*/
11339: hPijx(p, bage, fage);
1.145 brouard 11340: fclose(ficrespij);
1.227 brouard 11341:
1.220 brouard 11342: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 11343: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 11344: k=1;
1.126 brouard 11345: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 11346:
1.219 brouard 11347: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 11348: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 11349: for(i=1;i<=AGESUP;i++)
1.219 brouard 11350: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 11351: for(k=1;k<=ncovcombmax;k++)
11352: probs[i][j][k]=0.;
1.219 brouard 11353: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
11354: if (mobilav!=0 ||mobilavproj !=0 ) {
11355: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 11356: for(i=1;i<=AGESUP;i++)
11357: for(j=1;j<=nlstate;j++)
11358: for(k=1;k<=ncovcombmax;k++)
11359: mobaverages[i][j][k]=0.;
1.219 brouard 11360: mobaverage=mobaverages;
11361: if (mobilav!=0) {
1.235 brouard 11362: printf("Movingaveraging observed prevalence\n");
1.258 brouard 11363: fprintf(ficlog,"Movingaveraging observed prevalence\n");
1.227 brouard 11364: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
11365: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
11366: printf(" Error in movingaverage mobilav=%d\n",mobilav);
11367: }
1.219 brouard 11368: }
11369: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
11370: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
11371: else if (mobilavproj !=0) {
1.235 brouard 11372: printf("Movingaveraging projected observed prevalence\n");
1.258 brouard 11373: fprintf(ficlog,"Movingaveraging projected observed prevalence\n");
1.227 brouard 11374: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
11375: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
11376: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
11377: }
1.219 brouard 11378: }
11379: }/* end if moving average */
1.227 brouard 11380:
1.126 brouard 11381: /*---------- Forecasting ------------------*/
11382: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
11383: if(prevfcast==1){
11384: /* if(stepm ==1){*/
1.225 brouard 11385: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 11386: }
1.217 brouard 11387: if(backcast==1){
1.219 brouard 11388: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11389: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11390: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11391:
11392: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
11393:
11394: bprlim=matrix(1,nlstate,1,nlstate);
11395: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
11396: fclose(ficresplb);
11397:
1.222 brouard 11398: hBijx(p, bage, fage, mobaverage);
11399: fclose(ficrespijb);
1.219 brouard 11400: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
11401:
11402: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 11403: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 11404: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11405: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11406: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11407: }
1.217 brouard 11408:
1.186 brouard 11409:
11410: /* ------ Other prevalence ratios------------ */
1.126 brouard 11411:
1.215 brouard 11412: free_ivector(wav,1,imx);
11413: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
11414: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
11415: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 11416:
11417:
1.127 brouard 11418: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 11419:
1.201 brouard 11420: strcpy(filerese,"E_");
11421: strcat(filerese,fileresu);
1.126 brouard 11422: if((ficreseij=fopen(filerese,"w"))==NULL) {
11423: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11424: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11425: }
1.208 brouard 11426: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
11427: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.238 brouard 11428:
11429: pstamp(ficreseij);
1.219 brouard 11430:
1.235 brouard 11431: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11432: if (cptcovn < 1){i1=1;}
11433:
11434: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11435: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11436: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11437: continue;
1.219 brouard 11438: fprintf(ficreseij,"\n#****** ");
1.235 brouard 11439: printf("\n#****** ");
1.225 brouard 11440: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11441: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11442: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11443: }
11444: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11445: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11446: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 11447: }
11448: fprintf(ficreseij,"******\n");
1.235 brouard 11449: printf("******\n");
1.219 brouard 11450:
11451: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11452: oldm=oldms;savm=savms;
1.235 brouard 11453: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 11454:
1.219 brouard 11455: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 11456: }
11457: fclose(ficreseij);
1.208 brouard 11458: printf("done evsij\n");fflush(stdout);
11459: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 11460:
1.227 brouard 11461: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 11462:
11463:
1.201 brouard 11464: strcpy(filerest,"T_");
11465: strcat(filerest,fileresu);
1.127 brouard 11466: if((ficrest=fopen(filerest,"w"))==NULL) {
11467: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
11468: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
11469: }
1.208 brouard 11470: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
11471: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 11472:
1.126 brouard 11473:
1.201 brouard 11474: strcpy(fileresstde,"STDE_");
11475: strcat(fileresstde,fileresu);
1.126 brouard 11476: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 11477: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
11478: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 11479: }
1.227 brouard 11480: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
11481: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 11482:
1.201 brouard 11483: strcpy(filerescve,"CVE_");
11484: strcat(filerescve,fileresu);
1.126 brouard 11485: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 11486: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
11487: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 11488: }
1.227 brouard 11489: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
11490: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 11491:
1.201 brouard 11492: strcpy(fileresv,"V_");
11493: strcat(fileresv,fileresu);
1.126 brouard 11494: if((ficresvij=fopen(fileresv,"w"))==NULL) {
11495: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
11496: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
11497: }
1.227 brouard 11498: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
11499: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 11500:
1.145 brouard 11501: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11502: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
11503:
1.235 brouard 11504: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11505: if (cptcovn < 1){i1=1;}
11506:
11507: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11508: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11509: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11510: continue;
1.242 brouard 11511: printf("\n#****** Result for:");
11512: fprintf(ficrest,"\n#****** Result for:");
11513: fprintf(ficlog,"\n#****** Result for:");
1.227 brouard 11514: for(j=1;j<=cptcoveff;j++){
11515: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11516: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11517: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11518: }
1.235 brouard 11519: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11520: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11521: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11522: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11523: }
1.208 brouard 11524: fprintf(ficrest,"******\n");
1.227 brouard 11525: fprintf(ficlog,"******\n");
11526: printf("******\n");
1.208 brouard 11527:
11528: fprintf(ficresstdeij,"\n#****** ");
11529: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 11530: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11531: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11532: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 11533: }
1.235 brouard 11534: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11535: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11536: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11537: }
1.208 brouard 11538: fprintf(ficresstdeij,"******\n");
11539: fprintf(ficrescveij,"******\n");
11540:
11541: fprintf(ficresvij,"\n#****** ");
1.238 brouard 11542: /* pstamp(ficresvij); */
1.225 brouard 11543: for(j=1;j<=cptcoveff;j++)
1.227 brouard 11544: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11545: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11546: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11547: }
1.208 brouard 11548: fprintf(ficresvij,"******\n");
11549:
11550: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11551: oldm=oldms;savm=savms;
1.235 brouard 11552: printf(" cvevsij ");
11553: fprintf(ficlog, " cvevsij ");
11554: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 11555: printf(" end cvevsij \n ");
11556: fprintf(ficlog, " end cvevsij \n ");
11557:
11558: /*
11559: */
11560: /* goto endfree; */
11561:
11562: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11563: pstamp(ficrest);
11564:
11565:
11566: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 11567: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
11568: cptcod= 0; /* To be deleted */
11569: printf("varevsij vpopbased=%d \n",vpopbased);
11570: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 11571: varevsij(optionfilefiname, vareij, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, estepm, cptcov,cptcod,vpopbased,mobilav, strstart, nres); /* cptcod not initialized Intel */
1.227 brouard 11572: fprintf(ficrest,"# Total life expectancy with std error and decomposition into time to be expected in each health state\n# (weighted average of eij where weights are ");
11573: if(vpopbased==1)
11574: fprintf(ficrest,"the age specific prevalence observed (cross-sectionally) in the population i.e cross-sectionally\n in each health state (popbased=1) (mobilav=%d)\n",mobilav);
11575: else
11576: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
11577: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
11578: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
11579: fprintf(ficrest,"\n");
11580: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
11581: epj=vector(1,nlstate+1);
11582: printf("Computing age specific period (stable) prevalences in each health state \n");
11583: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11584: for(age=bage; age <=fage ;age++){
1.235 brouard 11585: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11586: if (vpopbased==1) {
11587: if(mobilav ==0){
11588: for(i=1; i<=nlstate;i++)
11589: prlim[i][i]=probs[(int)age][i][k];
11590: }else{ /* mobilav */
11591: for(i=1; i<=nlstate;i++)
11592: prlim[i][i]=mobaverage[(int)age][i][k];
11593: }
11594: }
1.219 brouard 11595:
1.227 brouard 11596: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11597: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11598: /* printf(" age %4.0f ",age); */
11599: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11600: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11601: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11602: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11603: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11604: }
11605: epj[nlstate+1] +=epj[j];
11606: }
11607: /* printf(" age %4.0f \n",age); */
1.219 brouard 11608:
1.227 brouard 11609: for(i=1, vepp=0.;i <=nlstate;i++)
11610: for(j=1;j <=nlstate;j++)
11611: vepp += vareij[i][j][(int)age];
11612: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11613: for(j=1;j <=nlstate;j++){
11614: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11615: }
11616: fprintf(ficrest,"\n");
11617: }
1.208 brouard 11618: } /* End vpopbased */
11619: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11620: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11621: free_vector(epj,1,nlstate+1);
1.235 brouard 11622: printf("done selection\n");fflush(stdout);
11623: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11624:
1.145 brouard 11625: /*}*/
1.235 brouard 11626: } /* End k selection */
1.227 brouard 11627:
11628: printf("done State-specific expectancies\n");fflush(stdout);
11629: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11630:
1.126 brouard 11631: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11632:
1.201 brouard 11633: strcpy(fileresvpl,"VPL_");
11634: strcat(fileresvpl,fileresu);
1.126 brouard 11635: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11636: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11637: exit(0);
11638: }
1.208 brouard 11639: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11640: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11641:
1.145 brouard 11642: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11643: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11644:
1.235 brouard 11645: i1=pow(2,cptcoveff);
11646: if (cptcovn < 1){i1=1;}
11647:
11648: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11649: for(k=1; k<=i1;k++){
1.253 brouard 11650: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11651: continue;
1.227 brouard 11652: fprintf(ficresvpl,"\n#****** ");
11653: printf("\n#****** ");
11654: fprintf(ficlog,"\n#****** ");
11655: for(j=1;j<=cptcoveff;j++) {
11656: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11657: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11658: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11659: }
1.235 brouard 11660: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11661: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11662: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11663: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11664: }
1.227 brouard 11665: fprintf(ficresvpl,"******\n");
11666: printf("******\n");
11667: fprintf(ficlog,"******\n");
11668:
11669: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11670: oldm=oldms;savm=savms;
1.235 brouard 11671: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11672: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11673: /*}*/
1.126 brouard 11674: }
1.227 brouard 11675:
1.126 brouard 11676: fclose(ficresvpl);
1.208 brouard 11677: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11678: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11679:
11680: free_vector(weight,1,n);
11681: free_imatrix(Tvard,1,NCOVMAX,1,2);
11682: free_imatrix(s,1,maxwav+1,1,n);
11683: free_matrix(anint,1,maxwav,1,n);
11684: free_matrix(mint,1,maxwav,1,n);
11685: free_ivector(cod,1,n);
11686: free_ivector(tab,1,NCOVMAX);
11687: fclose(ficresstdeij);
11688: fclose(ficrescveij);
11689: fclose(ficresvij);
11690: fclose(ficrest);
11691: fclose(ficpar);
11692:
11693:
1.126 brouard 11694: /*---------- End : free ----------------*/
1.219 brouard 11695: if (mobilav!=0 ||mobilavproj !=0)
11696: free_ma3x(mobaverages,1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax); /* We need to have a squared matrix with prevalence of the dead! */
1.218 brouard 11697: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11698: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11699: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11700: } /* mle==-3 arrives here for freeing */
1.227 brouard 11701: /* endfree:*/
11702: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11703: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11704: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11705: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11706: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11707: free_matrix(coqvar,1,maxwav,1,n);
11708: free_matrix(covar,0,NCOVMAX,1,n);
11709: free_matrix(matcov,1,npar,1,npar);
11710: free_matrix(hess,1,npar,1,npar);
11711: /*free_vector(delti,1,npar);*/
11712: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11713: free_matrix(agev,1,maxwav,1,imx);
11714: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11715:
11716: free_ivector(ncodemax,1,NCOVMAX);
11717: free_ivector(ncodemaxwundef,1,NCOVMAX);
11718: free_ivector(Dummy,-1,NCOVMAX);
11719: free_ivector(Fixed,-1,NCOVMAX);
1.238 brouard 11720: free_ivector(DummyV,1,NCOVMAX);
11721: free_ivector(FixedV,1,NCOVMAX);
1.227 brouard 11722: free_ivector(Typevar,-1,NCOVMAX);
11723: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11724: free_ivector(TvarsQ,1,NCOVMAX);
11725: free_ivector(TvarsQind,1,NCOVMAX);
11726: free_ivector(TvarsD,1,NCOVMAX);
11727: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11728: free_ivector(TvarFD,1,NCOVMAX);
11729: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11730: free_ivector(TvarF,1,NCOVMAX);
11731: free_ivector(TvarFind,1,NCOVMAX);
11732: free_ivector(TvarV,1,NCOVMAX);
11733: free_ivector(TvarVind,1,NCOVMAX);
11734: free_ivector(TvarA,1,NCOVMAX);
11735: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11736: free_ivector(TvarFQ,1,NCOVMAX);
11737: free_ivector(TvarFQind,1,NCOVMAX);
11738: free_ivector(TvarVD,1,NCOVMAX);
11739: free_ivector(TvarVDind,1,NCOVMAX);
11740: free_ivector(TvarVQ,1,NCOVMAX);
11741: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11742: free_ivector(Tvarsel,1,NCOVMAX);
11743: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11744: free_ivector(Tposprod,1,NCOVMAX);
11745: free_ivector(Tprod,1,NCOVMAX);
11746: free_ivector(Tvaraff,1,NCOVMAX);
11747: free_ivector(invalidvarcomb,1,ncovcombmax);
11748: free_ivector(Tage,1,NCOVMAX);
11749: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11750: free_ivector(TmodelInvind,1,NCOVMAX);
11751: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11752:
11753: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11754: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11755: fflush(fichtm);
11756: fflush(ficgp);
11757:
1.227 brouard 11758:
1.126 brouard 11759: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11760: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11761: fprintf(ficlog,"End of Imach with %d errors and/or warnings %d. Please look at the log file for details.\n",nberr,nbwarn);
1.126 brouard 11762: }else{
11763: printf("End of Imach\n");
11764: fprintf(ficlog,"End of Imach\n");
11765: }
11766: printf("See log file on %s\n",filelog);
11767: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11768: /*(void) gettimeofday(&end_time,&tzp);*/
11769: rend_time = time(NULL);
11770: end_time = *localtime(&rend_time);
11771: /* tml = *localtime(&end_time.tm_sec); */
11772: strcpy(strtend,asctime(&end_time));
1.126 brouard 11773: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11774: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11775: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11776:
1.157 brouard 11777: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11778: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11779: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11780: /* printf("Total time was %d uSec.\n", total_usecs);*/
11781: /* if(fileappend(fichtm,optionfilehtm)){ */
11782: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11783: fclose(fichtm);
11784: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11785: fclose(fichtmcov);
11786: fclose(ficgp);
11787: fclose(ficlog);
11788: /*------ End -----------*/
1.227 brouard 11789:
11790:
11791: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11792: #ifdef WIN32
1.227 brouard 11793: if (_chdir(pathcd) != 0)
11794: printf("Can't move to directory %s!\n",path);
11795: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11796: #else
1.227 brouard 11797: if(chdir(pathcd) != 0)
11798: printf("Can't move to directory %s!\n", path);
11799: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11800: #endif
1.126 brouard 11801: printf("Current directory %s!\n",pathcd);
11802: /*strcat(plotcmd,CHARSEPARATOR);*/
11803: sprintf(plotcmd,"gnuplot");
1.157 brouard 11804: #ifdef _WIN32
1.126 brouard 11805: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11806: #endif
11807: if(!stat(plotcmd,&info)){
1.158 brouard 11808: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11809: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11810: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11811: }else
11812: strcpy(pplotcmd,plotcmd);
1.157 brouard 11813: #ifdef __unix
1.126 brouard 11814: strcpy(plotcmd,GNUPLOTPROGRAM);
11815: if(!stat(plotcmd,&info)){
1.158 brouard 11816: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11817: }else
11818: strcpy(pplotcmd,plotcmd);
11819: #endif
11820: }else
11821: strcpy(pplotcmd,plotcmd);
11822:
11823: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 11824: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 11825:
1.126 brouard 11826: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 11827: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 11828: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 11829: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 11830: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 11831: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 11832: }
1.158 brouard 11833: printf(" Successful, please wait...");
1.126 brouard 11834: while (z[0] != 'q') {
11835: /* chdir(path); */
1.154 brouard 11836: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 11837: scanf("%s",z);
11838: /* if (z[0] == 'c') system("./imach"); */
11839: if (z[0] == 'e') {
1.158 brouard 11840: #ifdef __APPLE__
1.152 brouard 11841: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 11842: #elif __linux
11843: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 11844: #else
1.152 brouard 11845: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 11846: #endif
11847: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
11848: system(pplotcmd);
1.126 brouard 11849: }
11850: else if (z[0] == 'g') system(plotcmd);
11851: else if (z[0] == 'q') exit(0);
11852: }
1.227 brouard 11853: end:
1.126 brouard 11854: while (z[0] != 'q') {
1.195 brouard 11855: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 11856: scanf("%s",z);
11857: }
11858: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/small/back.gif){kind=icon}
Return to imach.c CVS log ![[TXT]](/icons/small/text.gif){kind=icon}
![[DIR]](/icons/small/dir.gif){kind=icon}
Up to [Local Repository] / imach / src