Annotation of imach/src/imach.c, revision 1.257
1.257 ! brouard 1: /* $Id: imach.c,v 1.256 2017/03/27 05:50:23 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.257 ! brouard 4: Revision 1.256 2017/03/27 05:50:23 brouard
! 5: Summary: Temporary
! 6:
1.256 brouard 7: Revision 1.255 2017/03/08 16:02:28 brouard
8: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed
9:
1.255 brouard 10: Revision 1.254 2017/03/08 07:13:00 brouard
11: Summary: Fixing data parameter line
12:
1.254 brouard 13: Revision 1.253 2016/12/15 11:59:41 brouard
14: Summary: 0.99 in progress
15:
1.253 brouard 16: Revision 1.252 2016/09/15 21:15:37 brouard
17: *** empty log message ***
18:
1.252 brouard 19: Revision 1.251 2016/09/15 15:01:13 brouard
20: Summary: not working
21:
1.251 brouard 22: Revision 1.250 2016/09/08 16:07:27 brouard
23: Summary: continue
24:
1.250 brouard 25: Revision 1.249 2016/09/07 17:14:18 brouard
26: Summary: Starting values from frequencies
27:
1.249 brouard 28: Revision 1.248 2016/09/07 14:10:18 brouard
29: *** empty log message ***
30:
1.248 brouard 31: Revision 1.247 2016/09/02 11:11:21 brouard
32: *** empty log message ***
33:
1.247 brouard 34: Revision 1.246 2016/09/02 08:49:22 brouard
35: *** empty log message ***
36:
1.246 brouard 37: Revision 1.245 2016/09/02 07:25:01 brouard
38: *** empty log message ***
39:
1.245 brouard 40: Revision 1.244 2016/09/02 07:17:34 brouard
41: *** empty log message ***
42:
1.244 brouard 43: Revision 1.243 2016/09/02 06:45:35 brouard
44: *** empty log message ***
45:
1.243 brouard 46: Revision 1.242 2016/08/30 15:01:20 brouard
47: Summary: Fixing a lots
48:
1.242 brouard 49: Revision 1.241 2016/08/29 17:17:25 brouard
50: Summary: gnuplot problem in Back projection to fix
51:
1.241 brouard 52: Revision 1.240 2016/08/29 07:53:18 brouard
53: Summary: Better
54:
1.240 brouard 55: Revision 1.239 2016/08/26 15:51:03 brouard
56: Summary: Improvement in Powell output in order to copy and paste
57:
58: Author:
59:
1.239 brouard 60: Revision 1.238 2016/08/26 14:23:35 brouard
61: Summary: Starting tests of 0.99
62:
1.238 brouard 63: Revision 1.237 2016/08/26 09:20:19 brouard
64: Summary: to valgrind
65:
1.237 brouard 66: Revision 1.236 2016/08/25 10:50:18 brouard
67: *** empty log message ***
68:
1.236 brouard 69: Revision 1.235 2016/08/25 06:59:23 brouard
70: *** empty log message ***
71:
1.235 brouard 72: Revision 1.234 2016/08/23 16:51:20 brouard
73: *** empty log message ***
74:
1.234 brouard 75: Revision 1.233 2016/08/23 07:40:50 brouard
76: Summary: not working
77:
1.233 brouard 78: Revision 1.232 2016/08/22 14:20:21 brouard
79: Summary: not working
80:
1.232 brouard 81: Revision 1.231 2016/08/22 07:17:15 brouard
82: Summary: not working
83:
1.231 brouard 84: Revision 1.230 2016/08/22 06:55:53 brouard
85: Summary: Not working
86:
1.230 brouard 87: Revision 1.229 2016/07/23 09:45:53 brouard
88: Summary: Completing for func too
89:
1.229 brouard 90: Revision 1.228 2016/07/22 17:45:30 brouard
91: Summary: Fixing some arrays, still debugging
92:
1.227 brouard 93: Revision 1.226 2016/07/12 18:42:34 brouard
94: Summary: temp
95:
1.226 brouard 96: Revision 1.225 2016/07/12 08:40:03 brouard
97: Summary: saving but not running
98:
1.225 brouard 99: Revision 1.224 2016/07/01 13:16:01 brouard
100: Summary: Fixes
101:
1.224 brouard 102: Revision 1.223 2016/02/19 09:23:35 brouard
103: Summary: temporary
104:
1.223 brouard 105: Revision 1.222 2016/02/17 08:14:50 brouard
106: Summary: Probably last 0.98 stable version 0.98r6
107:
1.222 brouard 108: Revision 1.221 2016/02/15 23:35:36 brouard
109: Summary: minor bug
110:
1.220 brouard 111: Revision 1.219 2016/02/15 00:48:12 brouard
112: *** empty log message ***
113:
1.219 brouard 114: Revision 1.218 2016/02/12 11:29:23 brouard
115: Summary: 0.99 Back projections
116:
1.218 brouard 117: Revision 1.217 2015/12/23 17:18:31 brouard
118: Summary: Experimental backcast
119:
1.217 brouard 120: Revision 1.216 2015/12/18 17:32:11 brouard
121: Summary: 0.98r4 Warning and status=-2
122:
123: Version 0.98r4 is now:
124: - displaying an error when status is -1, date of interview unknown and date of death known;
125: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
126: Older changes concerning s=-2, dating from 2005 have been supersed.
127:
1.216 brouard 128: Revision 1.215 2015/12/16 08:52:24 brouard
129: Summary: 0.98r4 working
130:
1.215 brouard 131: Revision 1.214 2015/12/16 06:57:54 brouard
132: Summary: temporary not working
133:
1.214 brouard 134: Revision 1.213 2015/12/11 18:22:17 brouard
135: Summary: 0.98r4
136:
1.213 brouard 137: Revision 1.212 2015/11/21 12:47:24 brouard
138: Summary: minor typo
139:
1.212 brouard 140: Revision 1.211 2015/11/21 12:41:11 brouard
141: Summary: 0.98r3 with some graph of projected cross-sectional
142:
143: Author: Nicolas Brouard
144:
1.211 brouard 145: Revision 1.210 2015/11/18 17:41:20 brouard
1.252 brouard 146: Summary: Start working on projected prevalences Revision 1.209 2015/11/17 22:12:03 brouard
1.210 brouard 147: Summary: Adding ftolpl parameter
148: Author: N Brouard
149:
150: We had difficulties to get smoothed confidence intervals. It was due
151: to the period prevalence which wasn't computed accurately. The inner
152: parameter ftolpl is now an outer parameter of the .imach parameter
153: file after estepm. If ftolpl is small 1.e-4 and estepm too,
154: computation are long.
155:
1.209 brouard 156: Revision 1.208 2015/11/17 14:31:57 brouard
157: Summary: temporary
158:
1.208 brouard 159: Revision 1.207 2015/10/27 17:36:57 brouard
160: *** empty log message ***
161:
1.207 brouard 162: Revision 1.206 2015/10/24 07:14:11 brouard
163: *** empty log message ***
164:
1.206 brouard 165: Revision 1.205 2015/10/23 15:50:53 brouard
166: Summary: 0.98r3 some clarification for graphs on likelihood contributions
167:
1.205 brouard 168: Revision 1.204 2015/10/01 16:20:26 brouard
169: Summary: Some new graphs of contribution to likelihood
170:
1.204 brouard 171: Revision 1.203 2015/09/30 17:45:14 brouard
172: Summary: looking at better estimation of the hessian
173:
174: Also a better criteria for convergence to the period prevalence And
175: therefore adding the number of years needed to converge. (The
176: prevalence in any alive state shold sum to one
177:
1.203 brouard 178: Revision 1.202 2015/09/22 19:45:16 brouard
179: Summary: Adding some overall graph on contribution to likelihood. Might change
180:
1.202 brouard 181: Revision 1.201 2015/09/15 17:34:58 brouard
182: Summary: 0.98r0
183:
184: - Some new graphs like suvival functions
185: - Some bugs fixed like model=1+age+V2.
186:
1.201 brouard 187: Revision 1.200 2015/09/09 16:53:55 brouard
188: Summary: Big bug thanks to Flavia
189:
190: Even model=1+age+V2. did not work anymore
191:
1.200 brouard 192: Revision 1.199 2015/09/07 14:09:23 brouard
193: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
194:
1.199 brouard 195: Revision 1.198 2015/09/03 07:14:39 brouard
196: Summary: 0.98q5 Flavia
197:
1.198 brouard 198: Revision 1.197 2015/09/01 18:24:39 brouard
199: *** empty log message ***
200:
1.197 brouard 201: Revision 1.196 2015/08/18 23:17:52 brouard
202: Summary: 0.98q5
203:
1.196 brouard 204: Revision 1.195 2015/08/18 16:28:39 brouard
205: Summary: Adding a hack for testing purpose
206:
207: After reading the title, ftol and model lines, if the comment line has
208: a q, starting with #q, the answer at the end of the run is quit. It
209: permits to run test files in batch with ctest. The former workaround was
210: $ echo q | imach foo.imach
211:
1.195 brouard 212: Revision 1.194 2015/08/18 13:32:00 brouard
213: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
214:
1.194 brouard 215: Revision 1.193 2015/08/04 07:17:42 brouard
216: Summary: 0.98q4
217:
1.193 brouard 218: Revision 1.192 2015/07/16 16:49:02 brouard
219: Summary: Fixing some outputs
220:
1.192 brouard 221: Revision 1.191 2015/07/14 10:00:33 brouard
222: Summary: Some fixes
223:
1.191 brouard 224: Revision 1.190 2015/05/05 08:51:13 brouard
225: Summary: Adding digits in output parameters (7 digits instead of 6)
226:
227: Fix 1+age+.
228:
1.190 brouard 229: Revision 1.189 2015/04/30 14:45:16 brouard
230: Summary: 0.98q2
231:
1.189 brouard 232: Revision 1.188 2015/04/30 08:27:53 brouard
233: *** empty log message ***
234:
1.188 brouard 235: Revision 1.187 2015/04/29 09:11:15 brouard
236: *** empty log message ***
237:
1.187 brouard 238: Revision 1.186 2015/04/23 12:01:52 brouard
239: Summary: V1*age is working now, version 0.98q1
240:
241: Some codes had been disabled in order to simplify and Vn*age was
242: working in the optimization phase, ie, giving correct MLE parameters,
243: but, as usual, outputs were not correct and program core dumped.
244:
1.186 brouard 245: Revision 1.185 2015/03/11 13:26:42 brouard
246: Summary: Inclusion of compile and links command line for Intel Compiler
247:
1.185 brouard 248: Revision 1.184 2015/03/11 11:52:39 brouard
249: Summary: Back from Windows 8. Intel Compiler
250:
1.184 brouard 251: Revision 1.183 2015/03/10 20:34:32 brouard
252: Summary: 0.98q0, trying with directest, mnbrak fixed
253:
254: We use directest instead of original Powell test; probably no
255: incidence on the results, but better justifications;
256: We fixed Numerical Recipes mnbrak routine which was wrong and gave
257: wrong results.
258:
1.183 brouard 259: Revision 1.182 2015/02/12 08:19:57 brouard
260: Summary: Trying to keep directest which seems simpler and more general
261: Author: Nicolas Brouard
262:
1.182 brouard 263: Revision 1.181 2015/02/11 23:22:24 brouard
264: Summary: Comments on Powell added
265:
266: Author:
267:
1.181 brouard 268: Revision 1.180 2015/02/11 17:33:45 brouard
269: Summary: Finishing move from main to function (hpijx and prevalence_limit)
270:
1.180 brouard 271: Revision 1.179 2015/01/04 09:57:06 brouard
272: Summary: back to OS/X
273:
1.179 brouard 274: Revision 1.178 2015/01/04 09:35:48 brouard
275: *** empty log message ***
276:
1.178 brouard 277: Revision 1.177 2015/01/03 18:40:56 brouard
278: Summary: Still testing ilc32 on OSX
279:
1.177 brouard 280: Revision 1.176 2015/01/03 16:45:04 brouard
281: *** empty log message ***
282:
1.176 brouard 283: Revision 1.175 2015/01/03 16:33:42 brouard
284: *** empty log message ***
285:
1.175 brouard 286: Revision 1.174 2015/01/03 16:15:49 brouard
287: Summary: Still in cross-compilation
288:
1.174 brouard 289: Revision 1.173 2015/01/03 12:06:26 brouard
290: Summary: trying to detect cross-compilation
291:
1.173 brouard 292: Revision 1.172 2014/12/27 12:07:47 brouard
293: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
294:
1.172 brouard 295: Revision 1.171 2014/12/23 13:26:59 brouard
296: Summary: Back from Visual C
297:
298: Still problem with utsname.h on Windows
299:
1.171 brouard 300: Revision 1.170 2014/12/23 11:17:12 brouard
301: Summary: Cleaning some \%% back to %%
302:
303: The escape was mandatory for a specific compiler (which one?), but too many warnings.
304:
1.170 brouard 305: Revision 1.169 2014/12/22 23:08:31 brouard
306: Summary: 0.98p
307:
308: Outputs some informations on compiler used, OS etc. Testing on different platforms.
309:
1.169 brouard 310: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 311: Summary: update
1.169 brouard 312:
1.168 brouard 313: Revision 1.167 2014/12/22 13:50:56 brouard
314: Summary: Testing uname and compiler version and if compiled 32 or 64
315:
316: Testing on Linux 64
317:
1.167 brouard 318: Revision 1.166 2014/12/22 11:40:47 brouard
319: *** empty log message ***
320:
1.166 brouard 321: Revision 1.165 2014/12/16 11:20:36 brouard
322: Summary: After compiling on Visual C
323:
324: * imach.c (Module): Merging 1.61 to 1.162
325:
1.165 brouard 326: Revision 1.164 2014/12/16 10:52:11 brouard
327: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
328:
329: * imach.c (Module): Merging 1.61 to 1.162
330:
1.164 brouard 331: Revision 1.163 2014/12/16 10:30:11 brouard
332: * imach.c (Module): Merging 1.61 to 1.162
333:
1.163 brouard 334: Revision 1.162 2014/09/25 11:43:39 brouard
335: Summary: temporary backup 0.99!
336:
1.162 brouard 337: Revision 1.1 2014/09/16 11:06:58 brouard
338: Summary: With some code (wrong) for nlopt
339:
340: Author:
341:
342: Revision 1.161 2014/09/15 20:41:41 brouard
343: Summary: Problem with macro SQR on Intel compiler
344:
1.161 brouard 345: Revision 1.160 2014/09/02 09:24:05 brouard
346: *** empty log message ***
347:
1.160 brouard 348: Revision 1.159 2014/09/01 10:34:10 brouard
349: Summary: WIN32
350: Author: Brouard
351:
1.159 brouard 352: Revision 1.158 2014/08/27 17:11:51 brouard
353: *** empty log message ***
354:
1.158 brouard 355: Revision 1.157 2014/08/27 16:26:55 brouard
356: Summary: Preparing windows Visual studio version
357: Author: Brouard
358:
359: In order to compile on Visual studio, time.h is now correct and time_t
360: and tm struct should be used. difftime should be used but sometimes I
361: just make the differences in raw time format (time(&now).
362: Trying to suppress #ifdef LINUX
363: Add xdg-open for __linux in order to open default browser.
364:
1.157 brouard 365: Revision 1.156 2014/08/25 20:10:10 brouard
366: *** empty log message ***
367:
1.156 brouard 368: Revision 1.155 2014/08/25 18:32:34 brouard
369: Summary: New compile, minor changes
370: Author: Brouard
371:
1.155 brouard 372: Revision 1.154 2014/06/20 17:32:08 brouard
373: Summary: Outputs now all graphs of convergence to period prevalence
374:
1.154 brouard 375: Revision 1.153 2014/06/20 16:45:46 brouard
376: Summary: If 3 live state, convergence to period prevalence on same graph
377: Author: Brouard
378:
1.153 brouard 379: Revision 1.152 2014/06/18 17:54:09 brouard
380: Summary: open browser, use gnuplot on same dir than imach if not found in the path
381:
1.152 brouard 382: Revision 1.151 2014/06/18 16:43:30 brouard
383: *** empty log message ***
384:
1.151 brouard 385: Revision 1.150 2014/06/18 16:42:35 brouard
386: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
387: Author: brouard
388:
1.150 brouard 389: Revision 1.149 2014/06/18 15:51:14 brouard
390: Summary: Some fixes in parameter files errors
391: Author: Nicolas Brouard
392:
1.149 brouard 393: Revision 1.148 2014/06/17 17:38:48 brouard
394: Summary: Nothing new
395: Author: Brouard
396:
397: Just a new packaging for OS/X version 0.98nS
398:
1.148 brouard 399: Revision 1.147 2014/06/16 10:33:11 brouard
400: *** empty log message ***
401:
1.147 brouard 402: Revision 1.146 2014/06/16 10:20:28 brouard
403: Summary: Merge
404: Author: Brouard
405:
406: Merge, before building revised version.
407:
1.146 brouard 408: Revision 1.145 2014/06/10 21:23:15 brouard
409: Summary: Debugging with valgrind
410: Author: Nicolas Brouard
411:
412: Lot of changes in order to output the results with some covariates
413: After the Edimburgh REVES conference 2014, it seems mandatory to
414: improve the code.
415: No more memory valgrind error but a lot has to be done in order to
416: continue the work of splitting the code into subroutines.
417: Also, decodemodel has been improved. Tricode is still not
418: optimal. nbcode should be improved. Documentation has been added in
419: the source code.
420:
1.144 brouard 421: Revision 1.143 2014/01/26 09:45:38 brouard
422: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
423:
424: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
425: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
426:
1.143 brouard 427: Revision 1.142 2014/01/26 03:57:36 brouard
428: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
429:
430: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
431:
1.142 brouard 432: Revision 1.141 2014/01/26 02:42:01 brouard
433: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
434:
1.141 brouard 435: Revision 1.140 2011/09/02 10:37:54 brouard
436: Summary: times.h is ok with mingw32 now.
437:
1.140 brouard 438: Revision 1.139 2010/06/14 07:50:17 brouard
439: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
440: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
441:
1.139 brouard 442: Revision 1.138 2010/04/30 18:19:40 brouard
443: *** empty log message ***
444:
1.138 brouard 445: Revision 1.137 2010/04/29 18:11:38 brouard
446: (Module): Checking covariates for more complex models
447: than V1+V2. A lot of change to be done. Unstable.
448:
1.137 brouard 449: Revision 1.136 2010/04/26 20:30:53 brouard
450: (Module): merging some libgsl code. Fixing computation
451: of likelione (using inter/intrapolation if mle = 0) in order to
452: get same likelihood as if mle=1.
453: Some cleaning of code and comments added.
454:
1.136 brouard 455: Revision 1.135 2009/10/29 15:33:14 brouard
456: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
457:
1.135 brouard 458: Revision 1.134 2009/10/29 13:18:53 brouard
459: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
460:
1.134 brouard 461: Revision 1.133 2009/07/06 10:21:25 brouard
462: just nforces
463:
1.133 brouard 464: Revision 1.132 2009/07/06 08:22:05 brouard
465: Many tings
466:
1.132 brouard 467: Revision 1.131 2009/06/20 16:22:47 brouard
468: Some dimensions resccaled
469:
1.131 brouard 470: Revision 1.130 2009/05/26 06:44:34 brouard
471: (Module): Max Covariate is now set to 20 instead of 8. A
472: lot of cleaning with variables initialized to 0. Trying to make
473: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
474:
1.130 brouard 475: Revision 1.129 2007/08/31 13:49:27 lievre
476: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
477:
1.129 lievre 478: Revision 1.128 2006/06/30 13:02:05 brouard
479: (Module): Clarifications on computing e.j
480:
1.128 brouard 481: Revision 1.127 2006/04/28 18:11:50 brouard
482: (Module): Yes the sum of survivors was wrong since
483: imach-114 because nhstepm was no more computed in the age
484: loop. Now we define nhstepma in the age loop.
485: (Module): In order to speed up (in case of numerous covariates) we
486: compute health expectancies (without variances) in a first step
487: and then all the health expectancies with variances or standard
488: deviation (needs data from the Hessian matrices) which slows the
489: computation.
490: In the future we should be able to stop the program is only health
491: expectancies and graph are needed without standard deviations.
492:
1.127 brouard 493: Revision 1.126 2006/04/28 17:23:28 brouard
494: (Module): Yes the sum of survivors was wrong since
495: imach-114 because nhstepm was no more computed in the age
496: loop. Now we define nhstepma in the age loop.
497: Version 0.98h
498:
1.126 brouard 499: Revision 1.125 2006/04/04 15:20:31 lievre
500: Errors in calculation of health expectancies. Age was not initialized.
501: Forecasting file added.
502:
503: Revision 1.124 2006/03/22 17:13:53 lievre
504: Parameters are printed with %lf instead of %f (more numbers after the comma).
505: The log-likelihood is printed in the log file
506:
507: Revision 1.123 2006/03/20 10:52:43 brouard
508: * imach.c (Module): <title> changed, corresponds to .htm file
509: name. <head> headers where missing.
510:
511: * imach.c (Module): Weights can have a decimal point as for
512: English (a comma might work with a correct LC_NUMERIC environment,
513: otherwise the weight is truncated).
514: Modification of warning when the covariates values are not 0 or
515: 1.
516: Version 0.98g
517:
518: Revision 1.122 2006/03/20 09:45:41 brouard
519: (Module): Weights can have a decimal point as for
520: English (a comma might work with a correct LC_NUMERIC environment,
521: otherwise the weight is truncated).
522: Modification of warning when the covariates values are not 0 or
523: 1.
524: Version 0.98g
525:
526: Revision 1.121 2006/03/16 17:45:01 lievre
527: * imach.c (Module): Comments concerning covariates added
528:
529: * imach.c (Module): refinements in the computation of lli if
530: status=-2 in order to have more reliable computation if stepm is
531: not 1 month. Version 0.98f
532:
533: Revision 1.120 2006/03/16 15:10:38 lievre
534: (Module): refinements in the computation of lli if
535: status=-2 in order to have more reliable computation if stepm is
536: not 1 month. Version 0.98f
537:
538: Revision 1.119 2006/03/15 17:42:26 brouard
539: (Module): Bug if status = -2, the loglikelihood was
540: computed as likelihood omitting the logarithm. Version O.98e
541:
542: Revision 1.118 2006/03/14 18:20:07 brouard
543: (Module): varevsij Comments added explaining the second
544: table of variances if popbased=1 .
545: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
546: (Module): Function pstamp added
547: (Module): Version 0.98d
548:
549: Revision 1.117 2006/03/14 17:16:22 brouard
550: (Module): varevsij Comments added explaining the second
551: table of variances if popbased=1 .
552: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
553: (Module): Function pstamp added
554: (Module): Version 0.98d
555:
556: Revision 1.116 2006/03/06 10:29:27 brouard
557: (Module): Variance-covariance wrong links and
558: varian-covariance of ej. is needed (Saito).
559:
560: Revision 1.115 2006/02/27 12:17:45 brouard
561: (Module): One freematrix added in mlikeli! 0.98c
562:
563: Revision 1.114 2006/02/26 12:57:58 brouard
564: (Module): Some improvements in processing parameter
565: filename with strsep.
566:
567: Revision 1.113 2006/02/24 14:20:24 brouard
568: (Module): Memory leaks checks with valgrind and:
569: datafile was not closed, some imatrix were not freed and on matrix
570: allocation too.
571:
572: Revision 1.112 2006/01/30 09:55:26 brouard
573: (Module): Back to gnuplot.exe instead of wgnuplot.exe
574:
575: Revision 1.111 2006/01/25 20:38:18 brouard
576: (Module): Lots of cleaning and bugs added (Gompertz)
577: (Module): Comments can be added in data file. Missing date values
578: can be a simple dot '.'.
579:
580: Revision 1.110 2006/01/25 00:51:50 brouard
581: (Module): Lots of cleaning and bugs added (Gompertz)
582:
583: Revision 1.109 2006/01/24 19:37:15 brouard
584: (Module): Comments (lines starting with a #) are allowed in data.
585:
586: Revision 1.108 2006/01/19 18:05:42 lievre
587: Gnuplot problem appeared...
588: To be fixed
589:
590: Revision 1.107 2006/01/19 16:20:37 brouard
591: Test existence of gnuplot in imach path
592:
593: Revision 1.106 2006/01/19 13:24:36 brouard
594: Some cleaning and links added in html output
595:
596: Revision 1.105 2006/01/05 20:23:19 lievre
597: *** empty log message ***
598:
599: Revision 1.104 2005/09/30 16:11:43 lievre
600: (Module): sump fixed, loop imx fixed, and simplifications.
601: (Module): If the status is missing at the last wave but we know
602: that the person is alive, then we can code his/her status as -2
603: (instead of missing=-1 in earlier versions) and his/her
604: contributions to the likelihood is 1 - Prob of dying from last
605: health status (= 1-p13= p11+p12 in the easiest case of somebody in
606: the healthy state at last known wave). Version is 0.98
607:
608: Revision 1.103 2005/09/30 15:54:49 lievre
609: (Module): sump fixed, loop imx fixed, and simplifications.
610:
611: Revision 1.102 2004/09/15 17:31:30 brouard
612: Add the possibility to read data file including tab characters.
613:
614: Revision 1.101 2004/09/15 10:38:38 brouard
615: Fix on curr_time
616:
617: Revision 1.100 2004/07/12 18:29:06 brouard
618: Add version for Mac OS X. Just define UNIX in Makefile
619:
620: Revision 1.99 2004/06/05 08:57:40 brouard
621: *** empty log message ***
622:
623: Revision 1.98 2004/05/16 15:05:56 brouard
624: New version 0.97 . First attempt to estimate force of mortality
625: directly from the data i.e. without the need of knowing the health
626: state at each age, but using a Gompertz model: log u =a + b*age .
627: This is the basic analysis of mortality and should be done before any
628: other analysis, in order to test if the mortality estimated from the
629: cross-longitudinal survey is different from the mortality estimated
630: from other sources like vital statistic data.
631:
632: The same imach parameter file can be used but the option for mle should be -3.
633:
1.133 brouard 634: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 635: former routines in order to include the new code within the former code.
636:
637: The output is very simple: only an estimate of the intercept and of
638: the slope with 95% confident intervals.
639:
640: Current limitations:
641: A) Even if you enter covariates, i.e. with the
642: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
643: B) There is no computation of Life Expectancy nor Life Table.
644:
645: Revision 1.97 2004/02/20 13:25:42 lievre
646: Version 0.96d. Population forecasting command line is (temporarily)
647: suppressed.
648:
649: Revision 1.96 2003/07/15 15:38:55 brouard
650: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
651: rewritten within the same printf. Workaround: many printfs.
652:
653: Revision 1.95 2003/07/08 07:54:34 brouard
654: * imach.c (Repository):
655: (Repository): Using imachwizard code to output a more meaningful covariance
656: matrix (cov(a12,c31) instead of numbers.
657:
658: Revision 1.94 2003/06/27 13:00:02 brouard
659: Just cleaning
660:
661: Revision 1.93 2003/06/25 16:33:55 brouard
662: (Module): On windows (cygwin) function asctime_r doesn't
663: exist so I changed back to asctime which exists.
664: (Module): Version 0.96b
665:
666: Revision 1.92 2003/06/25 16:30:45 brouard
667: (Module): On windows (cygwin) function asctime_r doesn't
668: exist so I changed back to asctime which exists.
669:
670: Revision 1.91 2003/06/25 15:30:29 brouard
671: * imach.c (Repository): Duplicated warning errors corrected.
672: (Repository): Elapsed time after each iteration is now output. It
673: helps to forecast when convergence will be reached. Elapsed time
674: is stamped in powell. We created a new html file for the graphs
675: concerning matrix of covariance. It has extension -cov.htm.
676:
677: Revision 1.90 2003/06/24 12:34:15 brouard
678: (Module): Some bugs corrected for windows. Also, when
679: mle=-1 a template is output in file "or"mypar.txt with the design
680: of the covariance matrix to be input.
681:
682: Revision 1.89 2003/06/24 12:30:52 brouard
683: (Module): Some bugs corrected for windows. Also, when
684: mle=-1 a template is output in file "or"mypar.txt with the design
685: of the covariance matrix to be input.
686:
687: Revision 1.88 2003/06/23 17:54:56 brouard
688: * 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.
689:
690: Revision 1.87 2003/06/18 12:26:01 brouard
691: Version 0.96
692:
693: Revision 1.86 2003/06/17 20:04:08 brouard
694: (Module): Change position of html and gnuplot routines and added
695: routine fileappend.
696:
697: Revision 1.85 2003/06/17 13:12:43 brouard
698: * imach.c (Repository): Check when date of death was earlier that
699: current date of interview. It may happen when the death was just
700: prior to the death. In this case, dh was negative and likelihood
701: was wrong (infinity). We still send an "Error" but patch by
702: assuming that the date of death was just one stepm after the
703: interview.
704: (Repository): Because some people have very long ID (first column)
705: we changed int to long in num[] and we added a new lvector for
706: memory allocation. But we also truncated to 8 characters (left
707: truncation)
708: (Repository): No more line truncation errors.
709:
710: Revision 1.84 2003/06/13 21:44:43 brouard
711: * imach.c (Repository): Replace "freqsummary" at a correct
712: place. It differs from routine "prevalence" which may be called
713: many times. Probs is memory consuming and must be used with
714: parcimony.
715: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
716:
717: Revision 1.83 2003/06/10 13:39:11 lievre
718: *** empty log message ***
719:
720: Revision 1.82 2003/06/05 15:57:20 brouard
721: Add log in imach.c and fullversion number is now printed.
722:
723: */
724: /*
725: Interpolated Markov Chain
726:
727: Short summary of the programme:
728:
1.227 brouard 729: This program computes Healthy Life Expectancies or State-specific
730: (if states aren't health statuses) Expectancies from
731: cross-longitudinal data. Cross-longitudinal data consist in:
732:
733: -1- a first survey ("cross") where individuals from different ages
734: are interviewed on their health status or degree of disability (in
735: the case of a health survey which is our main interest)
736:
737: -2- at least a second wave of interviews ("longitudinal") which
738: measure each change (if any) in individual health status. Health
739: expectancies are computed from the time spent in each health state
740: according to a model. More health states you consider, more time is
741: necessary to reach the Maximum Likelihood of the parameters involved
742: in the model. The simplest model is the multinomial logistic model
743: where pij is the probability to be observed in state j at the second
744: wave conditional to be observed in state i at the first
745: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
746: etc , where 'age' is age and 'sex' is a covariate. If you want to
747: have a more complex model than "constant and age", you should modify
748: the program where the markup *Covariates have to be included here
749: again* invites you to do it. More covariates you add, slower the
1.126 brouard 750: convergence.
751:
752: The advantage of this computer programme, compared to a simple
753: multinomial logistic model, is clear when the delay between waves is not
754: identical for each individual. Also, if a individual missed an
755: intermediate interview, the information is lost, but taken into
756: account using an interpolation or extrapolation.
757:
758: hPijx is the probability to be observed in state i at age x+h
759: conditional to the observed state i at age x. The delay 'h' can be
760: split into an exact number (nh*stepm) of unobserved intermediate
761: states. This elementary transition (by month, quarter,
762: semester or year) is modelled as a multinomial logistic. The hPx
763: matrix is simply the matrix product of nh*stepm elementary matrices
764: and the contribution of each individual to the likelihood is simply
765: hPijx.
766:
767: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 768: of the life expectancies. It also computes the period (stable) prevalence.
769:
770: Back prevalence and projections:
1.227 brouard 771:
772: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
773: double agemaxpar, double ftolpl, int *ncvyearp, double
774: dateprev1,double dateprev2, int firstpass, int lastpass, int
775: mobilavproj)
776:
777: Computes the back prevalence limit for any combination of
778: covariate values k at any age between ageminpar and agemaxpar and
779: returns it in **bprlim. In the loops,
780:
781: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
782: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
783:
784: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 785: Computes for any combination of covariates k and any age between bage and fage
786: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
787: oldm=oldms;savm=savms;
1.227 brouard 788:
789: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 790: Computes the transition matrix starting at age 'age' over
791: 'nhstepm*hstepm*stepm' months (i.e. until
792: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 793: nhstepm*hstepm matrices.
794:
795: Returns p3mat[i][j][h] after calling
796: p3mat[i][j][h]=matprod2(newm,
797: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
798: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
799: oldm);
1.226 brouard 800:
801: Important routines
802:
803: - func (or funcone), computes logit (pij) distinguishing
804: o fixed variables (single or product dummies or quantitative);
805: o varying variables by:
806: (1) wave (single, product dummies, quantitative),
807: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
808: % fixed dummy (treated) or quantitative (not done because time-consuming);
809: % varying dummy (not done) or quantitative (not done);
810: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
811: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
812: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
813: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
814: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 815:
1.226 brouard 816:
817:
1.133 brouard 818: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
819: Institut national d'études démographiques, Paris.
1.126 brouard 820: This software have been partly granted by Euro-REVES, a concerted action
821: from the European Union.
822: It is copyrighted identically to a GNU software product, ie programme and
823: software can be distributed freely for non commercial use. Latest version
824: can be accessed at http://euroreves.ined.fr/imach .
825:
826: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
827: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
828:
829: **********************************************************************/
830: /*
831: main
832: read parameterfile
833: read datafile
834: concatwav
835: freqsummary
836: if (mle >= 1)
837: mlikeli
838: print results files
839: if mle==1
840: computes hessian
841: read end of parameter file: agemin, agemax, bage, fage, estepm
842: begin-prev-date,...
843: open gnuplot file
844: open html file
1.145 brouard 845: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
846: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
847: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
848: freexexit2 possible for memory heap.
849:
850: h Pij x | pij_nom ficrestpij
851: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
852: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
853: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
854:
855: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
856: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
857: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
858: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
859: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
860:
1.126 brouard 861: forecasting if prevfcast==1 prevforecast call prevalence()
862: health expectancies
863: Variance-covariance of DFLE
864: prevalence()
865: movingaverage()
866: varevsij()
867: if popbased==1 varevsij(,popbased)
868: total life expectancies
869: Variance of period (stable) prevalence
870: end
871: */
872:
1.187 brouard 873: /* #define DEBUG */
874: /* #define DEBUGBRENT */
1.203 brouard 875: /* #define DEBUGLINMIN */
876: /* #define DEBUGHESS */
877: #define DEBUGHESSIJ
1.224 brouard 878: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 879: #define POWELL /* Instead of NLOPT */
1.224 brouard 880: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 881: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
882: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 883:
884: #include <math.h>
885: #include <stdio.h>
886: #include <stdlib.h>
887: #include <string.h>
1.226 brouard 888: #include <ctype.h>
1.159 brouard 889:
890: #ifdef _WIN32
891: #include <io.h>
1.172 brouard 892: #include <windows.h>
893: #include <tchar.h>
1.159 brouard 894: #else
1.126 brouard 895: #include <unistd.h>
1.159 brouard 896: #endif
1.126 brouard 897:
898: #include <limits.h>
899: #include <sys/types.h>
1.171 brouard 900:
901: #if defined(__GNUC__)
902: #include <sys/utsname.h> /* Doesn't work on Windows */
903: #endif
904:
1.126 brouard 905: #include <sys/stat.h>
906: #include <errno.h>
1.159 brouard 907: /* extern int errno; */
1.126 brouard 908:
1.157 brouard 909: /* #ifdef LINUX */
910: /* #include <time.h> */
911: /* #include "timeval.h" */
912: /* #else */
913: /* #include <sys/time.h> */
914: /* #endif */
915:
1.126 brouard 916: #include <time.h>
917:
1.136 brouard 918: #ifdef GSL
919: #include <gsl/gsl_errno.h>
920: #include <gsl/gsl_multimin.h>
921: #endif
922:
1.167 brouard 923:
1.162 brouard 924: #ifdef NLOPT
925: #include <nlopt.h>
926: typedef struct {
927: double (* function)(double [] );
928: } myfunc_data ;
929: #endif
930:
1.126 brouard 931: /* #include <libintl.h> */
932: /* #define _(String) gettext (String) */
933:
1.251 brouard 934: #define MAXLINE 2048 /* Was 256 and 1024. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 935:
936: #define GNUPLOTPROGRAM "gnuplot"
937: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
938: #define FILENAMELENGTH 132
939:
940: #define GLOCK_ERROR_NOPATH -1 /* empty path */
941: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
942:
1.144 brouard 943: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
944: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 945:
946: #define NINTERVMAX 8
1.144 brouard 947: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
948: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
949: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 950: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 951: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
952: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 953: #define MAXN 20000
1.144 brouard 954: #define YEARM 12. /**< Number of months per year */
1.218 brouard 955: /* #define AGESUP 130 */
956: #define AGESUP 150
957: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 958: #define AGEBASE 40
1.194 brouard 959: #define AGEOVERFLOW 1.e20
1.164 brouard 960: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 961: #ifdef _WIN32
962: #define DIRSEPARATOR '\\'
963: #define CHARSEPARATOR "\\"
964: #define ODIRSEPARATOR '/'
965: #else
1.126 brouard 966: #define DIRSEPARATOR '/'
967: #define CHARSEPARATOR "/"
968: #define ODIRSEPARATOR '\\'
969: #endif
970:
1.257 ! brouard 971: /* $Id: imach.c,v 1.256 2017/03/27 05:50:23 brouard Exp $ */
1.126 brouard 972: /* $State: Exp $ */
1.196 brouard 973: #include "version.h"
974: char version[]=__IMACH_VERSION__;
1.224 brouard 975: 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";
1.257 ! brouard 976: char fullversion[]="$Revision: 1.256 $ $Date: 2017/03/27 05:50:23 $";
1.126 brouard 977: char strstart[80];
978: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 979: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 980: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 981: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
982: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
983: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 984: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
985: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 986: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
987: int cptcovprodnoage=0; /**< Number of covariate products without age */
988: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 989: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
990: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 991: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 992: int nsd=0; /**< Total number of single dummy variables (output) */
993: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 994: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 995: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 996: int ntveff=0; /**< ntveff number of effective time varying variables */
997: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 998: int cptcov=0; /* Working variable */
1.218 brouard 999: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 1000: int npar=NPARMAX;
1001: int nlstate=2; /* Number of live states */
1002: int ndeath=1; /* Number of dead states */
1.130 brouard 1003: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 1004: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 1005: int popbased=0;
1006:
1007: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 1008: int maxwav=0; /* Maxim number of waves */
1009: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
1010: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
1011: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 1012: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 1013: int mle=1, weightopt=0;
1.126 brouard 1014: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
1015: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
1016: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
1017: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 1018: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 1019: int selected(int kvar); /* Is covariate kvar selected for printing results */
1020:
1.130 brouard 1021: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 1022: double **matprod2(); /* test */
1.126 brouard 1023: double **oldm, **newm, **savm; /* Working pointers to matrices */
1024: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 1025: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
1026:
1.136 brouard 1027: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 1028: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 1029: FILE *ficlog, *ficrespow;
1.130 brouard 1030: int globpr=0; /* Global variable for printing or not */
1.126 brouard 1031: double fretone; /* Only one call to likelihood */
1.130 brouard 1032: long ipmx=0; /* Number of contributions */
1.126 brouard 1033: double sw; /* Sum of weights */
1034: char filerespow[FILENAMELENGTH];
1035: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
1036: FILE *ficresilk;
1037: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
1038: FILE *ficresprobmorprev;
1039: FILE *fichtm, *fichtmcov; /* Html File */
1040: FILE *ficreseij;
1041: char filerese[FILENAMELENGTH];
1042: FILE *ficresstdeij;
1043: char fileresstde[FILENAMELENGTH];
1044: FILE *ficrescveij;
1045: char filerescve[FILENAMELENGTH];
1046: FILE *ficresvij;
1047: char fileresv[FILENAMELENGTH];
1048: FILE *ficresvpl;
1049: char fileresvpl[FILENAMELENGTH];
1050: char title[MAXLINE];
1.234 brouard 1051: char model[MAXLINE]; /**< The model line */
1.217 brouard 1052: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 1053: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
1054: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
1055: char command[FILENAMELENGTH];
1056: int outcmd=0;
1057:
1.217 brouard 1058: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 1059: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 1060: char filelog[FILENAMELENGTH]; /* Log file */
1061: char filerest[FILENAMELENGTH];
1062: char fileregp[FILENAMELENGTH];
1063: char popfile[FILENAMELENGTH];
1064:
1065: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1066:
1.157 brouard 1067: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1068: /* struct timezone tzp; */
1069: /* extern int gettimeofday(); */
1070: struct tm tml, *gmtime(), *localtime();
1071:
1072: extern time_t time();
1073:
1074: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1075: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1076: struct tm tm;
1077:
1.126 brouard 1078: char strcurr[80], strfor[80];
1079:
1080: char *endptr;
1081: long lval;
1082: double dval;
1083:
1084: #define NR_END 1
1085: #define FREE_ARG char*
1086: #define FTOL 1.0e-10
1087:
1088: #define NRANSI
1.240 brouard 1089: #define ITMAX 200
1090: #define ITPOWMAX 20 /* This is now multiplied by the number of parameters */
1.126 brouard 1091:
1092: #define TOL 2.0e-4
1093:
1094: #define CGOLD 0.3819660
1095: #define ZEPS 1.0e-10
1096: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1097:
1098: #define GOLD 1.618034
1099: #define GLIMIT 100.0
1100: #define TINY 1.0e-20
1101:
1102: static double maxarg1,maxarg2;
1103: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1104: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1105:
1106: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1107: #define rint(a) floor(a+0.5)
1.166 brouard 1108: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1109: #define mytinydouble 1.0e-16
1.166 brouard 1110: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1111: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1112: /* static double dsqrarg; */
1113: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1114: static double sqrarg;
1115: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1116: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1117: int agegomp= AGEGOMP;
1118:
1119: int imx;
1120: int stepm=1;
1121: /* Stepm, step in month: minimum step interpolation*/
1122:
1123: int estepm;
1124: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1125:
1126: int m,nb;
1127: long *num;
1.197 brouard 1128: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1129: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1130: covariate for which somebody answered excluding
1131: undefined. Usually 2: 0 and 1. */
1132: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1133: covariate for which somebody answered including
1134: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1135: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1136: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1137: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1138: double *ageexmed,*agecens;
1139: double dateintmean=0;
1140:
1141: double *weight;
1142: int **s; /* Status */
1.141 brouard 1143: double *agedc;
1.145 brouard 1144: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1145: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1146: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1147: double **coqvar; /* Fixed quantitative covariate iqv */
1148: double ***cotvar; /* Time varying covariate itv */
1149: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1150: double idx;
1151: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1152: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1153: /*k 1 2 3 4 5 6 7 8 9 */
1154: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1155: /* Tndvar[k] 1 2 3 4 5 */
1156: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1157: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1158: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1159: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1160: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1161: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1162: /* Tprod[i]=k 4 7 */
1163: /* Tage[i]=k 5 8 */
1164: /* */
1165: /* Type */
1166: /* V 1 2 3 4 5 */
1167: /* F F V V V */
1168: /* D Q D D Q */
1169: /* */
1170: int *TvarsD;
1171: int *TvarsDind;
1172: int *TvarsQ;
1173: int *TvarsQind;
1174:
1.235 brouard 1175: #define MAXRESULTLINES 10
1176: int nresult=0;
1177: int TKresult[MAXRESULTLINES];
1.237 brouard 1178: int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1179: int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1.235 brouard 1180: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1181: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.237 brouard 1182: double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.235 brouard 1183: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1184:
1.234 brouard 1185: /* 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 1186: 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 */
1187: 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 */
1188: 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 */
1189: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1190: 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 */
1191: 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 1192: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1193: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1194: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1195: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1196: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1197: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1198: 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 */
1199: 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 */
1200:
1.230 brouard 1201: int *Tvarsel; /**< Selected covariates for output */
1202: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1203: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1204: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1205: 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 1206: int *DummyV; /** Dummy[v] 0=dummy (0 1), 1 quantitative */
1207: int *FixedV; /** FixedV[v] 0 fixed, 1 varying */
1.197 brouard 1208: int *Tage;
1.227 brouard 1209: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1210: 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 1211: 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*/
1212: 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 1213: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1214: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1215: int **Tvard;
1216: int *Tprod;/**< Gives the k position of the k1 product */
1.238 brouard 1217: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3 */
1.227 brouard 1218: int *Tposprod; /**< Gives the k1 product from the k position */
1.238 brouard 1219: /* if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2) */
1220: /* Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5(V3*V2)]=2 (2nd product without age) */
1.227 brouard 1221: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1222: double *lsurv, *lpop, *tpop;
1223:
1.231 brouard 1224: #define FD 1; /* Fixed dummy covariate */
1225: #define FQ 2; /* Fixed quantitative covariate */
1226: #define FP 3; /* Fixed product covariate */
1227: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1228: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1229: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1230: #define VD 10; /* Varying dummy covariate */
1231: #define VQ 11; /* Varying quantitative covariate */
1232: #define VP 12; /* Varying product covariate */
1233: #define VPDD 13; /* Varying product dummy*dummy covariate */
1234: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1235: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1236: #define APFD 16; /* Age product * fixed dummy covariate */
1237: #define APFQ 17; /* Age product * fixed quantitative covariate */
1238: #define APVD 18; /* Age product * varying dummy covariate */
1239: #define APVQ 19; /* Age product * varying quantitative covariate */
1240:
1241: #define FTYPE 1; /* Fixed covariate */
1242: #define VTYPE 2; /* Varying covariate (loop in wave) */
1243: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1244:
1245: struct kmodel{
1246: int maintype; /* main type */
1247: int subtype; /* subtype */
1248: };
1249: struct kmodel modell[NCOVMAX];
1250:
1.143 brouard 1251: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1252: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1253:
1254: /**************** split *************************/
1255: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1256: {
1257: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1258: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1259: */
1260: char *ss; /* pointer */
1.186 brouard 1261: int l1=0, l2=0; /* length counters */
1.126 brouard 1262:
1263: l1 = strlen(path ); /* length of path */
1264: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1265: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1266: if ( ss == NULL ) { /* no directory, so determine current directory */
1267: strcpy( name, path ); /* we got the fullname name because no directory */
1268: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1269: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1270: /* get current working directory */
1271: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1272: #ifdef WIN32
1273: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1274: #else
1275: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1276: #endif
1.126 brouard 1277: return( GLOCK_ERROR_GETCWD );
1278: }
1279: /* got dirc from getcwd*/
1280: printf(" DIRC = %s \n",dirc);
1.205 brouard 1281: } else { /* strip directory from path */
1.126 brouard 1282: ss++; /* after this, the filename */
1283: l2 = strlen( ss ); /* length of filename */
1284: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1285: strcpy( name, ss ); /* save file name */
1286: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1287: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1288: printf(" DIRC2 = %s \n",dirc);
1289: }
1290: /* We add a separator at the end of dirc if not exists */
1291: l1 = strlen( dirc ); /* length of directory */
1292: if( dirc[l1-1] != DIRSEPARATOR ){
1293: dirc[l1] = DIRSEPARATOR;
1294: dirc[l1+1] = 0;
1295: printf(" DIRC3 = %s \n",dirc);
1296: }
1297: ss = strrchr( name, '.' ); /* find last / */
1298: if (ss >0){
1299: ss++;
1300: strcpy(ext,ss); /* save extension */
1301: l1= strlen( name);
1302: l2= strlen(ss)+1;
1303: strncpy( finame, name, l1-l2);
1304: finame[l1-l2]= 0;
1305: }
1306:
1307: return( 0 ); /* we're done */
1308: }
1309:
1310:
1311: /******************************************/
1312:
1313: void replace_back_to_slash(char *s, char*t)
1314: {
1315: int i;
1316: int lg=0;
1317: i=0;
1318: lg=strlen(t);
1319: for(i=0; i<= lg; i++) {
1320: (s[i] = t[i]);
1321: if (t[i]== '\\') s[i]='/';
1322: }
1323: }
1324:
1.132 brouard 1325: char *trimbb(char *out, char *in)
1.137 brouard 1326: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1327: char *s;
1328: s=out;
1329: while (*in != '\0'){
1.137 brouard 1330: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1331: in++;
1332: }
1333: *out++ = *in++;
1334: }
1335: *out='\0';
1336: return s;
1337: }
1338:
1.187 brouard 1339: /* char *substrchaine(char *out, char *in, char *chain) */
1340: /* { */
1341: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1342: /* char *s, *t; */
1343: /* t=in;s=out; */
1344: /* while ((*in != *chain) && (*in != '\0')){ */
1345: /* *out++ = *in++; */
1346: /* } */
1347:
1348: /* /\* *in matches *chain *\/ */
1349: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1350: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1351: /* } */
1352: /* in--; chain--; */
1353: /* while ( (*in != '\0')){ */
1354: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1355: /* *out++ = *in++; */
1356: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1357: /* } */
1358: /* *out='\0'; */
1359: /* out=s; */
1360: /* return out; */
1361: /* } */
1362: char *substrchaine(char *out, char *in, char *chain)
1363: {
1364: /* Substract chain 'chain' from 'in', return and output 'out' */
1365: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1366:
1367: char *strloc;
1368:
1369: strcpy (out, in);
1370: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1371: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1372: if(strloc != NULL){
1373: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1374: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1375: /* strcpy (strloc, strloc +strlen(chain));*/
1376: }
1377: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1378: return out;
1379: }
1380:
1381:
1.145 brouard 1382: char *cutl(char *blocc, char *alocc, char *in, char occ)
1383: {
1.187 brouard 1384: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1385: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1386: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1387: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1388: */
1.160 brouard 1389: char *s, *t;
1.145 brouard 1390: t=in;s=in;
1391: while ((*in != occ) && (*in != '\0')){
1392: *alocc++ = *in++;
1393: }
1394: if( *in == occ){
1395: *(alocc)='\0';
1396: s=++in;
1397: }
1398:
1399: if (s == t) {/* occ not found */
1400: *(alocc-(in-s))='\0';
1401: in=s;
1402: }
1403: while ( *in != '\0'){
1404: *blocc++ = *in++;
1405: }
1406:
1407: *blocc='\0';
1408: return t;
1409: }
1.137 brouard 1410: char *cutv(char *blocc, char *alocc, char *in, char occ)
1411: {
1.187 brouard 1412: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1413: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1414: gives blocc="abcdef2ghi" and alocc="j".
1415: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1416: */
1417: char *s, *t;
1418: t=in;s=in;
1419: while (*in != '\0'){
1420: while( *in == occ){
1421: *blocc++ = *in++;
1422: s=in;
1423: }
1424: *blocc++ = *in++;
1425: }
1426: if (s == t) /* occ not found */
1427: *(blocc-(in-s))='\0';
1428: else
1429: *(blocc-(in-s)-1)='\0';
1430: in=s;
1431: while ( *in != '\0'){
1432: *alocc++ = *in++;
1433: }
1434:
1435: *alocc='\0';
1436: return s;
1437: }
1438:
1.126 brouard 1439: int nbocc(char *s, char occ)
1440: {
1441: int i,j=0;
1442: int lg=20;
1443: i=0;
1444: lg=strlen(s);
1445: for(i=0; i<= lg; i++) {
1.234 brouard 1446: if (s[i] == occ ) j++;
1.126 brouard 1447: }
1448: return j;
1449: }
1450:
1.137 brouard 1451: /* void cutv(char *u,char *v, char*t, char occ) */
1452: /* { */
1453: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1454: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1455: /* gives u="abcdef2ghi" and v="j" *\/ */
1456: /* int i,lg,j,p=0; */
1457: /* i=0; */
1458: /* lg=strlen(t); */
1459: /* for(j=0; j<=lg-1; j++) { */
1460: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1461: /* } */
1.126 brouard 1462:
1.137 brouard 1463: /* for(j=0; j<p; j++) { */
1464: /* (u[j] = t[j]); */
1465: /* } */
1466: /* u[p]='\0'; */
1.126 brouard 1467:
1.137 brouard 1468: /* for(j=0; j<= lg; j++) { */
1469: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1470: /* } */
1471: /* } */
1.126 brouard 1472:
1.160 brouard 1473: #ifdef _WIN32
1474: char * strsep(char **pp, const char *delim)
1475: {
1476: char *p, *q;
1477:
1478: if ((p = *pp) == NULL)
1479: return 0;
1480: if ((q = strpbrk (p, delim)) != NULL)
1481: {
1482: *pp = q + 1;
1483: *q = '\0';
1484: }
1485: else
1486: *pp = 0;
1487: return p;
1488: }
1489: #endif
1490:
1.126 brouard 1491: /********************** nrerror ********************/
1492:
1493: void nrerror(char error_text[])
1494: {
1495: fprintf(stderr,"ERREUR ...\n");
1496: fprintf(stderr,"%s\n",error_text);
1497: exit(EXIT_FAILURE);
1498: }
1499: /*********************** vector *******************/
1500: double *vector(int nl, int nh)
1501: {
1502: double *v;
1503: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1504: if (!v) nrerror("allocation failure in vector");
1505: return v-nl+NR_END;
1506: }
1507:
1508: /************************ free vector ******************/
1509: void free_vector(double*v, int nl, int nh)
1510: {
1511: free((FREE_ARG)(v+nl-NR_END));
1512: }
1513:
1514: /************************ivector *******************************/
1515: int *ivector(long nl,long nh)
1516: {
1517: int *v;
1518: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1519: if (!v) nrerror("allocation failure in ivector");
1520: return v-nl+NR_END;
1521: }
1522:
1523: /******************free ivector **************************/
1524: void free_ivector(int *v, long nl, long nh)
1525: {
1526: free((FREE_ARG)(v+nl-NR_END));
1527: }
1528:
1529: /************************lvector *******************************/
1530: long *lvector(long nl,long nh)
1531: {
1532: long *v;
1533: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1534: if (!v) nrerror("allocation failure in ivector");
1535: return v-nl+NR_END;
1536: }
1537:
1538: /******************free lvector **************************/
1539: void free_lvector(long *v, long nl, long nh)
1540: {
1541: free((FREE_ARG)(v+nl-NR_END));
1542: }
1543:
1544: /******************* imatrix *******************************/
1545: int **imatrix(long nrl, long nrh, long ncl, long nch)
1546: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1547: {
1548: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1549: int **m;
1550:
1551: /* allocate pointers to rows */
1552: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1553: if (!m) nrerror("allocation failure 1 in matrix()");
1554: m += NR_END;
1555: m -= nrl;
1556:
1557:
1558: /* allocate rows and set pointers to them */
1559: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1560: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1561: m[nrl] += NR_END;
1562: m[nrl] -= ncl;
1563:
1564: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1565:
1566: /* return pointer to array of pointers to rows */
1567: return m;
1568: }
1569:
1570: /****************** free_imatrix *************************/
1571: void free_imatrix(m,nrl,nrh,ncl,nch)
1572: int **m;
1573: long nch,ncl,nrh,nrl;
1574: /* free an int matrix allocated by imatrix() */
1575: {
1576: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1577: free((FREE_ARG) (m+nrl-NR_END));
1578: }
1579:
1580: /******************* matrix *******************************/
1581: double **matrix(long nrl, long nrh, long ncl, long nch)
1582: {
1583: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1584: double **m;
1585:
1586: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1587: if (!m) nrerror("allocation failure 1 in matrix()");
1588: m += NR_END;
1589: m -= nrl;
1590:
1591: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1592: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1593: m[nrl] += NR_END;
1594: m[nrl] -= ncl;
1595:
1596: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1597: return m;
1.145 brouard 1598: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1599: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1600: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1601: */
1602: }
1603:
1604: /*************************free matrix ************************/
1605: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1606: {
1607: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1608: free((FREE_ARG)(m+nrl-NR_END));
1609: }
1610:
1611: /******************* ma3x *******************************/
1612: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1613: {
1614: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1615: double ***m;
1616:
1617: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1618: if (!m) nrerror("allocation failure 1 in matrix()");
1619: m += NR_END;
1620: m -= nrl;
1621:
1622: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1623: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1624: m[nrl] += NR_END;
1625: m[nrl] -= ncl;
1626:
1627: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1628:
1629: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1630: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1631: m[nrl][ncl] += NR_END;
1632: m[nrl][ncl] -= nll;
1633: for (j=ncl+1; j<=nch; j++)
1634: m[nrl][j]=m[nrl][j-1]+nlay;
1635:
1636: for (i=nrl+1; i<=nrh; i++) {
1637: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1638: for (j=ncl+1; j<=nch; j++)
1639: m[i][j]=m[i][j-1]+nlay;
1640: }
1641: return m;
1642: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1643: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1644: */
1645: }
1646:
1647: /*************************free ma3x ************************/
1648: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1649: {
1650: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1651: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1652: free((FREE_ARG)(m+nrl-NR_END));
1653: }
1654:
1655: /*************** function subdirf ***********/
1656: char *subdirf(char fileres[])
1657: {
1658: /* Caution optionfilefiname is hidden */
1659: strcpy(tmpout,optionfilefiname);
1660: strcat(tmpout,"/"); /* Add to the right */
1661: strcat(tmpout,fileres);
1662: return tmpout;
1663: }
1664:
1665: /*************** function subdirf2 ***********/
1666: char *subdirf2(char fileres[], char *preop)
1667: {
1668:
1669: /* Caution optionfilefiname is hidden */
1670: strcpy(tmpout,optionfilefiname);
1671: strcat(tmpout,"/");
1672: strcat(tmpout,preop);
1673: strcat(tmpout,fileres);
1674: return tmpout;
1675: }
1676:
1677: /*************** function subdirf3 ***********/
1678: char *subdirf3(char fileres[], char *preop, char *preop2)
1679: {
1680:
1681: /* Caution optionfilefiname is hidden */
1682: strcpy(tmpout,optionfilefiname);
1683: strcat(tmpout,"/");
1684: strcat(tmpout,preop);
1685: strcat(tmpout,preop2);
1686: strcat(tmpout,fileres);
1687: return tmpout;
1688: }
1.213 brouard 1689:
1690: /*************** function subdirfext ***********/
1691: char *subdirfext(char fileres[], char *preop, char *postop)
1692: {
1693:
1694: strcpy(tmpout,preop);
1695: strcat(tmpout,fileres);
1696: strcat(tmpout,postop);
1697: return tmpout;
1698: }
1.126 brouard 1699:
1.213 brouard 1700: /*************** function subdirfext3 ***********/
1701: char *subdirfext3(char fileres[], char *preop, char *postop)
1702: {
1703:
1704: /* Caution optionfilefiname is hidden */
1705: strcpy(tmpout,optionfilefiname);
1706: strcat(tmpout,"/");
1707: strcat(tmpout,preop);
1708: strcat(tmpout,fileres);
1709: strcat(tmpout,postop);
1710: return tmpout;
1711: }
1712:
1.162 brouard 1713: char *asc_diff_time(long time_sec, char ascdiff[])
1714: {
1715: long sec_left, days, hours, minutes;
1716: days = (time_sec) / (60*60*24);
1717: sec_left = (time_sec) % (60*60*24);
1718: hours = (sec_left) / (60*60) ;
1719: sec_left = (sec_left) %(60*60);
1720: minutes = (sec_left) /60;
1721: sec_left = (sec_left) % (60);
1722: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1723: return ascdiff;
1724: }
1725:
1.126 brouard 1726: /***************** f1dim *************************/
1727: extern int ncom;
1728: extern double *pcom,*xicom;
1729: extern double (*nrfunc)(double []);
1730:
1731: double f1dim(double x)
1732: {
1733: int j;
1734: double f;
1735: double *xt;
1736:
1737: xt=vector(1,ncom);
1738: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1739: f=(*nrfunc)(xt);
1740: free_vector(xt,1,ncom);
1741: return f;
1742: }
1743:
1744: /*****************brent *************************/
1745: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1746: {
1747: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1748: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1749: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1750: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1751: * returned function value.
1752: */
1.126 brouard 1753: int iter;
1754: double a,b,d,etemp;
1.159 brouard 1755: double fu=0,fv,fw,fx;
1.164 brouard 1756: double ftemp=0.;
1.126 brouard 1757: double p,q,r,tol1,tol2,u,v,w,x,xm;
1758: double e=0.0;
1759:
1760: a=(ax < cx ? ax : cx);
1761: b=(ax > cx ? ax : cx);
1762: x=w=v=bx;
1763: fw=fv=fx=(*f)(x);
1764: for (iter=1;iter<=ITMAX;iter++) {
1765: xm=0.5*(a+b);
1766: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1767: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1768: printf(".");fflush(stdout);
1769: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1770: #ifdef DEBUGBRENT
1.126 brouard 1771: 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);
1772: 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);
1773: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1774: #endif
1775: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1776: *xmin=x;
1777: return fx;
1778: }
1779: ftemp=fu;
1780: if (fabs(e) > tol1) {
1781: r=(x-w)*(fx-fv);
1782: q=(x-v)*(fx-fw);
1783: p=(x-v)*q-(x-w)*r;
1784: q=2.0*(q-r);
1785: if (q > 0.0) p = -p;
1786: q=fabs(q);
1787: etemp=e;
1788: e=d;
1789: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1790: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1791: else {
1.224 brouard 1792: d=p/q;
1793: u=x+d;
1794: if (u-a < tol2 || b-u < tol2)
1795: d=SIGN(tol1,xm-x);
1.126 brouard 1796: }
1797: } else {
1798: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1799: }
1800: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1801: fu=(*f)(u);
1802: if (fu <= fx) {
1803: if (u >= x) a=x; else b=x;
1804: SHFT(v,w,x,u)
1.183 brouard 1805: SHFT(fv,fw,fx,fu)
1806: } else {
1807: if (u < x) a=u; else b=u;
1808: if (fu <= fw || w == x) {
1.224 brouard 1809: v=w;
1810: w=u;
1811: fv=fw;
1812: fw=fu;
1.183 brouard 1813: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1814: v=u;
1815: fv=fu;
1.183 brouard 1816: }
1817: }
1.126 brouard 1818: }
1819: nrerror("Too many iterations in brent");
1820: *xmin=x;
1821: return fx;
1822: }
1823:
1824: /****************** mnbrak ***********************/
1825:
1826: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1827: double (*func)(double))
1.183 brouard 1828: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1829: the downhill direction (defined by the function as evaluated at the initial points) and returns
1830: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1831: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1832: */
1.126 brouard 1833: double ulim,u,r,q, dum;
1834: double fu;
1.187 brouard 1835:
1836: double scale=10.;
1837: int iterscale=0;
1838:
1839: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1840: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1841:
1842:
1843: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1844: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1845: /* *bx = *ax - (*ax - *bx)/scale; */
1846: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1847: /* } */
1848:
1.126 brouard 1849: if (*fb > *fa) {
1850: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1851: SHFT(dum,*fb,*fa,dum)
1852: }
1.126 brouard 1853: *cx=(*bx)+GOLD*(*bx-*ax);
1854: *fc=(*func)(*cx);
1.183 brouard 1855: #ifdef DEBUG
1.224 brouard 1856: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1857: 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 1858: #endif
1.224 brouard 1859: 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 1860: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1861: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1862: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1863: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1864: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1865: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1866: fu=(*func)(u);
1.163 brouard 1867: #ifdef DEBUG
1868: /* f(x)=A(x-u)**2+f(u) */
1869: double A, fparabu;
1870: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1871: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1872: 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);
1873: 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 1874: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1875: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1876: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1877: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1878: #endif
1.184 brouard 1879: #ifdef MNBRAKORIGINAL
1.183 brouard 1880: #else
1.191 brouard 1881: /* if (fu > *fc) { */
1882: /* #ifdef DEBUG */
1883: /* printf("mnbrak4 fu > fc \n"); */
1884: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1885: /* #endif */
1886: /* /\* 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 *\\/ *\/ */
1887: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1888: /* dum=u; /\* Shifting c and u *\/ */
1889: /* u = *cx; */
1890: /* *cx = dum; */
1891: /* dum = fu; */
1892: /* fu = *fc; */
1893: /* *fc =dum; */
1894: /* } else { /\* end *\/ */
1895: /* #ifdef DEBUG */
1896: /* printf("mnbrak3 fu < fc \n"); */
1897: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1898: /* #endif */
1899: /* dum=u; /\* Shifting c and u *\/ */
1900: /* u = *cx; */
1901: /* *cx = dum; */
1902: /* dum = fu; */
1903: /* fu = *fc; */
1904: /* *fc =dum; */
1905: /* } */
1.224 brouard 1906: #ifdef DEBUGMNBRAK
1907: double A, fparabu;
1908: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1909: fparabu= *fa - A*(*ax-u)*(*ax-u);
1910: 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);
1911: 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 1912: #endif
1.191 brouard 1913: dum=u; /* Shifting c and u */
1914: u = *cx;
1915: *cx = dum;
1916: dum = fu;
1917: fu = *fc;
1918: *fc =dum;
1.183 brouard 1919: #endif
1.162 brouard 1920: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1921: #ifdef DEBUG
1.224 brouard 1922: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1923: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1924: #endif
1.126 brouard 1925: fu=(*func)(u);
1926: if (fu < *fc) {
1.183 brouard 1927: #ifdef DEBUG
1.224 brouard 1928: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1929: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1930: #endif
1931: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1932: SHFT(*fb,*fc,fu,(*func)(u))
1933: #ifdef DEBUG
1934: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1935: #endif
1936: }
1.162 brouard 1937: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1938: #ifdef DEBUG
1.224 brouard 1939: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1940: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1941: #endif
1.126 brouard 1942: u=ulim;
1943: fu=(*func)(u);
1.183 brouard 1944: } else { /* u could be left to b (if r > q parabola has a maximum) */
1945: #ifdef DEBUG
1.224 brouard 1946: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1947: 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 1948: #endif
1.126 brouard 1949: u=(*cx)+GOLD*(*cx-*bx);
1950: fu=(*func)(u);
1.224 brouard 1951: #ifdef DEBUG
1952: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1953: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1954: #endif
1.183 brouard 1955: } /* end tests */
1.126 brouard 1956: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1957: SHFT(*fa,*fb,*fc,fu)
1958: #ifdef DEBUG
1.224 brouard 1959: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1960: 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 1961: #endif
1962: } /* 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 1963: }
1964:
1965: /*************** linmin ************************/
1.162 brouard 1966: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1967: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1968: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1969: the value of func at the returned location p . This is actually all accomplished by calling the
1970: routines mnbrak and brent .*/
1.126 brouard 1971: int ncom;
1972: double *pcom,*xicom;
1973: double (*nrfunc)(double []);
1974:
1.224 brouard 1975: #ifdef LINMINORIGINAL
1.126 brouard 1976: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1977: #else
1978: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1979: #endif
1.126 brouard 1980: {
1981: double brent(double ax, double bx, double cx,
1982: double (*f)(double), double tol, double *xmin);
1983: double f1dim(double x);
1984: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
1985: double *fc, double (*func)(double));
1986: int j;
1987: double xx,xmin,bx,ax;
1988: double fx,fb,fa;
1.187 brouard 1989:
1.203 brouard 1990: #ifdef LINMINORIGINAL
1991: #else
1992: double scale=10., axs, xxs; /* Scale added for infinity */
1993: #endif
1994:
1.126 brouard 1995: ncom=n;
1996: pcom=vector(1,n);
1997: xicom=vector(1,n);
1998: nrfunc=func;
1999: for (j=1;j<=n;j++) {
2000: pcom[j]=p[j];
1.202 brouard 2001: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 2002: }
1.187 brouard 2003:
1.203 brouard 2004: #ifdef LINMINORIGINAL
2005: xx=1.;
2006: #else
2007: axs=0.0;
2008: xxs=1.;
2009: do{
2010: xx= xxs;
2011: #endif
1.187 brouard 2012: ax=0.;
2013: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
2014: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
2015: /* 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)) */
2016: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
2017: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
2018: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
2019: /* 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 2020: #ifdef LINMINORIGINAL
2021: #else
2022: if (fx != fx){
1.224 brouard 2023: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
2024: printf("|");
2025: fprintf(ficlog,"|");
1.203 brouard 2026: #ifdef DEBUGLINMIN
1.224 brouard 2027: 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 2028: #endif
2029: }
1.224 brouard 2030: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 2031: #endif
2032:
1.191 brouard 2033: #ifdef DEBUGLINMIN
2034: 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 2035: 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 2036: #endif
1.224 brouard 2037: #ifdef LINMINORIGINAL
2038: #else
2039: if(fb == fx){ /* Flat function in the direction */
2040: xmin=xx;
2041: *flat=1;
2042: }else{
2043: *flat=0;
2044: #endif
2045: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 2046: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
2047: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
2048: /* fmin = f(p[j] + xmin * xi[j]) */
2049: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
2050: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 2051: #ifdef DEBUG
1.224 brouard 2052: 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);
2053: 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);
2054: #endif
2055: #ifdef LINMINORIGINAL
2056: #else
2057: }
1.126 brouard 2058: #endif
1.191 brouard 2059: #ifdef DEBUGLINMIN
2060: printf("linmin end ");
1.202 brouard 2061: fprintf(ficlog,"linmin end ");
1.191 brouard 2062: #endif
1.126 brouard 2063: for (j=1;j<=n;j++) {
1.203 brouard 2064: #ifdef LINMINORIGINAL
2065: xi[j] *= xmin;
2066: #else
2067: #ifdef DEBUGLINMIN
2068: if(xxs <1.0)
2069: printf(" before xi[%d]=%12.8f", j,xi[j]);
2070: #endif
2071: 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) */
2072: #ifdef DEBUGLINMIN
2073: if(xxs <1.0)
2074: 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 );
2075: #endif
2076: #endif
1.187 brouard 2077: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2078: }
1.191 brouard 2079: #ifdef DEBUGLINMIN
1.203 brouard 2080: printf("\n");
1.191 brouard 2081: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2082: 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 2083: for (j=1;j<=n;j++) {
1.202 brouard 2084: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2085: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2086: if(j % ncovmodel == 0){
1.191 brouard 2087: printf("\n");
1.202 brouard 2088: fprintf(ficlog,"\n");
2089: }
1.191 brouard 2090: }
1.203 brouard 2091: #else
1.191 brouard 2092: #endif
1.126 brouard 2093: free_vector(xicom,1,n);
2094: free_vector(pcom,1,n);
2095: }
2096:
2097:
2098: /*************** powell ************************/
1.162 brouard 2099: /*
2100: Minimization of a function func of n variables. Input consists of an initial starting point
2101: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2102: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2103: such that failure to decrease by more than this amount on one iteration signals doneness. On
2104: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2105: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2106: */
1.224 brouard 2107: #ifdef LINMINORIGINAL
2108: #else
2109: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2110: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2111: #endif
1.126 brouard 2112: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2113: double (*func)(double []))
2114: {
1.224 brouard 2115: #ifdef LINMINORIGINAL
2116: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2117: double (*func)(double []));
1.224 brouard 2118: #else
1.241 brouard 2119: void linmin(double p[], double xi[], int n, double *fret,
2120: double (*func)(double []),int *flat);
1.224 brouard 2121: #endif
1.239 brouard 2122: int i,ibig,j,jk,k;
1.126 brouard 2123: double del,t,*pt,*ptt,*xit;
1.181 brouard 2124: double directest;
1.126 brouard 2125: double fp,fptt;
2126: double *xits;
2127: int niterf, itmp;
1.224 brouard 2128: #ifdef LINMINORIGINAL
2129: #else
2130:
2131: flatdir=ivector(1,n);
2132: for (j=1;j<=n;j++) flatdir[j]=0;
2133: #endif
1.126 brouard 2134:
2135: pt=vector(1,n);
2136: ptt=vector(1,n);
2137: xit=vector(1,n);
2138: xits=vector(1,n);
2139: *fret=(*func)(p);
2140: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2141: rcurr_time = time(NULL);
1.126 brouard 2142: for (*iter=1;;++(*iter)) {
1.187 brouard 2143: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2144: ibig=0;
2145: del=0.0;
1.157 brouard 2146: rlast_time=rcurr_time;
2147: /* (void) gettimeofday(&curr_time,&tzp); */
2148: rcurr_time = time(NULL);
2149: curr_time = *localtime(&rcurr_time);
2150: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2151: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2152: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2153: for (i=1;i<=n;i++) {
1.126 brouard 2154: fprintf(ficrespow," %.12lf", p[i]);
2155: }
1.239 brouard 2156: fprintf(ficrespow,"\n");fflush(ficrespow);
2157: printf("\n#model= 1 + age ");
2158: fprintf(ficlog,"\n#model= 1 + age ");
2159: if(nagesqr==1){
1.241 brouard 2160: printf(" + age*age ");
2161: fprintf(ficlog," + age*age ");
1.239 brouard 2162: }
2163: for(j=1;j <=ncovmodel-2;j++){
2164: if(Typevar[j]==0) {
2165: printf(" + V%d ",Tvar[j]);
2166: fprintf(ficlog," + V%d ",Tvar[j]);
2167: }else if(Typevar[j]==1) {
2168: printf(" + V%d*age ",Tvar[j]);
2169: fprintf(ficlog," + V%d*age ",Tvar[j]);
2170: }else if(Typevar[j]==2) {
2171: printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2172: fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2173: }
2174: }
1.126 brouard 2175: printf("\n");
1.239 brouard 2176: /* printf("12 47.0114589 0.0154322 33.2424412 0.3279905 2.3731903 */
2177: /* 13 -21.5392400 0.1118147 1.2680506 1.2973408 -1.0663662 */
1.126 brouard 2178: fprintf(ficlog,"\n");
1.239 brouard 2179: for(i=1,jk=1; i <=nlstate; i++){
2180: for(k=1; k <=(nlstate+ndeath); k++){
2181: if (k != i) {
2182: printf("%d%d ",i,k);
2183: fprintf(ficlog,"%d%d ",i,k);
2184: for(j=1; j <=ncovmodel; j++){
2185: printf("%12.7f ",p[jk]);
2186: fprintf(ficlog,"%12.7f ",p[jk]);
2187: jk++;
2188: }
2189: printf("\n");
2190: fprintf(ficlog,"\n");
2191: }
2192: }
2193: }
1.241 brouard 2194: if(*iter <=3 && *iter >1){
1.157 brouard 2195: tml = *localtime(&rcurr_time);
2196: strcpy(strcurr,asctime(&tml));
2197: rforecast_time=rcurr_time;
1.126 brouard 2198: itmp = strlen(strcurr);
2199: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.241 brouard 2200: strcurr[itmp-1]='\0';
1.162 brouard 2201: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2202: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2203: for(niterf=10;niterf<=30;niterf+=10){
1.241 brouard 2204: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2205: forecast_time = *localtime(&rforecast_time);
2206: strcpy(strfor,asctime(&forecast_time));
2207: itmp = strlen(strfor);
2208: if(strfor[itmp-1]=='\n')
2209: strfor[itmp-1]='\0';
2210: 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);
2211: 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 2212: }
2213: }
1.187 brouard 2214: for (i=1;i<=n;i++) { /* For each direction i */
2215: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2216: fptt=(*fret);
2217: #ifdef DEBUG
1.203 brouard 2218: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2219: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2220: #endif
1.203 brouard 2221: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2222: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2223: #ifdef LINMINORIGINAL
1.188 brouard 2224: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2225: #else
2226: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2227: flatdir[i]=flat; /* Function is vanishing in that direction i */
2228: #endif
2229: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2230: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2231: /* because that direction will be replaced unless the gain del is small */
2232: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2233: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2234: /* with the new direction. */
2235: del=fabs(fptt-(*fret));
2236: ibig=i;
1.126 brouard 2237: }
2238: #ifdef DEBUG
2239: printf("%d %.12e",i,(*fret));
2240: fprintf(ficlog,"%d %.12e",i,(*fret));
2241: for (j=1;j<=n;j++) {
1.224 brouard 2242: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2243: printf(" x(%d)=%.12e",j,xit[j]);
2244: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2245: }
2246: for(j=1;j<=n;j++) {
1.225 brouard 2247: printf(" p(%d)=%.12e",j,p[j]);
2248: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2249: }
2250: printf("\n");
2251: fprintf(ficlog,"\n");
2252: #endif
1.187 brouard 2253: } /* end loop on each direction i */
2254: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2255: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2256: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2257: for(j=1;j<=n;j++) {
1.225 brouard 2258: if(flatdir[j] >0){
2259: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2260: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2261: }
2262: /* printf("\n"); */
2263: /* fprintf(ficlog,"\n"); */
2264: }
1.243 brouard 2265: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */
2266: if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */
1.188 brouard 2267: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2268: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2269: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2270: /* decreased of more than 3.84 */
2271: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2272: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2273: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2274:
1.188 brouard 2275: /* Starting the program with initial values given by a former maximization will simply change */
2276: /* the scales of the directions and the directions, because the are reset to canonical directions */
2277: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2278: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2279: #ifdef DEBUG
2280: int k[2],l;
2281: k[0]=1;
2282: k[1]=-1;
2283: printf("Max: %.12e",(*func)(p));
2284: fprintf(ficlog,"Max: %.12e",(*func)(p));
2285: for (j=1;j<=n;j++) {
2286: printf(" %.12e",p[j]);
2287: fprintf(ficlog," %.12e",p[j]);
2288: }
2289: printf("\n");
2290: fprintf(ficlog,"\n");
2291: for(l=0;l<=1;l++) {
2292: for (j=1;j<=n;j++) {
2293: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2294: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2295: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2296: }
2297: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2298: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2299: }
2300: #endif
2301:
1.224 brouard 2302: #ifdef LINMINORIGINAL
2303: #else
2304: free_ivector(flatdir,1,n);
2305: #endif
1.126 brouard 2306: free_vector(xit,1,n);
2307: free_vector(xits,1,n);
2308: free_vector(ptt,1,n);
2309: free_vector(pt,1,n);
2310: return;
1.192 brouard 2311: } /* enough precision */
1.240 brouard 2312: if (*iter == ITMAX*n) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2313: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2314: ptt[j]=2.0*p[j]-pt[j];
2315: xit[j]=p[j]-pt[j];
2316: pt[j]=p[j];
2317: }
1.181 brouard 2318: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2319: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2320: if (*iter <=4) {
1.225 brouard 2321: #else
2322: #endif
1.224 brouard 2323: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2324: #else
1.161 brouard 2325: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2326: #endif
1.162 brouard 2327: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2328: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2329: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2330: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2331: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2332: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2333: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2334: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2335: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2336: /* Even if f3 <f1, directest can be negative and t >0 */
2337: /* mu² and del² are equal when f3=f1 */
2338: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2339: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2340: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2341: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2342: #ifdef NRCORIGINAL
2343: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2344: #else
2345: 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 2346: t= t- del*SQR(fp-fptt);
1.183 brouard 2347: #endif
1.202 brouard 2348: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2349: #ifdef DEBUG
1.181 brouard 2350: 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);
2351: 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 2352: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2353: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2354: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2355: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2356: 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);
2357: 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);
2358: #endif
1.183 brouard 2359: #ifdef POWELLORIGINAL
2360: if (t < 0.0) { /* Then we use it for new direction */
2361: #else
1.182 brouard 2362: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2363: 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 2364: 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 2365: 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 2366: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2367: }
1.181 brouard 2368: if (directest < 0.0) { /* Then we use it for new direction */
2369: #endif
1.191 brouard 2370: #ifdef DEBUGLINMIN
1.234 brouard 2371: printf("Before linmin in direction P%d-P0\n",n);
2372: for (j=1;j<=n;j++) {
2373: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2374: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2375: if(j % ncovmodel == 0){
2376: printf("\n");
2377: fprintf(ficlog,"\n");
2378: }
2379: }
1.224 brouard 2380: #endif
2381: #ifdef LINMINORIGINAL
1.234 brouard 2382: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2383: #else
1.234 brouard 2384: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2385: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2386: #endif
1.234 brouard 2387:
1.191 brouard 2388: #ifdef DEBUGLINMIN
1.234 brouard 2389: for (j=1;j<=n;j++) {
2390: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2391: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2392: if(j % ncovmodel == 0){
2393: printf("\n");
2394: fprintf(ficlog,"\n");
2395: }
2396: }
1.224 brouard 2397: #endif
1.234 brouard 2398: for (j=1;j<=n;j++) {
2399: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2400: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2401: }
1.224 brouard 2402: #ifdef LINMINORIGINAL
2403: #else
1.234 brouard 2404: for (j=1, flatd=0;j<=n;j++) {
2405: if(flatdir[j]>0)
2406: flatd++;
2407: }
2408: if(flatd >0){
1.255 brouard 2409: printf("%d flat directions: ",flatd);
2410: fprintf(ficlog,"%d flat directions :",flatd);
1.234 brouard 2411: for (j=1;j<=n;j++) {
2412: if(flatdir[j]>0){
2413: printf("%d ",j);
2414: fprintf(ficlog,"%d ",j);
2415: }
2416: }
2417: printf("\n");
2418: fprintf(ficlog,"\n");
2419: }
1.191 brouard 2420: #endif
1.234 brouard 2421: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2422: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2423:
1.126 brouard 2424: #ifdef DEBUG
1.234 brouard 2425: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2426: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2427: for(j=1;j<=n;j++){
2428: printf(" %lf",xit[j]);
2429: fprintf(ficlog," %lf",xit[j]);
2430: }
2431: printf("\n");
2432: fprintf(ficlog,"\n");
1.126 brouard 2433: #endif
1.192 brouard 2434: } /* end of t or directest negative */
1.224 brouard 2435: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2436: #else
1.234 brouard 2437: } /* end if (fptt < fp) */
1.192 brouard 2438: #endif
1.225 brouard 2439: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2440: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2441: #else
1.224 brouard 2442: #endif
1.234 brouard 2443: } /* loop iteration */
1.126 brouard 2444: }
1.234 brouard 2445:
1.126 brouard 2446: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2447:
1.235 brouard 2448: 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 2449: {
1.235 brouard 2450: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2451: (and selected quantitative values in nres)
2452: by left multiplying the unit
1.234 brouard 2453: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2454: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2455: /* Wx is row vector: population in state 1, population in state 2, population dead */
2456: /* or prevalence in state 1, prevalence in state 2, 0 */
2457: /* newm is the matrix after multiplications, its rows are identical at a factor */
2458: /* Initial matrix pimij */
2459: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2460: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2461: /* 0, 0 , 1} */
2462: /*
2463: * and after some iteration: */
2464: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2465: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2466: /* 0, 0 , 1} */
2467: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2468: /* {0.51571254859325999, 0.4842874514067399, */
2469: /* 0.51326036147820708, 0.48673963852179264} */
2470: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2471:
1.126 brouard 2472: int i, ii,j,k;
1.209 brouard 2473: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2474: /* double **matprod2(); */ /* test */
1.218 brouard 2475: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2476: double **newm;
1.209 brouard 2477: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2478: int ncvloop=0;
1.169 brouard 2479:
1.209 brouard 2480: min=vector(1,nlstate);
2481: max=vector(1,nlstate);
2482: meandiff=vector(1,nlstate);
2483:
1.218 brouard 2484: /* Starting with matrix unity */
1.126 brouard 2485: for (ii=1;ii<=nlstate+ndeath;ii++)
2486: for (j=1;j<=nlstate+ndeath;j++){
2487: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2488: }
1.169 brouard 2489:
2490: cov[1]=1.;
2491:
2492: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2493: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2494: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2495: ncvloop++;
1.126 brouard 2496: newm=savm;
2497: /* Covariates have to be included here again */
1.138 brouard 2498: cov[2]=agefin;
1.187 brouard 2499: if(nagesqr==1)
2500: cov[3]= agefin*agefin;;
1.234 brouard 2501: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2502: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2503: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2504: /* 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 2505: }
2506: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2507: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2508: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2509: /* 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 2510: }
1.237 brouard 2511: for (k=1; k<=cptcovage;k++){ /* For product with age */
1.234 brouard 2512: if(Dummy[Tvar[Tage[k]]]){
2513: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2514: } else{
1.235 brouard 2515: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2516: }
1.235 brouard 2517: /* 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 2518: }
1.237 brouard 2519: for (k=1; k<=cptcovprod;k++){ /* For product without age */
1.235 brouard 2520: /* 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 2521: if(Dummy[Tvard[k][1]==0]){
2522: if(Dummy[Tvard[k][2]==0]){
2523: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2524: }else{
2525: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2526: }
2527: }else{
2528: if(Dummy[Tvard[k][2]==0]){
2529: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2530: }else{
2531: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2532: }
2533: }
1.234 brouard 2534: }
1.138 brouard 2535: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2536: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2537: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2538: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2539: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2540: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2541: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2542:
1.126 brouard 2543: savm=oldm;
2544: oldm=newm;
1.209 brouard 2545:
2546: for(j=1; j<=nlstate; j++){
2547: max[j]=0.;
2548: min[j]=1.;
2549: }
2550: for(i=1;i<=nlstate;i++){
2551: sumnew=0;
2552: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2553: for(j=1; j<=nlstate; j++){
2554: prlim[i][j]= newm[i][j]/(1-sumnew);
2555: max[j]=FMAX(max[j],prlim[i][j]);
2556: min[j]=FMIN(min[j],prlim[i][j]);
2557: }
2558: }
2559:
1.126 brouard 2560: maxmax=0.;
1.209 brouard 2561: for(j=1; j<=nlstate; j++){
2562: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2563: maxmax=FMAX(maxmax,meandiff[j]);
2564: /* 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 2565: } /* j loop */
1.203 brouard 2566: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2567: /* 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 2568: if(maxmax < ftolpl){
1.209 brouard 2569: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2570: free_vector(min,1,nlstate);
2571: free_vector(max,1,nlstate);
2572: free_vector(meandiff,1,nlstate);
1.126 brouard 2573: return prlim;
2574: }
1.169 brouard 2575: } /* age loop */
1.208 brouard 2576: /* After some age loop it doesn't converge */
1.209 brouard 2577: 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 2578: 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 2579: /* 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); */
2580: free_vector(min,1,nlstate);
2581: free_vector(max,1,nlstate);
2582: free_vector(meandiff,1,nlstate);
1.208 brouard 2583:
1.169 brouard 2584: return prlim; /* should not reach here */
1.126 brouard 2585: }
2586:
1.217 brouard 2587:
2588: /**** Back Prevalence limit (stable or period prevalence) ****************/
2589:
1.218 brouard 2590: /* 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) */
2591: /* 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 2592: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij, int nres)
1.217 brouard 2593: {
1.218 brouard 2594: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2595: matrix by transitions matrix until convergence is reached with precision ftolpl */
2596: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2597: /* Wx is row vector: population in state 1, population in state 2, population dead */
2598: /* or prevalence in state 1, prevalence in state 2, 0 */
2599: /* newm is the matrix after multiplications, its rows are identical at a factor */
2600: /* Initial matrix pimij */
2601: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2602: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2603: /* 0, 0 , 1} */
2604: /*
2605: * and after some iteration: */
2606: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2607: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2608: /* 0, 0 , 1} */
2609: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2610: /* {0.51571254859325999, 0.4842874514067399, */
2611: /* 0.51326036147820708, 0.48673963852179264} */
2612: /* If we start from prlim again, prlim tends to a constant matrix */
2613:
2614: int i, ii,j,k;
1.247 brouard 2615: int first=0;
1.217 brouard 2616: double *min, *max, *meandiff, maxmax,sumnew=0.;
2617: /* double **matprod2(); */ /* test */
2618: double **out, cov[NCOVMAX+1], **bmij();
2619: double **newm;
1.218 brouard 2620: double **dnewm, **doldm, **dsavm; /* for use */
2621: double **oldm, **savm; /* for use */
2622:
1.217 brouard 2623: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2624: int ncvloop=0;
2625:
2626: min=vector(1,nlstate);
2627: max=vector(1,nlstate);
2628: meandiff=vector(1,nlstate);
2629:
1.218 brouard 2630: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2631: oldm=oldms; savm=savms;
2632:
2633: /* Starting with matrix unity */
2634: for (ii=1;ii<=nlstate+ndeath;ii++)
2635: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2636: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2637: }
2638:
2639: cov[1]=1.;
2640:
2641: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2642: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2643: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2644: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2645: ncvloop++;
1.218 brouard 2646: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2647: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2648: /* Covariates have to be included here again */
2649: cov[2]=agefin;
2650: if(nagesqr==1)
2651: cov[3]= agefin*agefin;;
1.242 brouard 2652: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2653: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2654: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
2655: /* 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)); */
2656: }
2657: /* for (k=1; k<=cptcovn;k++) { */
2658: /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */
2659: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2660: /* /\* 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])]); *\/ */
2661: /* } */
2662: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2663: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2664: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2665: /* 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]); */
2666: }
2667: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2]; */
2668: /* for (k=1; k<=cptcovprod;k++) /\* Useless *\/ */
2669: /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */
2670: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */
2671: for (k=1; k<=cptcovage;k++){ /* For product with age */
2672: if(Dummy[Tvar[Tage[k]]]){
2673: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2674: } else{
2675: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2676: }
2677: /* 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]); */
2678: }
2679: for (k=1; k<=cptcovprod;k++){ /* For product without age */
2680: /* 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]); */
2681: if(Dummy[Tvard[k][1]==0]){
2682: if(Dummy[Tvard[k][2]==0]){
2683: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2684: }else{
2685: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2686: }
2687: }else{
2688: if(Dummy[Tvard[k][2]==0]){
2689: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2690: }else{
2691: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2692: }
2693: }
1.217 brouard 2694: }
2695:
2696: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2697: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2698: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2699: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2700: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2701: /* ij should be linked to the correct index of cov */
2702: /* age and covariate values ij are in 'cov', but we need to pass
2703: * ij for the observed prevalence at age and status and covariate
2704: * number: prevacurrent[(int)agefin][ii][ij]
2705: */
2706: /* 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 *\/ */
2707: /* 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 *\/ */
2708: 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 2709: savm=oldm;
2710: oldm=newm;
2711: for(j=1; j<=nlstate; j++){
2712: max[j]=0.;
2713: min[j]=1.;
2714: }
2715: for(j=1; j<=nlstate; j++){
2716: for(i=1;i<=nlstate;i++){
1.234 brouard 2717: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2718: bprlim[i][j]= newm[i][j];
2719: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2720: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2721: }
2722: }
1.218 brouard 2723:
1.217 brouard 2724: maxmax=0.;
2725: for(i=1; i<=nlstate; i++){
2726: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2727: maxmax=FMAX(maxmax,meandiff[i]);
2728: /* 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); */
2729: } /* j loop */
2730: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2731: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2732: if(maxmax < ftolpl){
1.220 brouard 2733: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2734: free_vector(min,1,nlstate);
2735: free_vector(max,1,nlstate);
2736: free_vector(meandiff,1,nlstate);
2737: return bprlim;
2738: }
2739: } /* age loop */
2740: /* After some age loop it doesn't converge */
1.247 brouard 2741: if(first){
2742: first=1;
2743: 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\
2744: 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);
2745: }
2746: 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 2747: 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);
2748: /* 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); */
2749: free_vector(min,1,nlstate);
2750: free_vector(max,1,nlstate);
2751: free_vector(meandiff,1,nlstate);
2752:
2753: return bprlim; /* should not reach here */
2754: }
2755:
1.126 brouard 2756: /*************** transition probabilities ***************/
2757:
2758: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2759: {
1.138 brouard 2760: /* According to parameters values stored in x and the covariate's values stored in cov,
2761: computes the probability to be observed in state j being in state i by appying the
2762: model to the ncovmodel covariates (including constant and age).
2763: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2764: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2765: ncth covariate in the global vector x is given by the formula:
2766: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2767: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2768: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2769: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2770: Outputs ps[i][j] the probability to be observed in j being in j according to
2771: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2772: */
2773: double s1, lnpijopii;
1.126 brouard 2774: /*double t34;*/
1.164 brouard 2775: int i,j, nc, ii, jj;
1.126 brouard 2776:
1.223 brouard 2777: for(i=1; i<= nlstate; i++){
2778: for(j=1; j<i;j++){
2779: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2780: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2781: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2782: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2783: }
2784: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2785: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2786: }
2787: for(j=i+1; j<=nlstate+ndeath;j++){
2788: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2789: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2790: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2791: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2792: }
2793: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2794: }
2795: }
1.218 brouard 2796:
1.223 brouard 2797: for(i=1; i<= nlstate; i++){
2798: s1=0;
2799: for(j=1; j<i; j++){
2800: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2801: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2802: }
2803: for(j=i+1; j<=nlstate+ndeath; j++){
2804: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2805: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2806: }
2807: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2808: ps[i][i]=1./(s1+1.);
2809: /* Computing other pijs */
2810: for(j=1; j<i; j++)
2811: ps[i][j]= exp(ps[i][j])*ps[i][i];
2812: for(j=i+1; j<=nlstate+ndeath; j++)
2813: ps[i][j]= exp(ps[i][j])*ps[i][i];
2814: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2815: } /* end i */
1.218 brouard 2816:
1.223 brouard 2817: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2818: for(jj=1; jj<= nlstate+ndeath; jj++){
2819: ps[ii][jj]=0;
2820: ps[ii][ii]=1;
2821: }
2822: }
1.218 brouard 2823:
2824:
1.223 brouard 2825: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2826: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2827: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2828: /* } */
2829: /* printf("\n "); */
2830: /* } */
2831: /* printf("\n ");printf("%lf ",cov[2]);*/
2832: /*
2833: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2834: goto end;*/
1.223 brouard 2835: return ps;
1.126 brouard 2836: }
2837:
1.218 brouard 2838: /*************** backward transition probabilities ***************/
2839:
2840: /* 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 ) */
2841: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2842: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2843: {
1.222 brouard 2844: /* Computes the backward probability at age agefin and covariate ij
2845: * and returns in **ps as well as **bmij.
2846: */
1.218 brouard 2847: int i, ii, j,k;
1.222 brouard 2848:
2849: double **out, **pmij();
2850: double sumnew=0.;
1.218 brouard 2851: double agefin;
1.222 brouard 2852:
2853: double **dnewm, **dsavm, **doldm;
2854: double **bbmij;
2855:
1.218 brouard 2856: doldm=ddoldms; /* global pointers */
1.222 brouard 2857: dnewm=ddnewms;
2858: dsavm=ddsavms;
2859:
2860: agefin=cov[2];
2861: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2862: the observed prevalence (with this covariate ij) */
2863: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2864: /* We do have the matrix Px in savm and we need pij */
2865: for (j=1;j<=nlstate+ndeath;j++){
2866: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2867: for (ii=1;ii<=nlstate;ii++){
2868: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2869: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2870: for (ii=1;ii<=nlstate+ndeath;ii++){
2871: if(sumnew >= 1.e-10){
2872: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2873: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2874: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2875: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2876: /* }else */
2877: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2878: }else{
1.242 brouard 2879: ;
2880: /* 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 2881: }
2882: } /*End ii */
2883: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2884: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2885: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2886: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2887: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2888: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2889: /* left Product of this matrix by diag matrix of prevalences (savm) */
2890: for (j=1;j<=nlstate+ndeath;j++){
2891: for (ii=1;ii<=nlstate+ndeath;ii++){
2892: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2893: }
2894: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2895: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2896: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2897: /* end bmij */
2898: return ps;
1.218 brouard 2899: }
1.217 brouard 2900: /*************** transition probabilities ***************/
2901:
1.218 brouard 2902: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2903: {
2904: /* According to parameters values stored in x and the covariate's values stored in cov,
2905: computes the probability to be observed in state j being in state i by appying the
2906: model to the ncovmodel covariates (including constant and age).
2907: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2908: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2909: ncth covariate in the global vector x is given by the formula:
2910: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2911: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2912: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2913: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2914: Outputs ps[i][j] the probability to be observed in j being in j according to
2915: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2916: */
2917: double s1, lnpijopii;
2918: /*double t34;*/
2919: int i,j, nc, ii, jj;
2920:
1.234 brouard 2921: for(i=1; i<= nlstate; i++){
2922: for(j=1; j<i;j++){
2923: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2924: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2925: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2926: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2927: }
2928: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2929: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2930: }
2931: for(j=i+1; j<=nlstate+ndeath;j++){
2932: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2933: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2934: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2935: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2936: }
2937: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2938: }
2939: }
2940:
2941: for(i=1; i<= nlstate; i++){
2942: s1=0;
2943: for(j=1; j<i; j++){
2944: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2945: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2946: }
2947: for(j=i+1; j<=nlstate+ndeath; j++){
2948: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2949: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2950: }
2951: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2952: ps[i][i]=1./(s1+1.);
2953: /* Computing other pijs */
2954: for(j=1; j<i; j++)
2955: ps[i][j]= exp(ps[i][j])*ps[i][i];
2956: for(j=i+1; j<=nlstate+ndeath; j++)
2957: ps[i][j]= exp(ps[i][j])*ps[i][i];
2958: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2959: } /* end i */
2960:
2961: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2962: for(jj=1; jj<= nlstate+ndeath; jj++){
2963: ps[ii][jj]=0;
2964: ps[ii][ii]=1;
2965: }
2966: }
2967: /* Added for backcast */ /* Transposed matrix too */
2968: for(jj=1; jj<= nlstate+ndeath; jj++){
2969: s1=0.;
2970: for(ii=1; ii<= nlstate+ndeath; ii++){
2971: s1+=ps[ii][jj];
2972: }
2973: for(ii=1; ii<= nlstate; ii++){
2974: ps[ii][jj]=ps[ii][jj]/s1;
2975: }
2976: }
2977: /* Transposition */
2978: for(jj=1; jj<= nlstate+ndeath; jj++){
2979: for(ii=jj; ii<= nlstate+ndeath; ii++){
2980: s1=ps[ii][jj];
2981: ps[ii][jj]=ps[jj][ii];
2982: ps[jj][ii]=s1;
2983: }
2984: }
2985: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2986: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2987: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2988: /* } */
2989: /* printf("\n "); */
2990: /* } */
2991: /* printf("\n ");printf("%lf ",cov[2]);*/
2992: /*
2993: for(i=1; i<= npar; i++) printf("%f ",x[i]);
2994: goto end;*/
2995: return ps;
1.217 brouard 2996: }
2997:
2998:
1.126 brouard 2999: /**************** Product of 2 matrices ******************/
3000:
1.145 brouard 3001: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 3002: {
3003: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
3004: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
3005: /* in, b, out are matrice of pointers which should have been initialized
3006: before: only the contents of out is modified. The function returns
3007: a pointer to pointers identical to out */
1.145 brouard 3008: int i, j, k;
1.126 brouard 3009: for(i=nrl; i<= nrh; i++)
1.145 brouard 3010: for(k=ncolol; k<=ncoloh; k++){
3011: out[i][k]=0.;
3012: for(j=ncl; j<=nch; j++)
3013: out[i][k] +=in[i][j]*b[j][k];
3014: }
1.126 brouard 3015: return out;
3016: }
3017:
3018:
3019: /************* Higher Matrix Product ***************/
3020:
1.235 brouard 3021: 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 3022: {
1.218 brouard 3023: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 3024: 'nhstepm*hstepm*stepm' months (i.e. until
3025: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3026: nhstepm*hstepm matrices.
3027: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3028: (typically every 2 years instead of every month which is too big
3029: for the memory).
3030: Model is determined by parameters x and covariates have to be
3031: included manually here.
3032:
3033: */
3034:
3035: int i, j, d, h, k;
1.131 brouard 3036: double **out, cov[NCOVMAX+1];
1.126 brouard 3037: double **newm;
1.187 brouard 3038: double agexact;
1.214 brouard 3039: double agebegin, ageend;
1.126 brouard 3040:
3041: /* Hstepm could be zero and should return the unit matrix */
3042: for (i=1;i<=nlstate+ndeath;i++)
3043: for (j=1;j<=nlstate+ndeath;j++){
3044: oldm[i][j]=(i==j ? 1.0 : 0.0);
3045: po[i][j][0]=(i==j ? 1.0 : 0.0);
3046: }
3047: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3048: for(h=1; h <=nhstepm; h++){
3049: for(d=1; d <=hstepm; d++){
3050: newm=savm;
3051: /* Covariates have to be included here again */
3052: cov[1]=1.;
1.214 brouard 3053: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 3054: cov[2]=agexact;
3055: if(nagesqr==1)
1.227 brouard 3056: cov[3]= agexact*agexact;
1.235 brouard 3057: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
3058: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
3059: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
3060: /* 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)); */
3061: }
3062: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
3063: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
3064: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
3065: /* 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]); */
3066: }
3067: for (k=1; k<=cptcovage;k++){
3068: if(Dummy[Tvar[Tage[k]]]){
3069: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3070: } else{
3071: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
3072: }
3073: /* 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]); */
3074: }
3075: for (k=1; k<=cptcovprod;k++){ /* */
3076: /* 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]); */
3077: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
3078: }
3079: /* for (k=1; k<=cptcovn;k++) */
3080: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
3081: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
3082: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
3083: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
3084: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 3085:
3086:
1.126 brouard 3087: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3088: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3089: /* right multiplication of oldm by the current matrix */
1.126 brouard 3090: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
3091: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 3092: /* if((int)age == 70){ */
3093: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3094: /* for(i=1; i<=nlstate+ndeath; i++) { */
3095: /* printf("%d pmmij ",i); */
3096: /* for(j=1;j<=nlstate+ndeath;j++) { */
3097: /* printf("%f ",pmmij[i][j]); */
3098: /* } */
3099: /* printf(" oldm "); */
3100: /* for(j=1;j<=nlstate+ndeath;j++) { */
3101: /* printf("%f ",oldm[i][j]); */
3102: /* } */
3103: /* printf("\n"); */
3104: /* } */
3105: /* } */
1.126 brouard 3106: savm=oldm;
3107: oldm=newm;
3108: }
3109: for(i=1; i<=nlstate+ndeath; i++)
3110: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 3111: po[i][j][h]=newm[i][j];
3112: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 3113: }
1.128 brouard 3114: /*printf("h=%d ",h);*/
1.126 brouard 3115: } /* end h */
1.218 brouard 3116: /* printf("\n H=%d \n",h); */
1.126 brouard 3117: return po;
3118: }
3119:
1.217 brouard 3120: /************* Higher Back Matrix Product ***************/
1.218 brouard 3121: /* 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 3122: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 3123: {
1.218 brouard 3124: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 3125: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 3126: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3127: nhstepm*hstepm matrices.
3128: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3129: (typically every 2 years instead of every month which is too big
1.217 brouard 3130: for the memory).
1.218 brouard 3131: Model is determined by parameters x and covariates have to be
3132: included manually here.
1.217 brouard 3133:
1.222 brouard 3134: */
1.217 brouard 3135:
3136: int i, j, d, h, k;
3137: double **out, cov[NCOVMAX+1];
3138: double **newm;
3139: double agexact;
3140: double agebegin, ageend;
1.222 brouard 3141: double **oldm, **savm;
1.217 brouard 3142:
1.222 brouard 3143: oldm=oldms;savm=savms;
1.217 brouard 3144: /* Hstepm could be zero and should return the unit matrix */
3145: for (i=1;i<=nlstate+ndeath;i++)
3146: for (j=1;j<=nlstate+ndeath;j++){
3147: oldm[i][j]=(i==j ? 1.0 : 0.0);
3148: po[i][j][0]=(i==j ? 1.0 : 0.0);
3149: }
3150: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3151: for(h=1; h <=nhstepm; h++){
3152: for(d=1; d <=hstepm; d++){
3153: newm=savm;
3154: /* Covariates have to be included here again */
3155: cov[1]=1.;
3156: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3157: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3158: cov[2]=agexact;
3159: if(nagesqr==1)
1.222 brouard 3160: cov[3]= agexact*agexact;
1.218 brouard 3161: for (k=1; k<=cptcovn;k++)
1.222 brouard 3162: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3163: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3164: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3165: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3166: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3167: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3168: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3169: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3170: /* 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 3171:
3172:
1.217 brouard 3173: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3174: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3175: /* Careful transposed matrix */
1.222 brouard 3176: /* age is in cov[2] */
1.218 brouard 3177: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3178: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3179: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3180: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3181: /* if((int)age == 70){ */
3182: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3183: /* for(i=1; i<=nlstate+ndeath; i++) { */
3184: /* printf("%d pmmij ",i); */
3185: /* for(j=1;j<=nlstate+ndeath;j++) { */
3186: /* printf("%f ",pmmij[i][j]); */
3187: /* } */
3188: /* printf(" oldm "); */
3189: /* for(j=1;j<=nlstate+ndeath;j++) { */
3190: /* printf("%f ",oldm[i][j]); */
3191: /* } */
3192: /* printf("\n"); */
3193: /* } */
3194: /* } */
3195: savm=oldm;
3196: oldm=newm;
3197: }
3198: for(i=1; i<=nlstate+ndeath; i++)
3199: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3200: po[i][j][h]=newm[i][j];
3201: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3202: }
3203: /*printf("h=%d ",h);*/
3204: } /* end h */
1.222 brouard 3205: /* printf("\n H=%d \n",h); */
1.217 brouard 3206: return po;
3207: }
3208:
3209:
1.162 brouard 3210: #ifdef NLOPT
3211: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3212: double fret;
3213: double *xt;
3214: int j;
3215: myfunc_data *d2 = (myfunc_data *) pd;
3216: /* xt = (p1-1); */
3217: xt=vector(1,n);
3218: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3219:
3220: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3221: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3222: printf("Function = %.12lf ",fret);
3223: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3224: printf("\n");
3225: free_vector(xt,1,n);
3226: return fret;
3227: }
3228: #endif
1.126 brouard 3229:
3230: /*************** log-likelihood *************/
3231: double func( double *x)
3232: {
1.226 brouard 3233: int i, ii, j, k, mi, d, kk;
3234: int ioffset=0;
3235: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3236: double **out;
3237: double lli; /* Individual log likelihood */
3238: int s1, s2;
1.228 brouard 3239: 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 3240: double bbh, survp;
3241: long ipmx;
3242: double agexact;
3243: /*extern weight */
3244: /* We are differentiating ll according to initial status */
3245: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3246: /*for(i=1;i<imx;i++)
3247: printf(" %d\n",s[4][i]);
3248: */
1.162 brouard 3249:
1.226 brouard 3250: ++countcallfunc;
1.162 brouard 3251:
1.226 brouard 3252: cov[1]=1.;
1.126 brouard 3253:
1.226 brouard 3254: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3255: ioffset=0;
1.226 brouard 3256: if(mle==1){
3257: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3258: /* Computes the values of the ncovmodel covariates of the model
3259: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3260: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3261: to be observed in j being in i according to the model.
3262: */
1.243 brouard 3263: ioffset=2+nagesqr ;
1.233 brouard 3264: /* Fixed */
1.234 brouard 3265: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3266: 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)*/
3267: }
1.226 brouard 3268: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3269: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3270: has been calculated etc */
3271: /* For an individual i, wav[i] gives the number of effective waves */
3272: /* We compute the contribution to Likelihood of each effective transition
3273: mw[mi][i] is real wave of the mi th effectve wave */
3274: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3275: s2=s[mw[mi+1][i]][i];
3276: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3277: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3278: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3279: */
3280: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3281: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3282: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3283: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
1.234 brouard 3284: }
3285: for (ii=1;ii<=nlstate+ndeath;ii++)
3286: for (j=1;j<=nlstate+ndeath;j++){
3287: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3288: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3289: }
3290: for(d=0; d<dh[mi][i]; d++){
3291: newm=savm;
3292: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3293: cov[2]=agexact;
3294: if(nagesqr==1)
3295: cov[3]= agexact*agexact; /* Should be changed here */
3296: for (kk=1; kk<=cptcovage;kk++) {
1.242 brouard 3297: if(!FixedV[Tvar[Tage[kk]]])
1.234 brouard 3298: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
1.242 brouard 3299: else
3300: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
1.234 brouard 3301: }
3302: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3303: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3304: savm=oldm;
3305: oldm=newm;
3306: } /* end mult */
3307:
3308: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3309: /* But now since version 0.9 we anticipate for bias at large stepm.
3310: * If stepm is larger than one month (smallest stepm) and if the exact delay
3311: * (in months) between two waves is not a multiple of stepm, we rounded to
3312: * the nearest (and in case of equal distance, to the lowest) interval but now
3313: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3314: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3315: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3316: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3317: * -stepm/2 to stepm/2 .
3318: * For stepm=1 the results are the same as for previous versions of Imach.
3319: * For stepm > 1 the results are less biased than in previous versions.
3320: */
1.234 brouard 3321: s1=s[mw[mi][i]][i];
3322: s2=s[mw[mi+1][i]][i];
3323: bbh=(double)bh[mi][i]/(double)stepm;
3324: /* bias bh is positive if real duration
3325: * is higher than the multiple of stepm and negative otherwise.
3326: */
3327: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3328: if( s2 > nlstate){
3329: /* i.e. if s2 is a death state and if the date of death is known
3330: then the contribution to the likelihood is the probability to
3331: die between last step unit time and current step unit time,
3332: which is also equal to probability to die before dh
3333: minus probability to die before dh-stepm .
3334: In version up to 0.92 likelihood was computed
3335: as if date of death was unknown. Death was treated as any other
3336: health state: the date of the interview describes the actual state
3337: and not the date of a change in health state. The former idea was
3338: to consider that at each interview the state was recorded
3339: (healthy, disable or death) and IMaCh was corrected; but when we
3340: introduced the exact date of death then we should have modified
3341: the contribution of an exact death to the likelihood. This new
3342: contribution is smaller and very dependent of the step unit
3343: stepm. It is no more the probability to die between last interview
3344: and month of death but the probability to survive from last
3345: interview up to one month before death multiplied by the
3346: probability to die within a month. Thanks to Chris
3347: Jackson for correcting this bug. Former versions increased
3348: mortality artificially. The bad side is that we add another loop
3349: which slows down the processing. The difference can be up to 10%
3350: lower mortality.
3351: */
3352: /* If, at the beginning of the maximization mostly, the
3353: cumulative probability or probability to be dead is
3354: constant (ie = 1) over time d, the difference is equal to
3355: 0. out[s1][3] = savm[s1][3]: probability, being at state
3356: s1 at precedent wave, to be dead a month before current
3357: wave is equal to probability, being at state s1 at
3358: precedent wave, to be dead at mont of the current
3359: wave. Then the observed probability (that this person died)
3360: is null according to current estimated parameter. In fact,
3361: it should be very low but not zero otherwise the log go to
3362: infinity.
3363: */
1.183 brouard 3364: /* #ifdef INFINITYORIGINAL */
3365: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3366: /* #else */
3367: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3368: /* lli=log(mytinydouble); */
3369: /* else */
3370: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3371: /* #endif */
1.226 brouard 3372: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3373:
1.226 brouard 3374: } else if ( s2==-1 ) { /* alive */
3375: for (j=1,survp=0. ; j<=nlstate; j++)
3376: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3377: /*survp += out[s1][j]; */
3378: lli= log(survp);
3379: }
3380: else if (s2==-4) {
3381: for (j=3,survp=0. ; j<=nlstate; j++)
3382: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3383: lli= log(survp);
3384: }
3385: else if (s2==-5) {
3386: for (j=1,survp=0. ; j<=2; j++)
3387: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3388: lli= log(survp);
3389: }
3390: else{
3391: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3392: /* 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 */
3393: }
3394: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3395: /*if(lli ==000.0)*/
3396: /*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); */
3397: ipmx +=1;
3398: sw += weight[i];
3399: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3400: /* if (lli < log(mytinydouble)){ */
3401: /* 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); */
3402: /* 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]); */
3403: /* } */
3404: } /* end of wave */
3405: } /* end of individual */
3406: } else if(mle==2){
3407: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3408: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3409: for(mi=1; mi<= wav[i]-1; mi++){
3410: for (ii=1;ii<=nlstate+ndeath;ii++)
3411: for (j=1;j<=nlstate+ndeath;j++){
3412: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3413: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3414: }
3415: for(d=0; d<=dh[mi][i]; d++){
3416: newm=savm;
3417: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3418: cov[2]=agexact;
3419: if(nagesqr==1)
3420: cov[3]= agexact*agexact;
3421: for (kk=1; kk<=cptcovage;kk++) {
3422: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3423: }
3424: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3425: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3426: savm=oldm;
3427: oldm=newm;
3428: } /* end mult */
3429:
3430: s1=s[mw[mi][i]][i];
3431: s2=s[mw[mi+1][i]][i];
3432: bbh=(double)bh[mi][i]/(double)stepm;
3433: 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 */
3434: ipmx +=1;
3435: sw += weight[i];
3436: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3437: } /* end of wave */
3438: } /* end of individual */
3439: } else if(mle==3){ /* exponential inter-extrapolation */
3440: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3441: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3442: for(mi=1; mi<= wav[i]-1; mi++){
3443: for (ii=1;ii<=nlstate+ndeath;ii++)
3444: for (j=1;j<=nlstate+ndeath;j++){
3445: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3446: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3447: }
3448: for(d=0; d<dh[mi][i]; d++){
3449: newm=savm;
3450: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3451: cov[2]=agexact;
3452: if(nagesqr==1)
3453: cov[3]= agexact*agexact;
3454: for (kk=1; kk<=cptcovage;kk++) {
3455: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3456: }
3457: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3458: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3459: savm=oldm;
3460: oldm=newm;
3461: } /* end mult */
3462:
3463: s1=s[mw[mi][i]][i];
3464: s2=s[mw[mi+1][i]][i];
3465: bbh=(double)bh[mi][i]/(double)stepm;
3466: 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 */
3467: ipmx +=1;
3468: sw += weight[i];
3469: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3470: } /* end of wave */
3471: } /* end of individual */
3472: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3473: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3474: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3475: for(mi=1; mi<= wav[i]-1; mi++){
3476: for (ii=1;ii<=nlstate+ndeath;ii++)
3477: for (j=1;j<=nlstate+ndeath;j++){
3478: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3479: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3480: }
3481: for(d=0; d<dh[mi][i]; d++){
3482: newm=savm;
3483: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3484: cov[2]=agexact;
3485: if(nagesqr==1)
3486: cov[3]= agexact*agexact;
3487: for (kk=1; kk<=cptcovage;kk++) {
3488: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3489: }
1.126 brouard 3490:
1.226 brouard 3491: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3492: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3493: savm=oldm;
3494: oldm=newm;
3495: } /* end mult */
3496:
3497: s1=s[mw[mi][i]][i];
3498: s2=s[mw[mi+1][i]][i];
3499: if( s2 > nlstate){
3500: lli=log(out[s1][s2] - savm[s1][s2]);
3501: } else if ( s2==-1 ) { /* alive */
3502: for (j=1,survp=0. ; j<=nlstate; j++)
3503: survp += out[s1][j];
3504: lli= log(survp);
3505: }else{
3506: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3507: }
3508: ipmx +=1;
3509: sw += weight[i];
3510: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3511: /* 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 3512: } /* end of wave */
3513: } /* end of individual */
3514: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3515: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3516: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3517: for(mi=1; mi<= wav[i]-1; mi++){
3518: for (ii=1;ii<=nlstate+ndeath;ii++)
3519: for (j=1;j<=nlstate+ndeath;j++){
3520: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3521: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3522: }
3523: for(d=0; d<dh[mi][i]; d++){
3524: newm=savm;
3525: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3526: cov[2]=agexact;
3527: if(nagesqr==1)
3528: cov[3]= agexact*agexact;
3529: for (kk=1; kk<=cptcovage;kk++) {
3530: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3531: }
1.126 brouard 3532:
1.226 brouard 3533: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3534: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3535: savm=oldm;
3536: oldm=newm;
3537: } /* end mult */
3538:
3539: s1=s[mw[mi][i]][i];
3540: s2=s[mw[mi+1][i]][i];
3541: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3542: ipmx +=1;
3543: sw += weight[i];
3544: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3545: /*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]);*/
3546: } /* end of wave */
3547: } /* end of individual */
3548: } /* End of if */
3549: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3550: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3551: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3552: return -l;
1.126 brouard 3553: }
3554:
3555: /*************** log-likelihood *************/
3556: double funcone( double *x)
3557: {
1.228 brouard 3558: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3559: int i, ii, j, k, mi, d, kk;
1.228 brouard 3560: int ioffset=0;
1.131 brouard 3561: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3562: double **out;
3563: double lli; /* Individual log likelihood */
3564: double llt;
3565: int s1, s2;
1.228 brouard 3566: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3567:
1.126 brouard 3568: double bbh, survp;
1.187 brouard 3569: double agexact;
1.214 brouard 3570: double agebegin, ageend;
1.126 brouard 3571: /*extern weight */
3572: /* We are differentiating ll according to initial status */
3573: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3574: /*for(i=1;i<imx;i++)
3575: printf(" %d\n",s[4][i]);
3576: */
3577: cov[1]=1.;
3578:
3579: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3580: ioffset=0;
3581: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.243 brouard 3582: /* ioffset=2+nagesqr+cptcovage; */
3583: ioffset=2+nagesqr;
1.232 brouard 3584: /* Fixed */
1.224 brouard 3585: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3586: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3587: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3588: 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)*/
3589: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3590: /* cov[2+6]=covar[Tvar[6]][i]; */
3591: /* cov[2+6]=covar[2][i]; V2 */
3592: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3593: /* cov[2+7]=covar[Tvar[7]][i]; */
3594: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3595: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3596: /* cov[2+9]=covar[Tvar[9]][i]; */
3597: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3598: }
1.232 brouard 3599: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3600: /* 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?)*\/ */
3601: /* } */
1.231 brouard 3602: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3603: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3604: /* } */
1.225 brouard 3605:
1.233 brouard 3606:
3607: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3608: /* Wave varying (but not age varying) */
3609: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3610: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3611: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
3612: }
1.232 brouard 3613: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.242 brouard 3614: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3615: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
3616: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3617: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
3618: /* 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 3619: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
1.242 brouard 3620: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3621: /* /\* 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]); *\/ */
3622: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
1.232 brouard 3623: /* } */
1.126 brouard 3624: for (ii=1;ii<=nlstate+ndeath;ii++)
1.242 brouard 3625: for (j=1;j<=nlstate+ndeath;j++){
3626: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3627: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3628: }
1.214 brouard 3629:
3630: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3631: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3632: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.247 brouard 3633: /* for(d=0; d<=0; d++){ /\* Delay between two effective waves Only one matrix to speed up*\/ */
1.242 brouard 3634: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3635: and mw[mi+1][i]. dh depends on stepm.*/
3636: newm=savm;
1.247 brouard 3637: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM; /* Here d is needed */
1.242 brouard 3638: cov[2]=agexact;
3639: if(nagesqr==1)
3640: cov[3]= agexact*agexact;
3641: for (kk=1; kk<=cptcovage;kk++) {
3642: if(!FixedV[Tvar[Tage[kk]]])
3643: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3644: else
3645: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
3646: }
3647: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3648: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3649: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3650: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3651: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3652: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3653: savm=oldm;
3654: oldm=newm;
1.126 brouard 3655: } /* end mult */
3656:
3657: s1=s[mw[mi][i]][i];
3658: s2=s[mw[mi+1][i]][i];
1.217 brouard 3659: /* if(s2==-1){ */
3660: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3661: /* /\* exit(1); *\/ */
3662: /* } */
1.126 brouard 3663: bbh=(double)bh[mi][i]/(double)stepm;
3664: /* bias is positive if real duration
3665: * is higher than the multiple of stepm and negative otherwise.
3666: */
3667: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.242 brouard 3668: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3669: } else if ( s2==-1 ) { /* alive */
1.242 brouard 3670: for (j=1,survp=0. ; j<=nlstate; j++)
3671: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3672: lli= log(survp);
1.126 brouard 3673: }else if (mle==1){
1.242 brouard 3674: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3675: } else if(mle==2){
1.242 brouard 3676: 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 3677: } else if(mle==3){ /* exponential inter-extrapolation */
1.242 brouard 3678: 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 3679: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.242 brouard 3680: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3681: } else{ /* mle=0 back to 1 */
1.242 brouard 3682: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3683: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3684: } /* End of if */
3685: ipmx +=1;
3686: sw += weight[i];
3687: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3688: /*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 3689: if(globpr){
1.246 brouard 3690: fprintf(ficresilk,"%09ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3691: %11.6f %11.6f %11.6f ", \
1.242 brouard 3692: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3693: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3694: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3695: llt +=ll[k]*gipmx/gsw;
3696: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3697: }
3698: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3699: }
1.232 brouard 3700: } /* end of wave */
3701: } /* end of individual */
3702: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3703: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3704: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3705: if(globpr==0){ /* First time we count the contributions and weights */
3706: gipmx=ipmx;
3707: gsw=sw;
3708: }
3709: return -l;
1.126 brouard 3710: }
3711:
3712:
3713: /*************** function likelione ***********/
3714: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3715: {
3716: /* This routine should help understanding what is done with
3717: the selection of individuals/waves and
3718: to check the exact contribution to the likelihood.
3719: Plotting could be done.
3720: */
3721: int k;
3722:
3723: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3724: strcpy(fileresilk,"ILK_");
1.202 brouard 3725: strcat(fileresilk,fileresu);
1.126 brouard 3726: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3727: printf("Problem with resultfile: %s\n", fileresilk);
3728: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3729: }
1.214 brouard 3730: 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");
3731: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3732: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3733: for(k=1; k<=nlstate; k++)
3734: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3735: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3736: }
3737:
3738: *fretone=(*funcone)(p);
3739: if(*globpri !=0){
3740: fclose(ficresilk);
1.205 brouard 3741: if (mle ==0)
3742: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3743: else if(mle >=1)
3744: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3745: 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 3746:
1.208 brouard 3747:
3748: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3749: 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 3750: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3751: }
1.207 brouard 3752: 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 3753: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3754: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3755: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3756: fflush(fichtm);
1.205 brouard 3757: }
1.126 brouard 3758: return;
3759: }
3760:
3761:
3762: /*********** Maximum Likelihood Estimation ***************/
3763:
3764: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3765: {
1.165 brouard 3766: int i,j, iter=0;
1.126 brouard 3767: double **xi;
3768: double fret;
3769: double fretone; /* Only one call to likelihood */
3770: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3771:
3772: #ifdef NLOPT
3773: int creturn;
3774: nlopt_opt opt;
3775: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3776: double *lb;
3777: double minf; /* the minimum objective value, upon return */
3778: double * p1; /* Shifted parameters from 0 instead of 1 */
3779: myfunc_data dinst, *d = &dinst;
3780: #endif
3781:
3782:
1.126 brouard 3783: xi=matrix(1,npar,1,npar);
3784: for (i=1;i<=npar;i++)
3785: for (j=1;j<=npar;j++)
3786: xi[i][j]=(i==j ? 1.0 : 0.0);
3787: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3788: strcpy(filerespow,"POW_");
1.126 brouard 3789: strcat(filerespow,fileres);
3790: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3791: printf("Problem with resultfile: %s\n", filerespow);
3792: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3793: }
3794: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3795: for (i=1;i<=nlstate;i++)
3796: for(j=1;j<=nlstate+ndeath;j++)
3797: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3798: fprintf(ficrespow,"\n");
1.162 brouard 3799: #ifdef POWELL
1.126 brouard 3800: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3801: #endif
1.126 brouard 3802:
1.162 brouard 3803: #ifdef NLOPT
3804: #ifdef NEWUOA
3805: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3806: #else
3807: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3808: #endif
3809: lb=vector(0,npar-1);
3810: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3811: nlopt_set_lower_bounds(opt, lb);
3812: nlopt_set_initial_step1(opt, 0.1);
3813:
3814: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3815: d->function = func;
3816: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3817: nlopt_set_min_objective(opt, myfunc, d);
3818: nlopt_set_xtol_rel(opt, ftol);
3819: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3820: printf("nlopt failed! %d\n",creturn);
3821: }
3822: else {
3823: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3824: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3825: iter=1; /* not equal */
3826: }
3827: nlopt_destroy(opt);
3828: #endif
1.126 brouard 3829: free_matrix(xi,1,npar,1,npar);
3830: fclose(ficrespow);
1.203 brouard 3831: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3832: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3833: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3834:
3835: }
3836:
3837: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3838: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3839: {
3840: double **a,**y,*x,pd;
1.203 brouard 3841: /* double **hess; */
1.164 brouard 3842: int i, j;
1.126 brouard 3843: int *indx;
3844:
3845: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3846: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3847: void lubksb(double **a, int npar, int *indx, double b[]) ;
3848: void ludcmp(double **a, int npar, int *indx, double *d) ;
3849: double gompertz(double p[]);
1.203 brouard 3850: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3851:
3852: printf("\nCalculation of the hessian matrix. Wait...\n");
3853: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3854: for (i=1;i<=npar;i++){
1.203 brouard 3855: printf("%d-",i);fflush(stdout);
3856: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3857:
3858: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3859:
3860: /* printf(" %f ",p[i]);
3861: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3862: }
3863:
3864: for (i=1;i<=npar;i++) {
3865: for (j=1;j<=npar;j++) {
3866: if (j>i) {
1.203 brouard 3867: printf(".%d-%d",i,j);fflush(stdout);
3868: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3869: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3870:
3871: hess[j][i]=hess[i][j];
3872: /*printf(" %lf ",hess[i][j]);*/
3873: }
3874: }
3875: }
3876: printf("\n");
3877: fprintf(ficlog,"\n");
3878:
3879: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3880: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3881:
3882: a=matrix(1,npar,1,npar);
3883: y=matrix(1,npar,1,npar);
3884: x=vector(1,npar);
3885: indx=ivector(1,npar);
3886: for (i=1;i<=npar;i++)
3887: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3888: ludcmp(a,npar,indx,&pd);
3889:
3890: for (j=1;j<=npar;j++) {
3891: for (i=1;i<=npar;i++) x[i]=0;
3892: x[j]=1;
3893: lubksb(a,npar,indx,x);
3894: for (i=1;i<=npar;i++){
3895: matcov[i][j]=x[i];
3896: }
3897: }
3898:
3899: printf("\n#Hessian matrix#\n");
3900: fprintf(ficlog,"\n#Hessian matrix#\n");
3901: for (i=1;i<=npar;i++) {
3902: for (j=1;j<=npar;j++) {
1.203 brouard 3903: printf("%.6e ",hess[i][j]);
3904: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3905: }
3906: printf("\n");
3907: fprintf(ficlog,"\n");
3908: }
3909:
1.203 brouard 3910: /* printf("\n#Covariance matrix#\n"); */
3911: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3912: /* for (i=1;i<=npar;i++) { */
3913: /* for (j=1;j<=npar;j++) { */
3914: /* printf("%.6e ",matcov[i][j]); */
3915: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3916: /* } */
3917: /* printf("\n"); */
3918: /* fprintf(ficlog,"\n"); */
3919: /* } */
3920:
1.126 brouard 3921: /* Recompute Inverse */
1.203 brouard 3922: /* for (i=1;i<=npar;i++) */
3923: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3924: /* ludcmp(a,npar,indx,&pd); */
3925:
3926: /* printf("\n#Hessian matrix recomputed#\n"); */
3927:
3928: /* for (j=1;j<=npar;j++) { */
3929: /* for (i=1;i<=npar;i++) x[i]=0; */
3930: /* x[j]=1; */
3931: /* lubksb(a,npar,indx,x); */
3932: /* for (i=1;i<=npar;i++){ */
3933: /* y[i][j]=x[i]; */
3934: /* printf("%.3e ",y[i][j]); */
3935: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3936: /* } */
3937: /* printf("\n"); */
3938: /* fprintf(ficlog,"\n"); */
3939: /* } */
3940:
3941: /* Verifying the inverse matrix */
3942: #ifdef DEBUGHESS
3943: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3944:
1.203 brouard 3945: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3946: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3947:
3948: for (j=1;j<=npar;j++) {
3949: for (i=1;i<=npar;i++){
1.203 brouard 3950: printf("%.2f ",y[i][j]);
3951: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3952: }
3953: printf("\n");
3954: fprintf(ficlog,"\n");
3955: }
1.203 brouard 3956: #endif
1.126 brouard 3957:
3958: free_matrix(a,1,npar,1,npar);
3959: free_matrix(y,1,npar,1,npar);
3960: free_vector(x,1,npar);
3961: free_ivector(indx,1,npar);
1.203 brouard 3962: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3963:
3964:
3965: }
3966:
3967: /*************** hessian matrix ****************/
3968: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3969: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3970: int i;
3971: int l=1, lmax=20;
1.203 brouard 3972: double k1,k2, res, fx;
1.132 brouard 3973: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3974: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3975: int k=0,kmax=10;
3976: double l1;
3977:
3978: fx=func(x);
3979: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3980: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3981: l1=pow(10,l);
3982: delts=delt;
3983: for(k=1 ; k <kmax; k=k+1){
3984: delt = delta*(l1*k);
3985: p2[theta]=x[theta] +delt;
1.145 brouard 3986: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 3987: p2[theta]=x[theta]-delt;
3988: k2=func(p2)-fx;
3989: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 3990: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 3991:
1.203 brouard 3992: #ifdef DEBUGHESSII
1.126 brouard 3993: 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);
3994: 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);
3995: #endif
3996: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
3997: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
3998: k=kmax;
3999: }
4000: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 4001: k=kmax; l=lmax*10;
1.126 brouard 4002: }
4003: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
4004: delts=delt;
4005: }
1.203 brouard 4006: } /* End loop k */
1.126 brouard 4007: }
4008: delti[theta]=delts;
4009: return res;
4010:
4011: }
4012:
1.203 brouard 4013: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 4014: {
4015: int i;
1.164 brouard 4016: int l=1, lmax=20;
1.126 brouard 4017: double k1,k2,k3,k4,res,fx;
1.132 brouard 4018: double p2[MAXPARM+1];
1.203 brouard 4019: int k, kmax=1;
4020: double v1, v2, cv12, lc1, lc2;
1.208 brouard 4021:
4022: int firstime=0;
1.203 brouard 4023:
1.126 brouard 4024: fx=func(x);
1.203 brouard 4025: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 4026: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 4027: p2[thetai]=x[thetai]+delti[thetai]*k;
4028: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4029: k1=func(p2)-fx;
4030:
1.203 brouard 4031: p2[thetai]=x[thetai]+delti[thetai]*k;
4032: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4033: k2=func(p2)-fx;
4034:
1.203 brouard 4035: p2[thetai]=x[thetai]-delti[thetai]*k;
4036: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4037: k3=func(p2)-fx;
4038:
1.203 brouard 4039: p2[thetai]=x[thetai]-delti[thetai]*k;
4040: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4041: k4=func(p2)-fx;
1.203 brouard 4042: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
4043: if(k1*k2*k3*k4 <0.){
1.208 brouard 4044: firstime=1;
1.203 brouard 4045: kmax=kmax+10;
1.208 brouard 4046: }
4047: if(kmax >=10 || firstime ==1){
1.246 brouard 4048: 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);
4049: 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 4050: 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);
4051: 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);
4052: }
4053: #ifdef DEBUGHESSIJ
4054: v1=hess[thetai][thetai];
4055: v2=hess[thetaj][thetaj];
4056: cv12=res;
4057: /* Computing eigen value of Hessian matrix */
4058: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4059: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4060: if ((lc2 <0) || (lc1 <0) ){
4061: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4062: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4063: 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);
4064: 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);
4065: }
1.126 brouard 4066: #endif
4067: }
4068: return res;
4069: }
4070:
1.203 brouard 4071: /* Not done yet: Was supposed to fix if not exactly at the maximum */
4072: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
4073: /* { */
4074: /* int i; */
4075: /* int l=1, lmax=20; */
4076: /* double k1,k2,k3,k4,res,fx; */
4077: /* double p2[MAXPARM+1]; */
4078: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
4079: /* int k=0,kmax=10; */
4080: /* double l1; */
4081:
4082: /* fx=func(x); */
4083: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
4084: /* l1=pow(10,l); */
4085: /* delts=delt; */
4086: /* for(k=1 ; k <kmax; k=k+1){ */
4087: /* delt = delti*(l1*k); */
4088: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
4089: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4090: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4091: /* k1=func(p2)-fx; */
4092:
4093: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4094: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4095: /* k2=func(p2)-fx; */
4096:
4097: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4098: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4099: /* k3=func(p2)-fx; */
4100:
4101: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4102: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4103: /* k4=func(p2)-fx; */
4104: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
4105: /* #ifdef DEBUGHESSIJ */
4106: /* 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); */
4107: /* 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); */
4108: /* #endif */
4109: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
4110: /* k=kmax; */
4111: /* } */
4112: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
4113: /* k=kmax; l=lmax*10; */
4114: /* } */
4115: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
4116: /* delts=delt; */
4117: /* } */
4118: /* } /\* End loop k *\/ */
4119: /* } */
4120: /* delti[theta]=delts; */
4121: /* return res; */
4122: /* } */
4123:
4124:
1.126 brouard 4125: /************** Inverse of matrix **************/
4126: void ludcmp(double **a, int n, int *indx, double *d)
4127: {
4128: int i,imax,j,k;
4129: double big,dum,sum,temp;
4130: double *vv;
4131:
4132: vv=vector(1,n);
4133: *d=1.0;
4134: for (i=1;i<=n;i++) {
4135: big=0.0;
4136: for (j=1;j<=n;j++)
4137: if ((temp=fabs(a[i][j])) > big) big=temp;
1.256 brouard 4138: if (big == 0.0){
4139: printf(" Singular Hessian matrix at row %d:\n",i);
4140: for (j=1;j<=n;j++) {
4141: printf(" a[%d][%d]=%f,",i,j,a[i][j]);
4142: fprintf(ficlog," a[%d][%d]=%f,",i,j,a[i][j]);
4143: }
4144: fflush(ficlog);
4145: fclose(ficlog);
4146: nrerror("Singular matrix in routine ludcmp");
4147: }
1.126 brouard 4148: vv[i]=1.0/big;
4149: }
4150: for (j=1;j<=n;j++) {
4151: for (i=1;i<j;i++) {
4152: sum=a[i][j];
4153: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
4154: a[i][j]=sum;
4155: }
4156: big=0.0;
4157: for (i=j;i<=n;i++) {
4158: sum=a[i][j];
4159: for (k=1;k<j;k++)
4160: sum -= a[i][k]*a[k][j];
4161: a[i][j]=sum;
4162: if ( (dum=vv[i]*fabs(sum)) >= big) {
4163: big=dum;
4164: imax=i;
4165: }
4166: }
4167: if (j != imax) {
4168: for (k=1;k<=n;k++) {
4169: dum=a[imax][k];
4170: a[imax][k]=a[j][k];
4171: a[j][k]=dum;
4172: }
4173: *d = -(*d);
4174: vv[imax]=vv[j];
4175: }
4176: indx[j]=imax;
4177: if (a[j][j] == 0.0) a[j][j]=TINY;
4178: if (j != n) {
4179: dum=1.0/(a[j][j]);
4180: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4181: }
4182: }
4183: free_vector(vv,1,n); /* Doesn't work */
4184: ;
4185: }
4186:
4187: void lubksb(double **a, int n, int *indx, double b[])
4188: {
4189: int i,ii=0,ip,j;
4190: double sum;
4191:
4192: for (i=1;i<=n;i++) {
4193: ip=indx[i];
4194: sum=b[ip];
4195: b[ip]=b[i];
4196: if (ii)
4197: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4198: else if (sum) ii=i;
4199: b[i]=sum;
4200: }
4201: for (i=n;i>=1;i--) {
4202: sum=b[i];
4203: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4204: b[i]=sum/a[i][i];
4205: }
4206: }
4207:
4208: void pstamp(FILE *fichier)
4209: {
1.196 brouard 4210: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4211: }
4212:
1.253 brouard 4213: int linreg(int ifi, int ila, int *no, const double x[], const double y[], double* a, double* b, double* r, double* sa, double * sb) {
4214:
4215: /* y=a+bx regression */
4216: double sumx = 0.0; /* sum of x */
4217: double sumx2 = 0.0; /* sum of x**2 */
4218: double sumxy = 0.0; /* sum of x * y */
4219: double sumy = 0.0; /* sum of y */
4220: double sumy2 = 0.0; /* sum of y**2 */
4221: double sume2; /* sum of square or residuals */
4222: double yhat;
4223:
4224: double denom=0;
4225: int i;
4226: int ne=*no;
4227:
4228: for ( i=ifi, ne=0;i<=ila;i++) {
4229: if(!isfinite(x[i]) || !isfinite(y[i])){
4230: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4231: continue;
4232: }
4233: ne=ne+1;
4234: sumx += x[i];
4235: sumx2 += x[i]*x[i];
4236: sumxy += x[i] * y[i];
4237: sumy += y[i];
4238: sumy2 += y[i]*y[i];
4239: denom = (ne * sumx2 - sumx*sumx);
4240: /* 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); */
4241: }
4242:
4243: denom = (ne * sumx2 - sumx*sumx);
4244: if (denom == 0) {
4245: // vertical, slope m is infinity
4246: *b = INFINITY;
4247: *a = 0;
4248: if (r) *r = 0;
4249: return 1;
4250: }
4251:
4252: *b = (ne * sumxy - sumx * sumy) / denom;
4253: *a = (sumy * sumx2 - sumx * sumxy) / denom;
4254: if (r!=NULL) {
4255: *r = (sumxy - sumx * sumy / ne) / /* compute correlation coeff */
4256: sqrt((sumx2 - sumx*sumx/ne) *
4257: (sumy2 - sumy*sumy/ne));
4258: }
4259: *no=ne;
4260: for ( i=ifi, ne=0;i<=ila;i++) {
4261: if(!isfinite(x[i]) || !isfinite(y[i])){
4262: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4263: continue;
4264: }
4265: ne=ne+1;
4266: yhat = y[i] - *a -*b* x[i];
4267: sume2 += yhat * yhat ;
4268:
4269: denom = (ne * sumx2 - sumx*sumx);
4270: /* 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); */
4271: }
4272: *sb = sqrt(sume2/(ne-2)/(sumx2 - sumx * sumx /ne));
4273: *sa= *sb * sqrt(sumx2/ne);
4274:
4275: return 0;
4276: }
4277:
1.126 brouard 4278: /************ Frequencies ********************/
1.251 brouard 4279: void freqsummary(char fileres[], double p[], double pstart[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
1.226 brouard 4280: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4281: int firstpass, int lastpass, int stepm, int weightopt, char model[])
1.250 brouard 4282: { /* Some frequencies as well as proposing some starting values */
1.226 brouard 4283:
1.253 brouard 4284: int i, m, jk, j1, bool, z1,j, nj, nl, k, iv, jj=0;
1.226 brouard 4285: int iind=0, iage=0;
4286: int mi; /* Effective wave */
4287: int first;
4288: double ***freq; /* Frequencies */
1.253 brouard 4289: double *x, *y, a,b,r, sa, sb; /* for regression, y=b+m*x and r is the correlation coefficient */
4290: int no;
1.226 brouard 4291: double *meanq;
4292: double **meanqt;
4293: double *pp, **prop, *posprop, *pospropt;
4294: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4295: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4296: double agebegin, ageend;
4297:
4298: pp=vector(1,nlstate);
1.251 brouard 4299: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4300: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4301: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4302: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4303: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4304: meanqt=matrix(1,lastpass,1,nqtveff);
4305: strcpy(fileresp,"P_");
4306: strcat(fileresp,fileresu);
4307: /*strcat(fileresphtm,fileresu);*/
4308: if((ficresp=fopen(fileresp,"w"))==NULL) {
4309: printf("Problem with prevalence resultfile: %s\n", fileresp);
4310: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4311: exit(0);
4312: }
1.240 brouard 4313:
1.226 brouard 4314: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4315: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4316: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4317: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4318: fflush(ficlog);
4319: exit(70);
4320: }
4321: else{
4322: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4323: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4324: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4325: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4326: }
1.237 brouard 4327: 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 4328:
1.226 brouard 4329: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4330: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4331: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4332: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4333: fflush(ficlog);
4334: exit(70);
1.240 brouard 4335: } else{
1.226 brouard 4336: 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 4337: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4338: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4339: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4340: }
1.240 brouard 4341: 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);
4342:
1.253 brouard 4343: y= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
4344: x= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.251 brouard 4345: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4346: j1=0;
1.126 brouard 4347:
1.227 brouard 4348: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4349: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4350: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.240 brouard 4351:
4352:
1.226 brouard 4353: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4354: reference=low_education V1=0,V2=0
4355: med_educ V1=1 V2=0,
4356: high_educ V1=0 V2=1
4357: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4358: */
1.249 brouard 4359: dateintsum=0;
4360: k2cpt=0;
4361:
1.253 brouard 4362: if(cptcoveff == 0 )
4363: nl=1; /* Constant model only */
4364: else
4365: nl=2;
4366: for (nj = 1; nj <= nl; nj++){ /* nj= 1 constant model, nl number of loops. */
4367: if(nj==1)
4368: j=0; /* First pass for the constant */
4369: else
4370: j=cptcoveff; /* Other passes for the covariate values */
1.251 brouard 4371: first=1;
4372: 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 */
4373: posproptt=0.;
4374: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4375: scanf("%d", i);*/
4376: for (i=-5; i<=nlstate+ndeath; i++)
4377: for (jk=-5; jk<=nlstate+ndeath; jk++)
4378: for(m=iagemin; m <= iagemax+3; m++)
4379: freq[i][jk][m]=0;
4380:
4381: for (i=1; i<=nlstate; i++) {
1.240 brouard 4382: for(m=iagemin; m <= iagemax+3; m++)
1.251 brouard 4383: prop[i][m]=0;
4384: posprop[i]=0;
4385: pospropt[i]=0;
4386: }
4387: /* for (z1=1; z1<= nqfveff; z1++) { */
4388: /* meanq[z1]+=0.; */
4389: /* for(m=1;m<=lastpass;m++){ */
4390: /* meanqt[m][z1]=0.; */
4391: /* } */
4392: /* } */
4393:
4394: /* dateintsum=0; */
4395: /* k2cpt=0; */
4396:
4397: /* For that combination of covariate j1, we count and print the frequencies in one pass */
4398: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4399: bool=1;
4400: if(j !=0){
4401: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
4402: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
4403: /* for (z1=1; z1<= nqfveff; z1++) { */
4404: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4405: /* } */
4406: for (z1=1; z1<=cptcoveff; z1++) { /* loops on covariates in the model */
4407: /* if(Tvaraff[z1] ==-20){ */
4408: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4409: /* }else if(Tvaraff[z1] ==-10){ */
4410: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4411: /* }else */
4412: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){ /* for combination j1 of covariates */
4413: /* Tests if this individual iind responded to combination j1 (V4=1 V3=0) */
4414: bool=0; /* bool should be equal to 1 to be selected, one covariate value failed */
4415: /* 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",
4416: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4417: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4418: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4419: } /* Onlyf fixed */
4420: } /* end z1 */
4421: } /* cptcovn > 0 */
4422: } /* end any */
4423: }/* end j==0 */
4424: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
4425: /* for(m=firstpass; m<=lastpass; m++){ */
4426: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4427: m=mw[mi][iind];
4428: if(j!=0){
4429: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4430: for (z1=1; z1<=cptcoveff; z1++) {
4431: if( Fixed[Tmodelind[z1]]==1){
4432: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4433: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality. If covariate's
4434: value is -1, we don't select. It differs from the
4435: constant and age model which counts them. */
4436: bool=0; /* not selected */
4437: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4438: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4439: bool=0;
4440: }
4441: }
4442: }
4443: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4444: } /* end j==0 */
4445: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4446: if(bool==1){
4447: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4448: and mw[mi+1][iind]. dh depends on stepm. */
4449: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4450: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4451: if(m >=firstpass && m <=lastpass){
4452: k2=anint[m][iind]+(mint[m][iind]/12.);
4453: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4454: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4455: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4456: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4457: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4458: if (m<lastpass) {
4459: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4460: /* 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]); */
4461: if(s[m][iind]==-1)
4462: 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.));
4463: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4464: /* if((int)agev[m][iind] == 55) */
4465: /* printf("j=%d, j1=%d Age %d, iind=%d, num=%09ld m=%d\n",j,j1,(int)agev[m][iind],iind, num[iind],m); */
4466: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4467: 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 4468: }
1.251 brouard 4469: } /* end if between passes */
4470: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99) && (j==0)) {
4471: dateintsum=dateintsum+k2; /* on all covariates ?*/
4472: k2cpt++;
4473: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
1.234 brouard 4474: }
1.251 brouard 4475: }else{
4476: bool=1;
4477: }/* end bool 2 */
4478: } /* end m */
4479: } /* end bool */
4480: } /* end iind = 1 to imx */
4481: /* prop[s][age] is feeded for any initial and valid live state as well as
4482: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
4483:
4484:
4485: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4486: pstamp(ficresp);
4487: if (cptcoveff>0 && j!=0){
4488: printf( "\n#********** Variable ");
4489: fprintf(ficresp, "\n#********** Variable ");
4490: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4491: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
4492: fprintf(ficlog, "\n#********** Variable ");
4493: for (z1=1; z1<=cptcoveff; z1++){
4494: if(!FixedV[Tvaraff[z1]]){
4495: printf( "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4496: fprintf(ficresp, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4497: fprintf(ficresphtm, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4498: fprintf(ficresphtmfr, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4499: fprintf(ficlog, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.250 brouard 4500: }else{
1.251 brouard 4501: printf( "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4502: fprintf(ficresp, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4503: fprintf(ficresphtm, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4504: fprintf(ficresphtmfr, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4505: fprintf(ficlog, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4506: }
4507: }
4508: printf( "**********\n#");
4509: fprintf(ficresp, "**********\n#");
4510: fprintf(ficresphtm, "**********</h3>\n");
4511: fprintf(ficresphtmfr, "**********</h3>\n");
4512: fprintf(ficlog, "**********\n");
4513: }
4514: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4515: for(i=1; i<=nlstate;i++) {
4516: fprintf(ficresp, " Age Prev(%d) N(%d) N ",i,i);
4517: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4518: }
4519: fprintf(ficresp, "\n");
4520: fprintf(ficresphtm, "\n");
4521:
4522: /* Header of frequency table by age */
4523: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4524: fprintf(ficresphtmfr,"<th>Age</th> ");
4525: for(jk=-1; jk <=nlstate+ndeath; jk++){
4526: for(m=-1; m <=nlstate+ndeath; m++){
4527: if(jk!=0 && m!=0)
4528: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.240 brouard 4529: }
1.226 brouard 4530: }
1.251 brouard 4531: fprintf(ficresphtmfr, "\n");
4532:
4533: /* For each age */
4534: for(iage=iagemin; iage <= iagemax+3; iage++){
4535: fprintf(ficresphtm,"<tr>");
4536: if(iage==iagemax+1){
4537: fprintf(ficlog,"1");
4538: fprintf(ficresphtmfr,"<tr><th>0</th> ");
4539: }else if(iage==iagemax+2){
4540: fprintf(ficlog,"0");
4541: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
4542: }else if(iage==iagemax+3){
4543: fprintf(ficlog,"Total");
4544: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
4545: }else{
1.240 brouard 4546: if(first==1){
1.251 brouard 4547: first=0;
4548: printf("See log file for details...\n");
4549: }
4550: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4551: fprintf(ficlog,"Age %d", iage);
4552: }
4553: for(jk=1; jk <=nlstate ; jk++){
4554: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4555: pp[jk] += freq[jk][m][iage];
4556: }
4557: for(jk=1; jk <=nlstate ; jk++){
4558: for(m=-1, pos=0; m <=0 ; m++)
4559: pos += freq[jk][m][iage];
4560: if(pp[jk]>=1.e-10){
4561: if(first==1){
4562: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4563: }
4564: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4565: }else{
4566: if(first==1)
4567: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4568: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
1.240 brouard 4569: }
4570: }
4571:
1.251 brouard 4572: for(jk=1; jk <=nlstate ; jk++){
4573: /* posprop[jk]=0; */
4574: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4575: pp[jk] += freq[jk][m][iage];
4576: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
4577:
4578: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
4579: pos += pp[jk]; /* pos is the total number of transitions until this age */
4580: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4581: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4582: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
1.240 brouard 4583: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4584: }
1.251 brouard 4585: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4586: if(pos>=1.e-5){
1.251 brouard 4587: if(first==1)
4588: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4589: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4590: }else{
4591: if(first==1)
4592: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4593: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4594: }
4595: if( iage <= iagemax){
4596: if(pos>=1.e-5){
4597: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4598: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4599: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4600: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4601: }
4602: else{
4603: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4604: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4605: }
1.240 brouard 4606: }
1.251 brouard 4607: pospropt[jk] +=posprop[jk];
4608: } /* end loop jk */
4609: /* pospropt=0.; */
4610: for(jk=-1; jk <=nlstate+ndeath; jk++){
4611: for(m=-1; m <=nlstate+ndeath; m++){
4612: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4613: if(first==1){
4614: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4615: }
1.253 brouard 4616: /* printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]); */
1.251 brouard 4617: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4618: }
4619: if(jk!=0 && m!=0)
4620: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
1.240 brouard 4621: }
1.251 brouard 4622: } /* end loop jk */
4623: posproptt=0.;
4624: for(jk=1; jk <=nlstate; jk++){
4625: posproptt += pospropt[jk];
4626: }
4627: fprintf(ficresphtmfr,"</tr>\n ");
4628: if(iage <= iagemax){
4629: fprintf(ficresp,"\n");
4630: fprintf(ficresphtm,"</tr>\n");
1.240 brouard 4631: }
1.251 brouard 4632: if(first==1)
4633: printf("Others in log...\n");
4634: fprintf(ficlog,"\n");
4635: } /* end loop age iage */
4636: fprintf(ficresphtm,"<tr><th>Tot</th>");
4637: for(jk=1; jk <=nlstate ; jk++){
4638: if(posproptt < 1.e-5){
4639: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
4640: }else{
4641: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.240 brouard 4642: }
1.226 brouard 4643: }
1.251 brouard 4644: fprintf(ficresphtm,"</tr>\n");
4645: fprintf(ficresphtm,"</table>\n");
4646: fprintf(ficresphtmfr,"</table>\n");
1.226 brouard 4647: if(posproptt < 1.e-5){
1.251 brouard 4648: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4649: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4650: fprintf(ficres,"\n This combination (%d) is not valid and no result will be produced\n\n",j1);
4651: invalidvarcomb[j1]=1;
1.226 brouard 4652: }else{
1.251 brouard 4653: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4654: invalidvarcomb[j1]=0;
1.226 brouard 4655: }
1.251 brouard 4656: fprintf(ficresphtmfr,"</table>\n");
4657: fprintf(ficlog,"\n");
4658: if(j!=0){
4659: printf("#Freqsummary: Starting values for combination j1=%d:\n", j1);
4660: for(i=1,jk=1; i <=nlstate; i++){
4661: for(k=1; k <=(nlstate+ndeath); k++){
4662: if (k != i) {
4663: for(jj=1; jj <=ncovmodel; jj++){ /* For counting jk */
1.253 brouard 4664: if(jj==1){ /* Constant case (in fact cste + age) */
1.251 brouard 4665: if(j1==1){ /* All dummy covariates to zero */
4666: freq[i][k][iagemax+4]=freq[i][k][iagemax+3]; /* Stores case 0 0 0 */
4667: freq[i][i][iagemax+4]=freq[i][i][iagemax+3]; /* Stores case 0 0 0 */
1.252 brouard 4668: printf("%d%d ",i,k);
4669: fprintf(ficlog,"%d%d ",i,k);
4670: 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]));
4671: 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]));
4672: pstart[jk]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
1.251 brouard 4673: }
1.253 brouard 4674: }else if((j1==1) && (jj==2 || nagesqr==1)){ /* age or age*age parameter without covariate V4*age (to be done later) */
4675: for(iage=iagemin; iage <= iagemax+3; iage++){
4676: x[iage]= (double)iage;
4677: y[iage]= log(freq[i][k][iage]/freq[i][i][iage]);
4678: /* printf("i=%d, k=%d, jk=%d, j1=%d, jj=%d, y[%d]=%f\n",i,k,jk,j1,jj, iage, y[iage]); */
4679: }
4680: linreg(iagemin,iagemax,&no,x,y,&a,&b,&r, &sa, &sb ); /* y= a+b*x with standard errors */
4681: pstart[jk]=b;
4682: pstart[jk-1]=a;
1.252 brouard 4683: }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 */
4684: 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]);
4685: 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 4686: 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 4687: printf("%d%d ",i,k);
4688: fprintf(ficlog,"%d%d ",i,k);
1.251 brouard 4689: 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]));
4690: }else{ /* Other cases, like quantitative fixed or varying covariates */
4691: ;
4692: }
4693: /* printf("%12.7f )", param[i][jj][k]); */
4694: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
4695: jk++;
4696: } /* end jj */
4697: } /* end k!= i */
4698: } /* end k */
4699: } /* end i, jk */
4700: } /* end j !=0 */
4701: } /* end selected combination of covariate j1 */
4702: if(j==0){ /* We can estimate starting values from the occurences in each case */
4703: printf("#Freqsummary: Starting values for the constants:\n");
4704: fprintf(ficlog,"\n");
4705: for(i=1,jk=1; i <=nlstate; i++){
4706: for(k=1; k <=(nlstate+ndeath); k++){
4707: if (k != i) {
4708: printf("%d%d ",i,k);
4709: fprintf(ficlog,"%d%d ",i,k);
4710: for(jj=1; jj <=ncovmodel; jj++){
1.253 brouard 4711: pstart[jk]=p[jk]; /* Setting pstart to p values by default */
4712: if(jj==1){ /* Age has to be done */
4713: pstart[jk]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
1.251 brouard 4714: 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]));
4715: 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]));
4716: }
4717: /* printf("%12.7f )", param[i][jj][k]); */
4718: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
4719: jk++;
1.250 brouard 4720: }
1.251 brouard 4721: printf("\n");
4722: fprintf(ficlog,"\n");
1.250 brouard 4723: }
4724: }
4725: }
1.251 brouard 4726: printf("#Freqsummary\n");
4727: fprintf(ficlog,"\n");
4728: for(jk=-1; jk <=nlstate+ndeath; jk++){
4729: for(m=-1; m <=nlstate+ndeath; m++){
4730: /* param[i]|j][k]= freq[jk][m][iagemax+3] */
1.250 brouard 4731: printf(" %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]);
4732: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]);
1.251 brouard 4733: /* if(freq[jk][m][iage] !=0 ) { /\* minimizing output *\/ */
4734: /* printf(" %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]); */
4735: /* fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iagemax+3]); */
4736: /* } */
4737: }
4738: } /* end loop jk */
4739:
4740: printf("\n");
4741: fprintf(ficlog,"\n");
4742: } /* end j=0 */
1.249 brouard 4743: } /* end j */
1.252 brouard 4744:
1.253 brouard 4745: if(mle == -2){ /* We want to use these values as starting values */
1.252 brouard 4746: for(i=1, jk=1; i <=nlstate; i++){
4747: for(j=1; j <=nlstate+ndeath; j++){
4748: if(j!=i){
4749: /*ca[0]= k+'a'-1;ca[1]='\0';*/
4750: printf("%1d%1d",i,j);
4751: fprintf(ficparo,"%1d%1d",i,j);
4752: for(k=1; k<=ncovmodel;k++){
4753: /* printf(" %lf",param[i][j][k]); */
4754: /* fprintf(ficparo," %lf",param[i][j][k]); */
4755: p[jk]=pstart[jk];
4756: printf(" %f ",pstart[jk]);
4757: fprintf(ficparo," %f ",pstart[jk]);
4758: jk++;
4759: }
4760: printf("\n");
4761: fprintf(ficparo,"\n");
4762: }
4763: }
4764: }
4765: } /* end mle=-2 */
1.226 brouard 4766: dateintmean=dateintsum/k2cpt;
1.240 brouard 4767:
1.226 brouard 4768: fclose(ficresp);
4769: fclose(ficresphtm);
4770: fclose(ficresphtmfr);
4771: free_vector(meanq,1,nqfveff);
4772: free_matrix(meanqt,1,lastpass,1,nqtveff);
1.253 brouard 4773: free_vector(x, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
4774: free_vector(y, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.251 brouard 4775: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4776: free_vector(pospropt,1,nlstate);
4777: free_vector(posprop,1,nlstate);
1.251 brouard 4778: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4779: free_vector(pp,1,nlstate);
4780: /* End of freqsummary */
4781: }
1.126 brouard 4782:
4783: /************ Prevalence ********************/
1.227 brouard 4784: 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)
4785: {
4786: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4787: in each health status at the date of interview (if between dateprev1 and dateprev2).
4788: We still use firstpass and lastpass as another selection.
4789: */
1.126 brouard 4790:
1.227 brouard 4791: int i, m, jk, j1, bool, z1,j, iv;
4792: int mi; /* Effective wave */
4793: int iage;
4794: double agebegin, ageend;
4795:
4796: double **prop;
4797: double posprop;
4798: double y2; /* in fractional years */
4799: int iagemin, iagemax;
4800: int first; /** to stop verbosity which is redirected to log file */
4801:
4802: iagemin= (int) agemin;
4803: iagemax= (int) agemax;
4804: /*pp=vector(1,nlstate);*/
1.251 brouard 4805: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4806: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4807: j1=0;
1.222 brouard 4808:
1.227 brouard 4809: /*j=cptcoveff;*/
4810: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4811:
1.227 brouard 4812: first=1;
4813: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4814: for (i=1; i<=nlstate; i++)
1.251 brouard 4815: for(iage=iagemin-AGEMARGE; iage <= iagemax+4+AGEMARGE; iage++)
1.227 brouard 4816: prop[i][iage]=0.0;
4817: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4818: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4819: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4820:
4821: for (i=1; i<=imx; i++) { /* Each individual */
4822: bool=1;
4823: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4824: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4825: m=mw[mi][i];
4826: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4827: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4828: for (z1=1; z1<=cptcoveff; z1++){
4829: if( Fixed[Tmodelind[z1]]==1){
4830: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4831: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4832: bool=0;
4833: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4834: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4835: bool=0;
4836: }
4837: }
4838: if(bool==1){ /* Otherwise we skip that wave/person */
4839: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4840: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4841: if(m >=firstpass && m <=lastpass){
4842: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4843: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4844: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4845: if(agev[m][i]==1) agev[m][i]=iagemax+2;
1.251 brouard 4846: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+4+AGEMARGE){
1.227 brouard 4847: 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);
4848: exit(1);
4849: }
4850: if (s[m][i]>0 && s[m][i]<=nlstate) {
4851: /*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]]);*/
4852: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4853: prop[s[m][i]][iagemax+3] += weight[i];
4854: } /* end valid statuses */
4855: } /* end selection of dates */
4856: } /* end selection of waves */
4857: } /* end bool */
4858: } /* end wave */
4859: } /* end individual */
4860: for(i=iagemin; i <= iagemax+3; i++){
4861: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4862: posprop += prop[jk][i];
4863: }
4864:
4865: for(jk=1; jk <=nlstate ; jk++){
4866: if( i <= iagemax){
4867: if(posprop>=1.e-5){
4868: probs[i][jk][j1]= prop[jk][i]/posprop;
4869: } else{
4870: if(first==1){
4871: first=0;
4872: 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]);
4873: }
4874: }
4875: }
4876: }/* end jk */
4877: }/* end i */
1.222 brouard 4878: /*} *//* end i1 */
1.227 brouard 4879: } /* end j1 */
1.222 brouard 4880:
1.227 brouard 4881: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4882: /*free_vector(pp,1,nlstate);*/
1.251 brouard 4883: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4884: } /* End of prevalence */
1.126 brouard 4885:
4886: /************* Waves Concatenation ***************/
4887:
4888: 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)
4889: {
4890: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4891: Death is a valid wave (if date is known).
4892: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4893: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4894: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4895: */
1.126 brouard 4896:
1.224 brouard 4897: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4898: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4899: double sum=0., jmean=0.;*/
1.224 brouard 4900: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4901: int j, k=0,jk, ju, jl;
4902: double sum=0.;
4903: first=0;
1.214 brouard 4904: firstwo=0;
1.217 brouard 4905: firsthree=0;
1.218 brouard 4906: firstfour=0;
1.164 brouard 4907: jmin=100000;
1.126 brouard 4908: jmax=-1;
4909: jmean=0.;
1.224 brouard 4910:
4911: /* Treating live states */
1.214 brouard 4912: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4913: mi=0; /* First valid wave */
1.227 brouard 4914: mli=0; /* Last valid wave */
1.126 brouard 4915: m=firstpass;
1.214 brouard 4916: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4917: 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 */
4918: mli=m-1;/* mw[++mi][i]=m-1; */
4919: }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 */
4920: mw[++mi][i]=m;
4921: mli=m;
1.224 brouard 4922: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4923: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4924: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4925: }
1.227 brouard 4926: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4927: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4928: break;
1.224 brouard 4929: #else
1.227 brouard 4930: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4931: if(firsthree == 0){
4932: 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 pi. .\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);
4933: firsthree=1;
4934: }
4935: 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 pi. .\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);
4936: mw[++mi][i]=m;
4937: mli=m;
4938: }
4939: if(s[m][i]==-2){ /* Vital status is really unknown */
4940: nbwarn++;
4941: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4942: 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);
4943: 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);
4944: }
4945: break;
4946: }
4947: break;
1.224 brouard 4948: #endif
1.227 brouard 4949: }/* End m >= lastpass */
1.126 brouard 4950: }/* end while */
1.224 brouard 4951:
1.227 brouard 4952: /* 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 4953: /* After last pass */
1.224 brouard 4954: /* Treating death states */
1.214 brouard 4955: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4956: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4957: /* } */
1.126 brouard 4958: mi++; /* Death is another wave */
4959: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4960: /* Only death is a correct wave */
1.126 brouard 4961: mw[mi][i]=m;
1.257 ! brouard 4962: } /* else not in a death state */
1.224 brouard 4963: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.257 ! brouard 4964: else if ((int) andc[i] != 9999) { /* Date of death is known */
1.218 brouard 4965: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4966: 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 */
4967: nbwarn++;
4968: if(firstfiv==0){
4969: 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 );
4970: firstfiv=1;
4971: }else{
4972: 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 );
4973: }
4974: }else{ /* Death occured afer last wave potential bias */
4975: nberr++;
4976: if(firstwo==0){
1.257 ! brouard 4977: 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 4978: firstwo=1;
4979: }
1.257 ! brouard 4980: 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 4981: }
1.257 ! brouard 4982: }else{ /* if date of interview is unknown */
1.227 brouard 4983: /* death is known but not confirmed by death status at any wave */
4984: if(firstfour==0){
4985: 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 );
4986: firstfour=1;
4987: }
4988: 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 4989: }
1.224 brouard 4990: } /* end if date of death is known */
4991: #endif
4992: wav[i]=mi; /* mi should be the last effective wave (or mli) */
4993: /* wav[i]=mw[mi][i]; */
1.126 brouard 4994: if(mi==0){
4995: nbwarn++;
4996: if(first==0){
1.227 brouard 4997: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
4998: first=1;
1.126 brouard 4999: }
5000: if(first==1){
1.227 brouard 5001: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 5002: }
5003: } /* end mi==0 */
5004: } /* End individuals */
1.214 brouard 5005: /* wav and mw are no more changed */
1.223 brouard 5006:
1.214 brouard 5007:
1.126 brouard 5008: for(i=1; i<=imx; i++){
5009: for(mi=1; mi<wav[i];mi++){
5010: if (stepm <=0)
1.227 brouard 5011: dh[mi][i]=1;
1.126 brouard 5012: else{
1.227 brouard 5013: if (s[mw[mi+1][i]][i] > nlstate) { /* A death */
5014: if (agedc[i] < 2*AGESUP) {
5015: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
5016: if(j==0) j=1; /* Survives at least one month after exam */
5017: else if(j<0){
5018: nberr++;
5019: 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]);
5020: j=1; /* Temporary Dangerous patch */
5021: 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);
5022: 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]);
5023: 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);
5024: }
5025: k=k+1;
5026: if (j >= jmax){
5027: jmax=j;
5028: ijmax=i;
5029: }
5030: if (j <= jmin){
5031: jmin=j;
5032: ijmin=i;
5033: }
5034: sum=sum+j;
5035: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
5036: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
5037: }
5038: }
5039: else{
5040: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 5041: /* 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 5042:
1.227 brouard 5043: k=k+1;
5044: if (j >= jmax) {
5045: jmax=j;
5046: ijmax=i;
5047: }
5048: else if (j <= jmin){
5049: jmin=j;
5050: ijmin=i;
5051: }
5052: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
5053: /*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]);*/
5054: if(j<0){
5055: nberr++;
5056: 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]);
5057: 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]);
5058: }
5059: sum=sum+j;
5060: }
5061: jk= j/stepm;
5062: jl= j -jk*stepm;
5063: ju= j -(jk+1)*stepm;
5064: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
5065: if(jl==0){
5066: dh[mi][i]=jk;
5067: bh[mi][i]=0;
5068: }else{ /* We want a negative bias in order to only have interpolation ie
5069: * to avoid the price of an extra matrix product in likelihood */
5070: dh[mi][i]=jk+1;
5071: bh[mi][i]=ju;
5072: }
5073: }else{
5074: if(jl <= -ju){
5075: dh[mi][i]=jk;
5076: bh[mi][i]=jl; /* bias is positive if real duration
5077: * is higher than the multiple of stepm and negative otherwise.
5078: */
5079: }
5080: else{
5081: dh[mi][i]=jk+1;
5082: bh[mi][i]=ju;
5083: }
5084: if(dh[mi][i]==0){
5085: dh[mi][i]=1; /* At least one step */
5086: bh[mi][i]=ju; /* At least one step */
5087: /* 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);*/
5088: }
5089: } /* end if mle */
1.126 brouard 5090: }
5091: } /* end wave */
5092: }
5093: jmean=sum/k;
5094: 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 5095: 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 5096: }
1.126 brouard 5097:
5098: /*********** Tricode ****************************/
1.220 brouard 5099: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.242 brouard 5100: {
5101: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
5102: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
5103: * Boring subroutine which should only output nbcode[Tvar[j]][k]
5104: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
5105: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
5106: */
1.130 brouard 5107:
1.242 brouard 5108: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
5109: int modmaxcovj=0; /* Modality max of covariates j */
5110: int cptcode=0; /* Modality max of covariates j */
5111: int modmincovj=0; /* Modality min of covariates j */
1.145 brouard 5112:
5113:
1.242 brouard 5114: /* cptcoveff=0; */
5115: /* *cptcov=0; */
1.126 brouard 5116:
1.242 brouard 5117: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 5118:
1.242 brouard 5119: /* Loop on covariates without age and products and no quantitative variable */
5120: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
5121: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
5122: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
5123: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
5124: switch(Fixed[k]) {
5125: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
5126: 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*/
5127: ij=(int)(covar[Tvar[k]][i]);
5128: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
5129: * If product of Vn*Vm, still boolean *:
5130: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
5131: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
5132: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
5133: modality of the nth covariate of individual i. */
5134: if (ij > modmaxcovj)
5135: modmaxcovj=ij;
5136: else if (ij < modmincovj)
5137: modmincovj=ij;
5138: if ((ij < -1) && (ij > NCOVMAX)){
5139: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
5140: exit(1);
5141: }else
5142: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
5143: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
5144: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
5145: /* getting the maximum value of the modality of the covariate
5146: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
5147: female ies 1, then modmaxcovj=1.
5148: */
5149: } /* end for loop on individuals i */
5150: printf(" Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5151: fprintf(ficlog," Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5152: cptcode=modmaxcovj;
5153: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
5154: /*for (i=0; i<=cptcode; i++) {*/
5155: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
5156: printf("Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5157: fprintf(ficlog, "Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5158: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
5159: if( j != -1){
5160: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
5161: covariate for which somebody answered excluding
5162: undefined. Usually 2: 0 and 1. */
5163: }
5164: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
5165: covariate for which somebody answered including
5166: undefined. Usually 3: -1, 0 and 1. */
5167: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
5168: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
5169: } /* Ndum[-1] number of undefined modalities */
1.231 brouard 5170:
1.242 brouard 5171: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
5172: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
5173: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
5174: /* modmincovj=3; modmaxcovj = 7; */
5175: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
5176: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
5177: /* defining two dummy variables: variables V1_1 and V1_2.*/
5178: /* nbcode[Tvar[j]][ij]=k; */
5179: /* nbcode[Tvar[j]][1]=0; */
5180: /* nbcode[Tvar[j]][2]=1; */
5181: /* nbcode[Tvar[j]][3]=2; */
5182: /* To be continued (not working yet). */
5183: ij=0; /* ij is similar to i but can jump over null modalities */
5184: 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*/
5185: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
5186: break;
5187: }
5188: ij++;
5189: 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*/
5190: cptcode = ij; /* New max modality for covar j */
5191: } /* end of loop on modality i=-1 to 1 or more */
5192: break;
5193: case 1: /* Testing on varying covariate, could be simple and
5194: * should look at waves or product of fixed *
5195: * varying. No time to test -1, assuming 0 and 1 only */
5196: ij=0;
5197: for(i=0; i<=1;i++){
5198: nbcode[Tvar[k]][++ij]=i;
5199: }
5200: break;
5201: default:
5202: break;
5203: } /* end switch */
5204: } /* end dummy test */
5205:
5206: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
5207: /* /\*recode from 0 *\/ */
5208: /* k is a modality. If we have model=V1+V1*sex */
5209: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
5210: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
5211: /* } */
5212: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
5213: /* if (ij > ncodemax[j]) { */
5214: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5215: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5216: /* break; */
5217: /* } */
5218: /* } /\* end of loop on modality k *\/ */
5219: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
5220:
5221: for (k=-1; k< maxncov; k++) Ndum[k]=0;
5222: /* Look at fixed dummy (single or product) covariates to check empty modalities */
5223: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
5224: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
5225: 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 */
5226: 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 */
5227: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
5228: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
5229:
5230: ij=0;
5231: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
5232: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
5233: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
5234: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
5235: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
5236: /* If product not in single variable we don't print results */
5237: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
5238: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
5239: 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*/
5240: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
5241: 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 */
5242: if(Fixed[k]!=0)
5243: anyvaryingduminmodel=1;
5244: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
5245: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
5246: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
5247: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
5248: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
5249: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
5250: }
5251: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
5252: /* ij--; */
5253: /* cptcoveff=ij; /\*Number of total covariates*\/ */
5254: *cptcov=ij; /*Number of total real effective covariates: effective
5255: * because they can be excluded from the model and real
5256: * if in the model but excluded because missing values, but how to get k from ij?*/
5257: for(j=ij+1; j<= cptcovt; j++){
5258: Tvaraff[j]=0;
5259: Tmodelind[j]=0;
5260: }
5261: for(j=ntveff+1; j<= cptcovt; j++){
5262: TmodelInvind[j]=0;
5263: }
5264: /* To be sorted */
5265: ;
5266: }
1.126 brouard 5267:
1.145 brouard 5268:
1.126 brouard 5269: /*********** Health Expectancies ****************/
5270:
1.235 brouard 5271: 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 5272:
5273: {
5274: /* Health expectancies, no variances */
1.164 brouard 5275: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 5276: int nhstepma, nstepma; /* Decreasing with age */
5277: double age, agelim, hf;
5278: double ***p3mat;
5279: double eip;
5280:
1.238 brouard 5281: /* pstamp(ficreseij); */
1.126 brouard 5282: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
5283: fprintf(ficreseij,"# Age");
5284: for(i=1; i<=nlstate;i++){
5285: for(j=1; j<=nlstate;j++){
5286: fprintf(ficreseij," e%1d%1d ",i,j);
5287: }
5288: fprintf(ficreseij," e%1d. ",i);
5289: }
5290: fprintf(ficreseij,"\n");
5291:
5292:
5293: if(estepm < stepm){
5294: printf ("Problem %d lower than %d\n",estepm, stepm);
5295: }
5296: else hstepm=estepm;
5297: /* We compute the life expectancy from trapezoids spaced every estepm months
5298: * This is mainly to measure the difference between two models: for example
5299: * if stepm=24 months pijx are given only every 2 years and by summing them
5300: * we are calculating an estimate of the Life Expectancy assuming a linear
5301: * progression in between and thus overestimating or underestimating according
5302: * to the curvature of the survival function. If, for the same date, we
5303: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5304: * to compare the new estimate of Life expectancy with the same linear
5305: * hypothesis. A more precise result, taking into account a more precise
5306: * curvature will be obtained if estepm is as small as stepm. */
5307:
5308: /* For example we decided to compute the life expectancy with the smallest unit */
5309: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5310: nhstepm is the number of hstepm from age to agelim
5311: nstepm is the number of stepm from age to agelin.
5312: Look at hpijx to understand the reason of that which relies in memory size
5313: and note for a fixed period like estepm months */
5314: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5315: survival function given by stepm (the optimization length). Unfortunately it
5316: means that if the survival funtion is printed only each two years of age and if
5317: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5318: results. So we changed our mind and took the option of the best precision.
5319: */
5320: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5321:
5322: agelim=AGESUP;
5323: /* If stepm=6 months */
5324: /* Computed by stepm unit matrices, product of hstepm matrices, stored
5325: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
5326:
5327: /* nhstepm age range expressed in number of stepm */
5328: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5329: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5330: /* if (stepm >= YEARM) hstepm=1;*/
5331: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5332: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5333:
5334: for (age=bage; age<=fage; age ++){
5335: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5336: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5337: /* if (stepm >= YEARM) hstepm=1;*/
5338: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
5339:
5340: /* If stepm=6 months */
5341: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5342: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5343:
1.235 brouard 5344: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 5345:
5346: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5347:
5348: printf("%d|",(int)age);fflush(stdout);
5349: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5350:
5351: /* Computing expectancies */
5352: for(i=1; i<=nlstate;i++)
5353: for(j=1; j<=nlstate;j++)
5354: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5355: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
5356:
5357: /* 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]);*/
5358:
5359: }
5360:
5361: fprintf(ficreseij,"%3.0f",age );
5362: for(i=1; i<=nlstate;i++){
5363: eip=0;
5364: for(j=1; j<=nlstate;j++){
5365: eip +=eij[i][j][(int)age];
5366: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
5367: }
5368: fprintf(ficreseij,"%9.4f", eip );
5369: }
5370: fprintf(ficreseij,"\n");
5371:
5372: }
5373: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5374: printf("\n");
5375: fprintf(ficlog,"\n");
5376:
5377: }
5378:
1.235 brouard 5379: 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 5380:
5381: {
5382: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5383: to initial status i, ei. .
1.126 brouard 5384: */
5385: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5386: int nhstepma, nstepma; /* Decreasing with age */
5387: double age, agelim, hf;
5388: double ***p3matp, ***p3matm, ***varhe;
5389: double **dnewm,**doldm;
5390: double *xp, *xm;
5391: double **gp, **gm;
5392: double ***gradg, ***trgradg;
5393: int theta;
5394:
5395: double eip, vip;
5396:
5397: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5398: xp=vector(1,npar);
5399: xm=vector(1,npar);
5400: dnewm=matrix(1,nlstate*nlstate,1,npar);
5401: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5402:
5403: pstamp(ficresstdeij);
5404: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5405: fprintf(ficresstdeij,"# Age");
5406: for(i=1; i<=nlstate;i++){
5407: for(j=1; j<=nlstate;j++)
5408: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5409: fprintf(ficresstdeij," e%1d. ",i);
5410: }
5411: fprintf(ficresstdeij,"\n");
5412:
5413: pstamp(ficrescveij);
5414: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5415: fprintf(ficrescveij,"# Age");
5416: for(i=1; i<=nlstate;i++)
5417: for(j=1; j<=nlstate;j++){
5418: cptj= (j-1)*nlstate+i;
5419: for(i2=1; i2<=nlstate;i2++)
5420: for(j2=1; j2<=nlstate;j2++){
5421: cptj2= (j2-1)*nlstate+i2;
5422: if(cptj2 <= cptj)
5423: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5424: }
5425: }
5426: fprintf(ficrescveij,"\n");
5427:
5428: if(estepm < stepm){
5429: printf ("Problem %d lower than %d\n",estepm, stepm);
5430: }
5431: else hstepm=estepm;
5432: /* We compute the life expectancy from trapezoids spaced every estepm months
5433: * This is mainly to measure the difference between two models: for example
5434: * if stepm=24 months pijx are given only every 2 years and by summing them
5435: * we are calculating an estimate of the Life Expectancy assuming a linear
5436: * progression in between and thus overestimating or underestimating according
5437: * to the curvature of the survival function. If, for the same date, we
5438: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5439: * to compare the new estimate of Life expectancy with the same linear
5440: * hypothesis. A more precise result, taking into account a more precise
5441: * curvature will be obtained if estepm is as small as stepm. */
5442:
5443: /* For example we decided to compute the life expectancy with the smallest unit */
5444: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5445: nhstepm is the number of hstepm from age to agelim
5446: nstepm is the number of stepm from age to agelin.
5447: Look at hpijx to understand the reason of that which relies in memory size
5448: and note for a fixed period like estepm months */
5449: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5450: survival function given by stepm (the optimization length). Unfortunately it
5451: means that if the survival funtion is printed only each two years of age and if
5452: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5453: results. So we changed our mind and took the option of the best precision.
5454: */
5455: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5456:
5457: /* If stepm=6 months */
5458: /* nhstepm age range expressed in number of stepm */
5459: agelim=AGESUP;
5460: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5461: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5462: /* if (stepm >= YEARM) hstepm=1;*/
5463: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5464:
5465: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5466: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5467: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5468: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5469: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5470: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5471:
5472: for (age=bage; age<=fage; age ++){
5473: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5474: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5475: /* if (stepm >= YEARM) hstepm=1;*/
5476: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5477:
1.126 brouard 5478: /* If stepm=6 months */
5479: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5480: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5481:
5482: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5483:
1.126 brouard 5484: /* Computing Variances of health expectancies */
5485: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5486: decrease memory allocation */
5487: for(theta=1; theta <=npar; theta++){
5488: for(i=1; i<=npar; i++){
1.222 brouard 5489: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5490: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5491: }
1.235 brouard 5492: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5493: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5494:
1.126 brouard 5495: for(j=1; j<= nlstate; j++){
1.222 brouard 5496: for(i=1; i<=nlstate; i++){
5497: for(h=0; h<=nhstepm-1; h++){
5498: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5499: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5500: }
5501: }
1.126 brouard 5502: }
1.218 brouard 5503:
1.126 brouard 5504: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5505: for(h=0; h<=nhstepm-1; h++){
5506: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5507: }
1.126 brouard 5508: }/* End theta */
5509:
5510:
5511: for(h=0; h<=nhstepm-1; h++)
5512: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5513: for(theta=1; theta <=npar; theta++)
5514: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5515:
1.218 brouard 5516:
1.222 brouard 5517: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5518: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5519: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5520:
1.222 brouard 5521: printf("%d|",(int)age);fflush(stdout);
5522: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5523: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5524: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5525: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5526: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5527: for(ij=1;ij<=nlstate*nlstate;ij++)
5528: for(ji=1;ji<=nlstate*nlstate;ji++)
5529: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5530: }
5531: }
1.218 brouard 5532:
1.126 brouard 5533: /* Computing expectancies */
1.235 brouard 5534: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5535: for(i=1; i<=nlstate;i++)
5536: for(j=1; j<=nlstate;j++)
1.222 brouard 5537: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5538: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5539:
1.222 brouard 5540: /* 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 5541:
1.222 brouard 5542: }
1.218 brouard 5543:
1.126 brouard 5544: fprintf(ficresstdeij,"%3.0f",age );
5545: for(i=1; i<=nlstate;i++){
5546: eip=0.;
5547: vip=0.;
5548: for(j=1; j<=nlstate;j++){
1.222 brouard 5549: eip += eij[i][j][(int)age];
5550: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5551: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5552: 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 5553: }
5554: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5555: }
5556: fprintf(ficresstdeij,"\n");
1.218 brouard 5557:
1.126 brouard 5558: fprintf(ficrescveij,"%3.0f",age );
5559: for(i=1; i<=nlstate;i++)
5560: for(j=1; j<=nlstate;j++){
1.222 brouard 5561: cptj= (j-1)*nlstate+i;
5562: for(i2=1; i2<=nlstate;i2++)
5563: for(j2=1; j2<=nlstate;j2++){
5564: cptj2= (j2-1)*nlstate+i2;
5565: if(cptj2 <= cptj)
5566: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5567: }
1.126 brouard 5568: }
5569: fprintf(ficrescveij,"\n");
1.218 brouard 5570:
1.126 brouard 5571: }
5572: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5573: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5574: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5575: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5576: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5577: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5578: printf("\n");
5579: fprintf(ficlog,"\n");
1.218 brouard 5580:
1.126 brouard 5581: free_vector(xm,1,npar);
5582: free_vector(xp,1,npar);
5583: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5584: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5585: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5586: }
1.218 brouard 5587:
1.126 brouard 5588: /************ Variance ******************/
1.235 brouard 5589: 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 5590: {
5591: /* Variance of health expectancies */
5592: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5593: /* double **newm;*/
5594: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5595:
5596: /* int movingaverage(); */
5597: double **dnewm,**doldm;
5598: double **dnewmp,**doldmp;
5599: int i, j, nhstepm, hstepm, h, nstepm ;
5600: int k;
5601: double *xp;
5602: double **gp, **gm; /* for var eij */
5603: double ***gradg, ***trgradg; /*for var eij */
5604: double **gradgp, **trgradgp; /* for var p point j */
5605: double *gpp, *gmp; /* for var p point j */
5606: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5607: double ***p3mat;
5608: double age,agelim, hf;
5609: /* double ***mobaverage; */
5610: int theta;
5611: char digit[4];
5612: char digitp[25];
5613:
5614: char fileresprobmorprev[FILENAMELENGTH];
5615:
5616: if(popbased==1){
5617: if(mobilav!=0)
5618: strcpy(digitp,"-POPULBASED-MOBILAV_");
5619: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5620: }
5621: else
5622: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5623:
1.218 brouard 5624: /* if (mobilav!=0) { */
5625: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5626: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5627: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5628: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5629: /* } */
5630: /* } */
5631:
5632: strcpy(fileresprobmorprev,"PRMORPREV-");
5633: sprintf(digit,"%-d",ij);
5634: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5635: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5636: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5637: strcat(fileresprobmorprev,fileresu);
5638: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5639: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5640: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5641: }
5642: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5643: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5644: pstamp(ficresprobmorprev);
5645: 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 5646: fprintf(ficresprobmorprev,"# Selected quantitative variables and dummies");
5647: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
5648: fprintf(ficresprobmorprev," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
5649: }
5650: for(j=1;j<=cptcoveff;j++)
5651: fprintf(ficresprobmorprev,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(ij,j)]);
5652: fprintf(ficresprobmorprev,"\n");
5653:
1.218 brouard 5654: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5655: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5656: fprintf(ficresprobmorprev," p.%-d SE",j);
5657: for(i=1; i<=nlstate;i++)
5658: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5659: }
5660: fprintf(ficresprobmorprev,"\n");
5661:
5662: fprintf(ficgp,"\n# Routine varevsij");
5663: fprintf(ficgp,"\nunset title \n");
5664: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5665: 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");
5666: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5667: /* } */
5668: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5669: pstamp(ficresvij);
5670: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5671: if(popbased==1)
5672: 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);
5673: else
5674: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5675: fprintf(ficresvij,"# Age");
5676: for(i=1; i<=nlstate;i++)
5677: for(j=1; j<=nlstate;j++)
5678: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5679: fprintf(ficresvij,"\n");
5680:
5681: xp=vector(1,npar);
5682: dnewm=matrix(1,nlstate,1,npar);
5683: doldm=matrix(1,nlstate,1,nlstate);
5684: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5685: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5686:
5687: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5688: gpp=vector(nlstate+1,nlstate+ndeath);
5689: gmp=vector(nlstate+1,nlstate+ndeath);
5690: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5691:
1.218 brouard 5692: if(estepm < stepm){
5693: printf ("Problem %d lower than %d\n",estepm, stepm);
5694: }
5695: else hstepm=estepm;
5696: /* For example we decided to compute the life expectancy with the smallest unit */
5697: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5698: nhstepm is the number of hstepm from age to agelim
5699: nstepm is the number of stepm from age to agelim.
5700: Look at function hpijx to understand why because of memory size limitations,
5701: we decided (b) to get a life expectancy respecting the most precise curvature of the
5702: survival function given by stepm (the optimization length). Unfortunately it
5703: means that if the survival funtion is printed every two years of age and if
5704: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5705: results. So we changed our mind and took the option of the best precision.
5706: */
5707: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5708: agelim = AGESUP;
5709: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5710: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5711: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5712: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5713: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5714: gp=matrix(0,nhstepm,1,nlstate);
5715: gm=matrix(0,nhstepm,1,nlstate);
5716:
5717:
5718: for(theta=1; theta <=npar; theta++){
5719: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5720: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5721: }
5722:
1.242 brouard 5723: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5724:
5725: if (popbased==1) {
5726: if(mobilav ==0){
5727: for(i=1; i<=nlstate;i++)
5728: prlim[i][i]=probs[(int)age][i][ij];
5729: }else{ /* mobilav */
5730: for(i=1; i<=nlstate;i++)
5731: prlim[i][i]=mobaverage[(int)age][i][ij];
5732: }
5733: }
5734:
1.235 brouard 5735: 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 5736: for(j=1; j<= nlstate; j++){
5737: for(h=0; h<=nhstepm; h++){
5738: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5739: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5740: }
5741: }
5742: /* Next for computing probability of death (h=1 means
5743: computed over hstepm matrices product = hstepm*stepm months)
5744: as a weighted average of prlim.
5745: */
5746: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5747: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5748: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5749: }
5750: /* end probability of death */
5751:
5752: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5753: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5754:
1.242 brouard 5755: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nres);
1.218 brouard 5756:
5757: if (popbased==1) {
5758: if(mobilav ==0){
5759: for(i=1; i<=nlstate;i++)
5760: prlim[i][i]=probs[(int)age][i][ij];
5761: }else{ /* mobilav */
5762: for(i=1; i<=nlstate;i++)
5763: prlim[i][i]=mobaverage[(int)age][i][ij];
5764: }
5765: }
5766:
1.235 brouard 5767: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5768:
5769: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5770: for(h=0; h<=nhstepm; h++){
5771: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5772: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5773: }
5774: }
5775: /* This for computing probability of death (h=1 means
5776: computed over hstepm matrices product = hstepm*stepm months)
5777: as a weighted average of prlim.
5778: */
5779: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5780: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5781: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5782: }
5783: /* end probability of death */
5784:
5785: for(j=1; j<= nlstate; j++) /* vareij */
5786: for(h=0; h<=nhstepm; h++){
5787: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5788: }
5789:
5790: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5791: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5792: }
5793:
5794: } /* End theta */
5795:
5796: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5797:
5798: for(h=0; h<=nhstepm; h++) /* veij */
5799: for(j=1; j<=nlstate;j++)
5800: for(theta=1; theta <=npar; theta++)
5801: trgradg[h][j][theta]=gradg[h][theta][j];
5802:
5803: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5804: for(theta=1; theta <=npar; theta++)
5805: trgradgp[j][theta]=gradgp[theta][j];
5806:
5807:
5808: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5809: for(i=1;i<=nlstate;i++)
5810: for(j=1;j<=nlstate;j++)
5811: vareij[i][j][(int)age] =0.;
5812:
5813: for(h=0;h<=nhstepm;h++){
5814: for(k=0;k<=nhstepm;k++){
5815: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5816: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5817: for(i=1;i<=nlstate;i++)
5818: for(j=1;j<=nlstate;j++)
5819: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5820: }
5821: }
5822:
5823: /* pptj */
5824: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5825: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5826: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5827: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5828: varppt[j][i]=doldmp[j][i];
5829: /* end ppptj */
5830: /* x centered again */
5831:
1.242 brouard 5832: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5833:
5834: if (popbased==1) {
5835: if(mobilav ==0){
5836: for(i=1; i<=nlstate;i++)
5837: prlim[i][i]=probs[(int)age][i][ij];
5838: }else{ /* mobilav */
5839: for(i=1; i<=nlstate;i++)
5840: prlim[i][i]=mobaverage[(int)age][i][ij];
5841: }
5842: }
5843:
5844: /* This for computing probability of death (h=1 means
5845: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5846: as a weighted average of prlim.
5847: */
1.235 brouard 5848: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5849: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5850: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5851: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5852: }
5853: /* end probability of death */
5854:
5855: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5856: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5857: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5858: for(i=1; i<=nlstate;i++){
5859: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5860: }
5861: }
5862: fprintf(ficresprobmorprev,"\n");
5863:
5864: fprintf(ficresvij,"%.0f ",age );
5865: for(i=1; i<=nlstate;i++)
5866: for(j=1; j<=nlstate;j++){
5867: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5868: }
5869: fprintf(ficresvij,"\n");
5870: free_matrix(gp,0,nhstepm,1,nlstate);
5871: free_matrix(gm,0,nhstepm,1,nlstate);
5872: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5873: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5874: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5875: } /* End age */
5876: free_vector(gpp,nlstate+1,nlstate+ndeath);
5877: free_vector(gmp,nlstate+1,nlstate+ndeath);
5878: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5879: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5880: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5881: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5882: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5883: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5884: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5885: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5886: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5887: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5888: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5889: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5890: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5891: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5892: 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);
5893: /* 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 5894: */
1.218 brouard 5895: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5896: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5897:
1.218 brouard 5898: free_vector(xp,1,npar);
5899: free_matrix(doldm,1,nlstate,1,nlstate);
5900: free_matrix(dnewm,1,nlstate,1,npar);
5901: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5902: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5903: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5904: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5905: fclose(ficresprobmorprev);
5906: fflush(ficgp);
5907: fflush(fichtm);
5908: } /* end varevsij */
1.126 brouard 5909:
5910: /************ Variance of prevlim ******************/
1.235 brouard 5911: 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 5912: {
1.205 brouard 5913: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5914: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5915:
1.126 brouard 5916: double **dnewm,**doldm;
5917: int i, j, nhstepm, hstepm;
5918: double *xp;
5919: double *gp, *gm;
5920: double **gradg, **trgradg;
1.208 brouard 5921: double **mgm, **mgp;
1.126 brouard 5922: double age,agelim;
5923: int theta;
5924:
5925: pstamp(ficresvpl);
5926: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
1.241 brouard 5927: fprintf(ficresvpl,"# Age ");
5928: if(nresult >=1)
5929: fprintf(ficresvpl," Result# ");
1.126 brouard 5930: for(i=1; i<=nlstate;i++)
5931: fprintf(ficresvpl," %1d-%1d",i,i);
5932: fprintf(ficresvpl,"\n");
5933:
5934: xp=vector(1,npar);
5935: dnewm=matrix(1,nlstate,1,npar);
5936: doldm=matrix(1,nlstate,1,nlstate);
5937:
5938: hstepm=1*YEARM; /* Every year of age */
5939: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5940: agelim = AGESUP;
5941: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5942: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5943: if (stepm >= YEARM) hstepm=1;
5944: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5945: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5946: mgp=matrix(1,npar,1,nlstate);
5947: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5948: gp=vector(1,nlstate);
5949: gm=vector(1,nlstate);
5950:
5951: for(theta=1; theta <=npar; theta++){
5952: for(i=1; i<=npar; i++){ /* Computes gradient */
5953: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5954: }
1.209 brouard 5955: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5956: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5957: else
1.235 brouard 5958: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5959: for(i=1;i<=nlstate;i++){
1.126 brouard 5960: gp[i] = prlim[i][i];
1.208 brouard 5961: mgp[theta][i] = prlim[i][i];
5962: }
1.126 brouard 5963: for(i=1; i<=npar; i++) /* Computes gradient */
5964: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5965: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5966: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5967: else
1.235 brouard 5968: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5969: for(i=1;i<=nlstate;i++){
1.126 brouard 5970: gm[i] = prlim[i][i];
1.208 brouard 5971: mgm[theta][i] = prlim[i][i];
5972: }
1.126 brouard 5973: for(i=1;i<=nlstate;i++)
5974: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5975: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5976: } /* End theta */
5977:
5978: trgradg =matrix(1,nlstate,1,npar);
5979:
5980: for(j=1; j<=nlstate;j++)
5981: for(theta=1; theta <=npar; theta++)
5982: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 5983: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5984: /* printf("\nmgm mgp %d ",(int)age); */
5985: /* for(j=1; j<=nlstate;j++){ */
5986: /* printf(" %d ",j); */
5987: /* for(theta=1; theta <=npar; theta++) */
5988: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
5989: /* printf("\n "); */
5990: /* } */
5991: /* } */
5992: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5993: /* printf("\n gradg %d ",(int)age); */
5994: /* for(j=1; j<=nlstate;j++){ */
5995: /* printf("%d ",j); */
5996: /* for(theta=1; theta <=npar; theta++) */
5997: /* printf("%d %lf ",theta,gradg[theta][j]); */
5998: /* printf("\n "); */
5999: /* } */
6000: /* } */
1.126 brouard 6001:
6002: for(i=1;i<=nlstate;i++)
6003: varpl[i][(int)age] =0.;
1.209 brouard 6004: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 6005: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6006: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
6007: }else{
1.126 brouard 6008: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6009: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 6010: }
1.126 brouard 6011: for(i=1;i<=nlstate;i++)
6012: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
6013:
6014: fprintf(ficresvpl,"%.0f ",age );
1.241 brouard 6015: if(nresult >=1)
6016: fprintf(ficresvpl,"%d ",nres );
1.126 brouard 6017: for(i=1; i<=nlstate;i++)
6018: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
6019: fprintf(ficresvpl,"\n");
6020: free_vector(gp,1,nlstate);
6021: free_vector(gm,1,nlstate);
1.208 brouard 6022: free_matrix(mgm,1,npar,1,nlstate);
6023: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 6024: free_matrix(gradg,1,npar,1,nlstate);
6025: free_matrix(trgradg,1,nlstate,1,npar);
6026: } /* End age */
6027:
6028: free_vector(xp,1,npar);
6029: free_matrix(doldm,1,nlstate,1,npar);
6030: free_matrix(dnewm,1,nlstate,1,nlstate);
6031:
6032: }
6033:
6034: /************ Variance of one-step probabilities ******************/
6035: 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 6036: {
6037: int i, j=0, k1, l1, tj;
6038: int k2, l2, j1, z1;
6039: int k=0, l;
6040: int first=1, first1, first2;
6041: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
6042: double **dnewm,**doldm;
6043: double *xp;
6044: double *gp, *gm;
6045: double **gradg, **trgradg;
6046: double **mu;
6047: double age, cov[NCOVMAX+1];
6048: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
6049: int theta;
6050: char fileresprob[FILENAMELENGTH];
6051: char fileresprobcov[FILENAMELENGTH];
6052: char fileresprobcor[FILENAMELENGTH];
6053: double ***varpij;
6054:
6055: strcpy(fileresprob,"PROB_");
6056: strcat(fileresprob,fileres);
6057: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
6058: printf("Problem with resultfile: %s\n", fileresprob);
6059: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
6060: }
6061: strcpy(fileresprobcov,"PROBCOV_");
6062: strcat(fileresprobcov,fileresu);
6063: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
6064: printf("Problem with resultfile: %s\n", fileresprobcov);
6065: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
6066: }
6067: strcpy(fileresprobcor,"PROBCOR_");
6068: strcat(fileresprobcor,fileresu);
6069: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
6070: printf("Problem with resultfile: %s\n", fileresprobcor);
6071: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
6072: }
6073: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6074: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6075: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6076: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6077: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6078: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6079: pstamp(ficresprob);
6080: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
6081: fprintf(ficresprob,"# Age");
6082: pstamp(ficresprobcov);
6083: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
6084: fprintf(ficresprobcov,"# Age");
6085: pstamp(ficresprobcor);
6086: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
6087: fprintf(ficresprobcor,"# Age");
1.126 brouard 6088:
6089:
1.222 brouard 6090: for(i=1; i<=nlstate;i++)
6091: for(j=1; j<=(nlstate+ndeath);j++){
6092: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
6093: fprintf(ficresprobcov," p%1d-%1d ",i,j);
6094: fprintf(ficresprobcor," p%1d-%1d ",i,j);
6095: }
6096: /* fprintf(ficresprob,"\n");
6097: fprintf(ficresprobcov,"\n");
6098: fprintf(ficresprobcor,"\n");
6099: */
6100: xp=vector(1,npar);
6101: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6102: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6103: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
6104: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
6105: first=1;
6106: fprintf(ficgp,"\n# Routine varprob");
6107: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
6108: fprintf(fichtm,"\n");
6109:
6110: 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);
6111: 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);
6112: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 6113: and drawn. It helps understanding how is the covariance between two incidences.\
6114: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 6115: 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 6116: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
6117: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
6118: standard deviations wide on each axis. <br>\
6119: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
6120: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
6121: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
6122:
1.222 brouard 6123: cov[1]=1;
6124: /* tj=cptcoveff; */
1.225 brouard 6125: tj = (int) pow(2,cptcoveff);
1.222 brouard 6126: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
6127: j1=0;
1.224 brouard 6128: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 6129: if (cptcovn>0) {
6130: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 6131: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6132: fprintf(ficresprob, "**********\n#\n");
6133: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 6134: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6135: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 6136:
1.222 brouard 6137: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 6138: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6139: fprintf(ficgp, "**********\n#\n");
1.220 brouard 6140:
6141:
1.222 brouard 6142: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 6143: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6144: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6145:
1.222 brouard 6146: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 6147: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6148: fprintf(ficresprobcor, "**********\n#");
6149: if(invalidvarcomb[j1]){
6150: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
6151: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
6152: continue;
6153: }
6154: }
6155: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
6156: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6157: gp=vector(1,(nlstate)*(nlstate+ndeath));
6158: gm=vector(1,(nlstate)*(nlstate+ndeath));
6159: for (age=bage; age<=fage; age ++){
6160: cov[2]=age;
6161: if(nagesqr==1)
6162: cov[3]= age*age;
6163: for (k=1; k<=cptcovn;k++) {
6164: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
6165: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
6166: * 1 1 1 1 1
6167: * 2 2 1 1 1
6168: * 3 1 2 1 1
6169: */
6170: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
6171: }
6172: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
6173: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
6174: for (k=1; k<=cptcovprod;k++)
6175: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 6176:
6177:
1.222 brouard 6178: for(theta=1; theta <=npar; theta++){
6179: for(i=1; i<=npar; i++)
6180: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 6181:
1.222 brouard 6182: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 6183:
1.222 brouard 6184: k=0;
6185: for(i=1; i<= (nlstate); i++){
6186: for(j=1; j<=(nlstate+ndeath);j++){
6187: k=k+1;
6188: gp[k]=pmmij[i][j];
6189: }
6190: }
1.220 brouard 6191:
1.222 brouard 6192: for(i=1; i<=npar; i++)
6193: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 6194:
1.222 brouard 6195: pmij(pmmij,cov,ncovmodel,xp,nlstate);
6196: k=0;
6197: for(i=1; i<=(nlstate); i++){
6198: for(j=1; j<=(nlstate+ndeath);j++){
6199: k=k+1;
6200: gm[k]=pmmij[i][j];
6201: }
6202: }
1.220 brouard 6203:
1.222 brouard 6204: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
6205: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
6206: }
1.126 brouard 6207:
1.222 brouard 6208: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
6209: for(theta=1; theta <=npar; theta++)
6210: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 6211:
1.222 brouard 6212: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
6213: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 6214:
1.222 brouard 6215: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 6216:
1.222 brouard 6217: k=0;
6218: for(i=1; i<=(nlstate); i++){
6219: for(j=1; j<=(nlstate+ndeath);j++){
6220: k=k+1;
6221: mu[k][(int) age]=pmmij[i][j];
6222: }
6223: }
6224: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
6225: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
6226: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 6227:
1.222 brouard 6228: /*printf("\n%d ",(int)age);
6229: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6230: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6231: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6232: }*/
1.220 brouard 6233:
1.222 brouard 6234: fprintf(ficresprob,"\n%d ",(int)age);
6235: fprintf(ficresprobcov,"\n%d ",(int)age);
6236: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 6237:
1.222 brouard 6238: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
6239: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
6240: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6241: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
6242: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
6243: }
6244: i=0;
6245: for (k=1; k<=(nlstate);k++){
6246: for (l=1; l<=(nlstate+ndeath);l++){
6247: i++;
6248: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
6249: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
6250: for (j=1; j<=i;j++){
6251: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
6252: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
6253: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
6254: }
6255: }
6256: }/* end of loop for state */
6257: } /* end of loop for age */
6258: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
6259: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
6260: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6261: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6262:
6263: /* Confidence intervalle of pij */
6264: /*
6265: fprintf(ficgp,"\nunset parametric;unset label");
6266: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
6267: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
6268: 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);
6269: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
6270: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
6271: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
6272: */
6273:
6274: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
6275: first1=1;first2=2;
6276: for (k2=1; k2<=(nlstate);k2++){
6277: for (l2=1; l2<=(nlstate+ndeath);l2++){
6278: if(l2==k2) continue;
6279: j=(k2-1)*(nlstate+ndeath)+l2;
6280: for (k1=1; k1<=(nlstate);k1++){
6281: for (l1=1; l1<=(nlstate+ndeath);l1++){
6282: if(l1==k1) continue;
6283: i=(k1-1)*(nlstate+ndeath)+l1;
6284: if(i<=j) continue;
6285: for (age=bage; age<=fage; age ++){
6286: if ((int)age %5==0){
6287: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
6288: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
6289: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
6290: mu1=mu[i][(int) age]/stepm*YEARM ;
6291: mu2=mu[j][(int) age]/stepm*YEARM;
6292: c12=cv12/sqrt(v1*v2);
6293: /* Computing eigen value of matrix of covariance */
6294: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6295: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6296: if ((lc2 <0) || (lc1 <0) ){
6297: if(first2==1){
6298: first1=0;
6299: 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);
6300: }
6301: 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);
6302: /* lc1=fabs(lc1); */ /* If we want to have them positive */
6303: /* lc2=fabs(lc2); */
6304: }
1.220 brouard 6305:
1.222 brouard 6306: /* Eigen vectors */
6307: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
6308: /*v21=sqrt(1.-v11*v11); *//* error */
6309: v21=(lc1-v1)/cv12*v11;
6310: v12=-v21;
6311: v22=v11;
6312: tnalp=v21/v11;
6313: if(first1==1){
6314: first1=0;
6315: 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);
6316: }
6317: 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);
6318: /*printf(fignu*/
6319: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
6320: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
6321: if(first==1){
6322: first=0;
6323: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
6324: fprintf(ficgp,"\nset parametric;unset label");
6325: 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);
6326: fprintf(ficgp,"\nset ter svg size 640, 480");
6327: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 6328: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 6329: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 6330: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
6331: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6332: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6333: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
6334: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6335: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6336: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6337: 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", \
6338: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6339: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6340: }else{
6341: first=0;
6342: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
6343: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6344: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6345: 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", \
6346: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6347: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6348: }/* if first */
6349: } /* age mod 5 */
6350: } /* end loop age */
6351: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6352: first=1;
6353: } /*l12 */
6354: } /* k12 */
6355: } /*l1 */
6356: }/* k1 */
6357: } /* loop on combination of covariates j1 */
6358: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
6359: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
6360: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6361: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
6362: free_vector(xp,1,npar);
6363: fclose(ficresprob);
6364: fclose(ficresprobcov);
6365: fclose(ficresprobcor);
6366: fflush(ficgp);
6367: fflush(fichtmcov);
6368: }
1.126 brouard 6369:
6370:
6371: /******************* Printing html file ***********/
1.201 brouard 6372: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 6373: int lastpass, int stepm, int weightopt, char model[],\
6374: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.217 brouard 6375: int popforecast, int prevfcast, int backcast, int estepm , \
1.213 brouard 6376: double jprev1, double mprev1,double anprev1, double dateprev1, \
6377: double jprev2, double mprev2,double anprev2, double dateprev2){
1.237 brouard 6378: int jj1, k1, i1, cpt, k4, nres;
1.126 brouard 6379:
6380: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
6381: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
6382: </ul>");
1.237 brouard 6383: fprintf(fichtm,"<ul><li> model=1+age+%s\n \
6384: </ul>", model);
1.214 brouard 6385: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
6386: 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",
6387: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
6388: 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 6389: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
6390: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 6391: fprintf(fichtm,"\
6392: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 6393: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6394: fprintf(fichtm,"\
1.217 brouard 6395: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6396: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6397: fprintf(fichtm,"\
1.126 brouard 6398: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6399: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6400: fprintf(fichtm,"\
1.217 brouard 6401: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6402: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6403: fprintf(fichtm,"\
1.211 brouard 6404: - (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 6405: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6406: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6407: if(prevfcast==1){
6408: fprintf(fichtm,"\
6409: - Prevalence projections by age and states: \
1.201 brouard 6410: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6411: }
1.126 brouard 6412:
1.222 brouard 6413: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 6414:
1.225 brouard 6415: m=pow(2,cptcoveff);
1.222 brouard 6416: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6417:
1.222 brouard 6418: jj1=0;
1.237 brouard 6419:
6420: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.241 brouard 6421: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
1.253 brouard 6422: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6423: continue;
1.220 brouard 6424:
1.222 brouard 6425: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6426: jj1++;
6427: if (cptcovn > 0) {
6428: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6429: for (cpt=1; cpt<=cptcoveff;cpt++){
1.237 brouard 6430: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6431: printf(" V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);fflush(stdout);
6432: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6433: /* printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout); */
1.222 brouard 6434: }
1.237 brouard 6435: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6436: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6437: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);fflush(stdout);
6438: }
6439:
1.230 brouard 6440: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6441: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6442: if(invalidvarcomb[k1]){
6443: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6444: printf("\nCombination (%d) ignored because no cases \n",k1);
6445: continue;
6446: }
6447: }
6448: /* aij, bij */
1.241 brouard 6449: fprintf(fichtm,"<br>- Logit model (yours is: 1+age+%s), for example: logit(pij)=log(pij/pii)= aij+ bij age + V1 age + etc. as a function of age: <a href=\"%s_%d-1-%d.svg\">%s_%d-1-%d.svg</a><br> \
6450: <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 6451: /* Pij */
1.241 brouard 6452: 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> \
6453: <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 6454: /* Quasi-incidences */
6455: 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 6456: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6457: 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 6458: 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> \
6459: <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 6460: /* Survival functions (period) in state j */
6461: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6462: 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> \
6463: <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 6464: }
6465: /* State specific survival functions (period) */
6466: for(cpt=1; cpt<=nlstate;cpt++){
6467: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6468: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.241 brouard 6469: <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 6470: }
6471: /* Period (stable) prevalence in each health state */
6472: for(cpt=1; cpt<=nlstate;cpt++){
1.255 brouard 6473: 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 6474: <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 6475: }
6476: if(backcast==1){
6477: /* Period (stable) back prevalence in each health state */
6478: for(cpt=1; cpt<=nlstate;cpt++){
1.255 brouard 6479: 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 6480: <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 6481: }
1.217 brouard 6482: }
1.222 brouard 6483: if(prevfcast==1){
6484: /* Projection of prevalence up to period (stable) prevalence in each health state */
6485: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6486: fprintf(fichtm,"<br>\n- Projection of cross-sectional prevalence (estimated with cases observed from %.1f to %.1f) 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> \
6487: <img src=\"%s_%d-%d-%d.svg\">", dateprev1, dateprev2, cpt, cpt, nlstate, subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.222 brouard 6488: }
6489: }
1.220 brouard 6490:
1.222 brouard 6491: for(cpt=1; cpt<=nlstate;cpt++) {
1.241 brouard 6492: 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> \
6493: <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 6494: }
6495: /* } /\* end i1 *\/ */
6496: }/* End k1 */
6497: fprintf(fichtm,"</ul>");
1.126 brouard 6498:
1.222 brouard 6499: fprintf(fichtm,"\
1.126 brouard 6500: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6501: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6502: - 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 6503: But because parameters are usually highly correlated (a higher incidence of disability \
6504: and a higher incidence of recovery can give very close observed transition) it might \
6505: be very useful to look not only at linear confidence intervals estimated from the \
6506: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6507: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6508: covariance matrix of the one-step probabilities. \
6509: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6510:
1.222 brouard 6511: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6512: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6513: fprintf(fichtm,"\
1.126 brouard 6514: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6515: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6516:
1.222 brouard 6517: fprintf(fichtm,"\
1.126 brouard 6518: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6519: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6520: fprintf(fichtm,"\
1.126 brouard 6521: - 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): \
6522: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6523: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6524: fprintf(fichtm,"\
1.126 brouard 6525: - (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): \
6526: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6527: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6528: fprintf(fichtm,"\
1.128 brouard 6529: - 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 6530: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6531: fprintf(fichtm,"\
1.128 brouard 6532: - 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 6533: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6534: fprintf(fichtm,"\
1.126 brouard 6535: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6536: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6537:
6538: /* if(popforecast==1) fprintf(fichtm,"\n */
6539: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6540: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6541: /* <br>",fileres,fileres,fileres,fileres); */
6542: /* else */
6543: /* 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 6544: fflush(fichtm);
6545: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6546:
1.225 brouard 6547: m=pow(2,cptcoveff);
1.222 brouard 6548: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6549:
1.222 brouard 6550: jj1=0;
1.237 brouard 6551:
1.241 brouard 6552: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.222 brouard 6553: for(k1=1; k1<=m;k1++){
1.253 brouard 6554: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6555: continue;
1.222 brouard 6556: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6557: jj1++;
1.126 brouard 6558: if (cptcovn > 0) {
6559: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6560: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.237 brouard 6561: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);
6562: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6563: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6564: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6565: }
6566:
1.126 brouard 6567: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6568:
1.222 brouard 6569: if(invalidvarcomb[k1]){
6570: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6571: continue;
6572: }
1.126 brouard 6573: }
6574: for(cpt=1; cpt<=nlstate;cpt++) {
1.218 brouard 6575: fprintf(fichtm,"\n<br>- Observed (cross-sectional) and period (incidence based) \
1.241 brouard 6576: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d-%d.svg\"> %s_%d-%d-%d.svg</a>\n <br>\
6577: <img src=\"%s_%d-%d-%d.svg\">",cpt,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
1.126 brouard 6578: }
6579: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6580: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6581: true period expectancies (those weighted with period prevalences are also\
6582: drawn in addition to the population based expectancies computed using\
1.241 brouard 6583: observed and cahotic prevalences: <a href=\"%s_%d-%d.svg\">%s_%d-%d.svg</a>\n<br>\
6584: <img src=\"%s_%d-%d.svg\">",subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres);
1.222 brouard 6585: /* } /\* end i1 *\/ */
6586: }/* End k1 */
1.241 brouard 6587: }/* End nres */
1.222 brouard 6588: fprintf(fichtm,"</ul>");
6589: fflush(fichtm);
1.126 brouard 6590: }
6591:
6592: /******************* Gnuplot file **************/
1.223 brouard 6593: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6594:
6595: char dirfileres[132],optfileres[132];
1.223 brouard 6596: char gplotcondition[132];
1.237 brouard 6597: 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 6598: int lv=0, vlv=0, kl=0;
1.130 brouard 6599: int ng=0;
1.201 brouard 6600: int vpopbased;
1.223 brouard 6601: int ioffset; /* variable offset for columns */
1.235 brouard 6602: int nres=0; /* Index of resultline */
1.219 brouard 6603:
1.126 brouard 6604: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6605: /* printf("Problem with file %s",optionfilegnuplot); */
6606: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6607: /* } */
6608:
6609: /*#ifdef windows */
6610: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6611: /*#endif */
1.225 brouard 6612: m=pow(2,cptcoveff);
1.126 brouard 6613:
1.202 brouard 6614: /* Contribution to likelihood */
6615: /* Plot the probability implied in the likelihood */
1.223 brouard 6616: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6617: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6618: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6619: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6620: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6621: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6622: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6623: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6624: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6625: 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));
6626: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6627: 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));
6628: for (i=1; i<= nlstate ; i ++) {
6629: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6630: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6631: 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);
6632: for (j=2; j<= nlstate+ndeath ; j ++) {
6633: 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);
6634: }
6635: fprintf(ficgp,";\nset out; unset ylabel;\n");
6636: }
6637: /* 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 */
6638: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6639: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6640: fprintf(ficgp,"\nset out;unset log\n");
6641: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6642:
1.126 brouard 6643: strcpy(dirfileres,optionfilefiname);
6644: strcpy(optfileres,"vpl");
1.223 brouard 6645: /* 1eme*/
1.238 brouard 6646: for (cpt=1; cpt<= nlstate ; cpt ++){ /* For each live state */
6647: for (k1=1; k1<= m ; k1 ++){ /* For each valid combination of covariate */
1.236 brouard 6648: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.238 brouard 6649: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
1.253 brouard 6650: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6651: continue;
6652: /* We are interested in selected combination by the resultline */
1.246 brouard 6653: /* printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt); */
1.238 brouard 6654: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
6655: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6656: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
6657: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6658: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6659: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6660: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6661: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.246 brouard 6662: /* printf(" V%d=%d ",Tvaraff[k],vlv); */
1.238 brouard 6663: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6664: }
6665: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.246 brouard 6666: /* printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); */
1.238 brouard 6667: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6668: }
1.246 brouard 6669: /* printf("\n#\n"); */
1.238 brouard 6670: fprintf(ficgp,"\n#\n");
6671: if(invalidvarcomb[k1]){
6672: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6673: continue;
6674: }
1.235 brouard 6675:
1.241 brouard 6676: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
6677: fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres);
6678: 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);
1.235 brouard 6679:
1.238 brouard 6680: for (i=1; i<= nlstate ; i ++) {
6681: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6682: else fprintf(ficgp," %%*lf (%%*lf)");
6683: }
1.242 brouard 6684: 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_"),k1-1,k1-1,nres);
1.238 brouard 6685: for (i=1; i<= nlstate ; i ++) {
6686: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6687: else fprintf(ficgp," %%*lf (%%*lf)");
6688: }
1.242 brouard 6689: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2==%d ? $3-1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres);
1.238 brouard 6690: for (i=1; i<= nlstate ; i ++) {
6691: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6692: else fprintf(ficgp," %%*lf (%%*lf)");
6693: }
6694: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6695: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6696: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
1.242 brouard 6697: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1, nres in 2 to be fixed */
1.238 brouard 6698: if(cptcoveff ==0){
1.245 brouard 6699: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
1.238 brouard 6700: }else{
6701: kl=0;
6702: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6703: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6704: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6705: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6706: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6707: vlv= nbcode[Tvaraff[k]][lv];
1.223 brouard 6708: kl++;
1.238 brouard 6709: /* 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 *\/ */
6710: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6711: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6712: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
6713: if(k==cptcoveff){
1.245 brouard 6714: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' w l lt 3",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
1.242 brouard 6715: 2+cptcoveff*2+(cpt-1), cpt ); /* 4 or 6 ?*/
1.238 brouard 6716: }else{
6717: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6718: kl++;
6719: }
6720: } /* end covariate */
6721: } /* end if no covariate */
6722: } /* end if backcast */
6723: fprintf(ficgp,"\nset out \n");
6724: } /* nres */
1.201 brouard 6725: } /* k1 */
6726: } /* cpt */
1.235 brouard 6727:
6728:
1.126 brouard 6729: /*2 eme*/
1.238 brouard 6730: for (k1=1; k1<= m ; k1 ++){
6731: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6732: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6733: continue;
6734: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
6735: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.225 brouard 6736: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6737: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6738: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6739: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6740: vlv= nbcode[Tvaraff[k]][lv];
6741: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6742: }
1.237 brouard 6743: /* for(k=1; k <= ncovds; k++){ */
1.236 brouard 6744: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 6745: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.236 brouard 6746: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6747: }
1.211 brouard 6748: fprintf(ficgp,"\n#\n");
1.223 brouard 6749: if(invalidvarcomb[k1]){
6750: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6751: continue;
6752: }
1.219 brouard 6753:
1.241 brouard 6754: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1,nres);
1.238 brouard 6755: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6756: if(vpopbased==0)
6757: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6758: else
6759: fprintf(ficgp,"\nreplot ");
6760: for (i=1; i<= nlstate+1 ; i ++) {
6761: k=2*i;
6762: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ?$4 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1, vpopbased);
6763: for (j=1; j<= nlstate+1 ; j ++) {
6764: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6765: else fprintf(ficgp," %%*lf (%%*lf)");
6766: }
6767: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6768: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
6769: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4-$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1,vpopbased);
6770: for (j=1; j<= nlstate+1 ; j ++) {
6771: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6772: else fprintf(ficgp," %%*lf (%%*lf)");
6773: }
6774: fprintf(ficgp,"\" t\"\" w l lt 0,");
6775: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4+$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1,vpopbased);
6776: for (j=1; j<= nlstate+1 ; j ++) {
6777: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6778: else fprintf(ficgp," %%*lf (%%*lf)");
6779: }
6780: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6781: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6782: } /* state */
6783: } /* vpopbased */
1.244 brouard 6784: fprintf(ficgp,"\nset out;set out \"%s_%d-%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1,nres); /* Buggy gnuplot */
1.238 brouard 6785: } /* end nres */
6786: } /* k1 end 2 eme*/
6787:
6788:
6789: /*3eme*/
6790: for (k1=1; k1<= m ; k1 ++){
6791: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6792: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6793: continue;
6794:
6795: for (cpt=1; cpt<= nlstate ; cpt ++) {
6796: fprintf(ficgp,"\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
6797: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6798: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6799: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6800: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6801: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6802: vlv= nbcode[Tvaraff[k]][lv];
6803: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6804: }
6805: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6806: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6807: }
6808: fprintf(ficgp,"\n#\n");
6809: if(invalidvarcomb[k1]){
6810: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6811: continue;
6812: }
6813:
6814: /* k=2+nlstate*(2*cpt-2); */
6815: k=2+(nlstate+1)*(cpt-1);
1.241 brouard 6816: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.238 brouard 6817: fprintf(ficgp,"set ter svg size 640, 480\n\
1.201 brouard 6818: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:%d t \"e%d1\" w l",ageminpar,fage,subdirf2(fileresu,"E_"),k1-1,k1-1,k,cpt);
1.238 brouard 6819: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6820: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6821: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6822: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6823: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6824: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6825:
1.238 brouard 6826: */
6827: for (i=1; i< nlstate ; i ++) {
6828: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+i,cpt,i+1);
6829: /* 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 6830:
1.238 brouard 6831: }
6832: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+nlstate,cpt);
6833: }
6834: } /* end nres */
6835: } /* end kl 3eme */
1.126 brouard 6836:
1.223 brouard 6837: /* 4eme */
1.201 brouard 6838: /* Survival functions (period) from state i in state j by initial state i */
1.238 brouard 6839: for (k1=1; k1<=m; k1++){ /* For each covariate and each value */
6840: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6841: if(m != 1 && TKresult[nres]!= k1)
1.223 brouard 6842: continue;
1.238 brouard 6843: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state cpt*/
6844: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
6845: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6846: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6847: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6848: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6849: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6850: vlv= nbcode[Tvaraff[k]][lv];
6851: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6852: }
6853: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6854: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6855: }
6856: fprintf(ficgp,"\n#\n");
6857: if(invalidvarcomb[k1]){
6858: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6859: continue;
1.223 brouard 6860: }
1.238 brouard 6861:
1.241 brouard 6862: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.238 brouard 6863: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6864: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6865: k=3;
6866: for (i=1; i<= nlstate ; i ++){
6867: if(i==1){
6868: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6869: }else{
6870: fprintf(ficgp,", '' ");
6871: }
6872: l=(nlstate+ndeath)*(i-1)+1;
6873: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6874: for (j=2; j<= nlstate+ndeath ; j ++)
6875: fprintf(ficgp,"+$%d",k+l+j-1);
6876: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
6877: } /* nlstate */
6878: fprintf(ficgp,"\nset out\n");
6879: } /* end cpt state*/
6880: } /* end nres */
6881: } /* end covariate k1 */
6882:
1.220 brouard 6883: /* 5eme */
1.201 brouard 6884: /* Survival functions (period) from state i in state j by final state j */
1.238 brouard 6885: for (k1=1; k1<= m ; k1++){ /* For each covariate combination if any */
6886: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6887: if(m != 1 && TKresult[nres]!= k1)
1.227 brouard 6888: continue;
1.238 brouard 6889: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
6890: 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);
6891: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6892: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6893: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6894: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6895: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6896: vlv= nbcode[Tvaraff[k]][lv];
6897: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6898: }
6899: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6900: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6901: }
6902: fprintf(ficgp,"\n#\n");
6903: if(invalidvarcomb[k1]){
6904: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6905: continue;
6906: }
1.227 brouard 6907:
1.241 brouard 6908: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.238 brouard 6909: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6910: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6911: k=3;
6912: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6913: if(j==1)
6914: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6915: else
6916: fprintf(ficgp,", '' ");
6917: l=(nlstate+ndeath)*(cpt-1) +j;
6918: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6919: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6920: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6921: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
6922: } /* nlstate */
6923: fprintf(ficgp,", '' ");
6924: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6925: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6926: l=(nlstate+ndeath)*(cpt-1) +j;
6927: if(j < nlstate)
6928: fprintf(ficgp,"$%d +",k+l);
6929: else
6930: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
6931: }
6932: fprintf(ficgp,"\nset out\n");
6933: } /* end cpt state*/
6934: } /* end covariate */
6935: } /* end nres */
1.227 brouard 6936:
1.220 brouard 6937: /* 6eme */
1.202 brouard 6938: /* CV preval stable (period) for each covariate */
1.237 brouard 6939: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6940: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6941: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6942: continue;
1.255 brouard 6943: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state of arrival */
1.227 brouard 6944:
1.211 brouard 6945: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6946: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6947: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6948: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6949: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6950: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6951: vlv= nbcode[Tvaraff[k]][lv];
6952: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6953: }
1.237 brouard 6954: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6955: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6956: }
1.211 brouard 6957: fprintf(ficgp,"\n#\n");
1.223 brouard 6958: if(invalidvarcomb[k1]){
1.227 brouard 6959: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6960: continue;
1.223 brouard 6961: }
1.227 brouard 6962:
1.241 brouard 6963: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.126 brouard 6964: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6965: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6966: k=3; /* Offset */
1.255 brouard 6967: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 6968: if(i==1)
6969: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6970: else
6971: fprintf(ficgp,", '' ");
1.255 brouard 6972: l=(nlstate+ndeath)*(i-1)+1; /* 1, 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 6973: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6974: for (j=2; j<= nlstate ; j ++)
6975: fprintf(ficgp,"+$%d",k+l+j-1);
6976: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6977: } /* nlstate */
1.201 brouard 6978: fprintf(ficgp,"\nset out\n");
1.153 brouard 6979: } /* end cpt state*/
6980: } /* end covariate */
1.227 brouard 6981:
6982:
1.220 brouard 6983: /* 7eme */
1.218 brouard 6984: if(backcast == 1){
1.217 brouard 6985: /* CV back preval stable (period) for each covariate */
1.237 brouard 6986: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6987: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6988: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6989: continue;
1.255 brouard 6990: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life ending state */
6991: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pijb' files, covariatecombination#=%d state=%d",k1, cpt);
1.227 brouard 6992: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6993: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6994: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6995: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 6996: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 6997: vlv= nbcode[Tvaraff[k]][lv];
6998: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6999: }
1.237 brouard 7000: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7001: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7002: }
1.227 brouard 7003: fprintf(ficgp,"\n#\n");
7004: if(invalidvarcomb[k1]){
7005: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7006: continue;
7007: }
7008:
1.241 brouard 7009: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.227 brouard 7010: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 7011: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7012: k=3; /* Offset */
1.255 brouard 7013: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 7014: if(i==1)
7015: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
7016: else
7017: fprintf(ficgp,", '' ");
7018: /* l=(nlstate+ndeath)*(i-1)+1; */
1.255 brouard 7019: l=(nlstate+ndeath)*(cpt-1)+1; /* fixed for i; cpt=1 1, cpt=2 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 7020: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
7021: /* 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 7022: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+i-1); /* To be verified */
1.227 brouard 7023: /* for (j=2; j<= nlstate ; j ++) */
7024: /* fprintf(ficgp,"+$%d",k+l+j-1); */
7025: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
7026: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
7027: } /* nlstate */
7028: fprintf(ficgp,"\nset out\n");
1.218 brouard 7029: } /* end cpt state*/
7030: } /* end covariate */
7031: } /* End if backcast */
7032:
1.223 brouard 7033: /* 8eme */
1.218 brouard 7034: if(prevfcast==1){
7035: /* Projection from cross-sectional to stable (period) for each covariate */
7036:
1.237 brouard 7037: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7038: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7039: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7040: continue;
1.211 brouard 7041: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 7042: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
7043: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
7044: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
7045: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7046: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7047: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7048: vlv= nbcode[Tvaraff[k]][lv];
7049: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
7050: }
1.237 brouard 7051: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7052: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7053: }
1.227 brouard 7054: fprintf(ficgp,"\n#\n");
7055: if(invalidvarcomb[k1]){
7056: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7057: continue;
7058: }
7059:
7060: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
1.241 brouard 7061: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.227 brouard 7062: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
1.238 brouard 7063: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7064: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
7065: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7066: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7067: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7068: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7069: if(i==1){
7070: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
7071: }else{
7072: fprintf(ficgp,",\\\n '' ");
7073: }
7074: if(cptcoveff ==0){ /* No covariate */
7075: ioffset=2; /* Age is in 2 */
7076: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7077: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7078: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7079: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7080: fprintf(ficgp," u %d:(", ioffset);
7081: if(i==nlstate+1)
7082: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
7083: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7084: else
7085: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
7086: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7087: }else{ /* more than 2 covariates */
7088: if(cptcoveff ==1){
7089: ioffset=4; /* Age is in 4 */
7090: }else{
7091: ioffset=6; /* Age is in 6 */
7092: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7093: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7094: }
7095: fprintf(ficgp," u %d:(",ioffset);
7096: kl=0;
7097: strcpy(gplotcondition,"(");
7098: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
7099: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
7100: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7101: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7102: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7103: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
7104: kl++;
7105: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
7106: kl++;
7107: if(k <cptcoveff && cptcoveff>1)
7108: sprintf(gplotcondition+strlen(gplotcondition)," && ");
7109: }
7110: strcpy(gplotcondition+strlen(gplotcondition),")");
7111: /* 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 *\/ */
7112: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
7113: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
7114: /* '' 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*/
7115: if(i==nlstate+1){
7116: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
7117: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7118: }else{
7119: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
7120: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7121: }
7122: } /* end if covariate */
7123: } /* nlstate */
7124: fprintf(ficgp,"\nset out\n");
1.223 brouard 7125: } /* end cpt state*/
7126: } /* end covariate */
7127: } /* End if prevfcast */
1.227 brouard 7128:
7129:
1.238 brouard 7130: /* 9eme writing MLE parameters */
7131: fprintf(ficgp,"\n##############\n#9eme MLE estimated parameters\n#############\n");
1.126 brouard 7132: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 7133: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 7134: for(k=1; k <=(nlstate+ndeath); k++){
7135: if (k != i) {
1.227 brouard 7136: fprintf(ficgp,"# current state %d\n",k);
7137: for(j=1; j <=ncovmodel; j++){
7138: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
7139: jk++;
7140: }
7141: fprintf(ficgp,"\n");
1.126 brouard 7142: }
7143: }
1.223 brouard 7144: }
1.187 brouard 7145: fprintf(ficgp,"##############\n#\n");
1.227 brouard 7146:
1.145 brouard 7147: /*goto avoid;*/
1.238 brouard 7148: /* 10eme Graphics of probabilities or incidences using written MLE parameters */
7149: fprintf(ficgp,"\n##############\n#10eme Graphics of probabilities or incidences\n#############\n");
1.187 brouard 7150: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
7151: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
7152: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
7153: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
7154: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7155: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7156: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7157: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7158: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
7159: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7160: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
7161: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
7162: fprintf(ficgp,"#\n");
1.223 brouard 7163: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
1.238 brouard 7164: fprintf(ficgp,"#Number of graphics: first is logit, 2nd is probabilities, third is incidences per year\n");
1.237 brouard 7165: fprintf(ficgp,"#model=%s \n",model);
1.238 brouard 7166: fprintf(ficgp,"# Type of graphic ng=%d\n",ng);
1.237 brouard 7167: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);/* to be checked */
7168: for(jk=1; jk <=m; jk++) /* For each combination of covariate */
7169: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7170: if(m != 1 && TKresult[nres]!= jk)
1.237 brouard 7171: continue;
7172: fprintf(ficgp,"# Combination of dummy jk=%d and ",jk);
7173: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7174: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7175: }
7176: fprintf(ficgp,"\n#\n");
1.241 brouard 7177: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng,nres);
1.223 brouard 7178: fprintf(ficgp,"\nset ter svg size 640, 480 ");
7179: if (ng==1){
7180: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
7181: fprintf(ficgp,"\nunset log y");
7182: }else if (ng==2){
7183: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
7184: fprintf(ficgp,"\nset log y");
7185: }else if (ng==3){
7186: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
7187: fprintf(ficgp,"\nset log y");
7188: }else
7189: fprintf(ficgp,"\nunset title ");
7190: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
7191: i=1;
7192: for(k2=1; k2<=nlstate; k2++) {
7193: k3=i;
7194: for(k=1; k<=(nlstate+ndeath); k++) {
7195: if (k != k2){
7196: switch( ng) {
7197: case 1:
7198: if(nagesqr==0)
7199: fprintf(ficgp," p%d+p%d*x",i,i+1);
7200: else /* nagesqr =1 */
7201: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7202: break;
7203: case 2: /* ng=2 */
7204: if(nagesqr==0)
7205: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
7206: else /* nagesqr =1 */
7207: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7208: break;
7209: case 3:
7210: if(nagesqr==0)
7211: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
7212: else /* nagesqr =1 */
7213: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
7214: break;
7215: }
7216: ij=1;/* To be checked else nbcode[0][0] wrong */
1.237 brouard 7217: ijp=1; /* product no age */
7218: /* for(j=3; j <=ncovmodel-nagesqr; j++) { */
7219: for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */
1.223 brouard 7220: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
1.237 brouard 7221: if(j==Tage[ij]) { /* Product by age */
7222: if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */
1.238 brouard 7223: if(DummyV[j]==0){
1.237 brouard 7224: fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);;
7225: }else{ /* quantitative */
7226: fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */
7227: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7228: }
7229: ij++;
7230: }
7231: }else if(j==Tprod[ijp]) { /* */
7232: /* printf("Tprod[%d]=%d, j=%d\n", ij, Tprod[ijp], j); */
7233: if(ijp <=cptcovprod) { /* Product */
1.238 brouard 7234: if(DummyV[Tvard[ijp][1]]==0){/* Vn is dummy */
7235: if(DummyV[Tvard[ijp][2]]==0){/* Vn and Vm are dummy */
1.237 brouard 7236: /* 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)]); */
7237: fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tinvresult[nres][Tvard[ijp][2]]);
7238: }else{ /* Vn is dummy and Vm is quanti */
7239: /* fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],Tqinvresult[nres][Tvard[ijp][2]]); */
7240: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7241: }
7242: }else{ /* Vn*Vm Vn is quanti */
1.238 brouard 7243: if(DummyV[Tvard[ijp][2]]==0){
1.237 brouard 7244: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][2]],Tqinvresult[nres][Tvard[ijp][1]]);
7245: }else{ /* Both quanti */
7246: fprintf(ficgp,"+p%d*%f*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7247: }
7248: }
1.238 brouard 7249: ijp++;
1.237 brouard 7250: }
7251: } else{ /* simple covariate */
7252: /* fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,nbcode[Tvar[j]][codtabm(jk,j)]); /\* Valgrind bug nbcode *\/ */
7253: if(Dummy[j]==0){
7254: fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]); /* */
7255: }else{ /* quantitative */
7256: fprintf(ficgp,"+p%d*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* */
1.223 brouard 7257: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7258: }
1.237 brouard 7259: } /* end simple */
7260: } /* end j */
1.223 brouard 7261: }else{
7262: i=i-ncovmodel;
7263: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
7264: fprintf(ficgp," (1.");
7265: }
1.227 brouard 7266:
1.223 brouard 7267: if(ng != 1){
7268: fprintf(ficgp,")/(1");
1.227 brouard 7269:
1.223 brouard 7270: for(k1=1; k1 <=nlstate; k1++){
7271: if(nagesqr==0)
7272: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
7273: else /* nagesqr =1 */
7274: 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 7275:
1.223 brouard 7276: ij=1;
7277: for(j=3; j <=ncovmodel-nagesqr; j++){
1.237 brouard 7278: if((j-2)==Tage[ij]) { /* Bug valgrind */
7279: if(ij <=cptcovage) { /* Bug valgrind */
1.223 brouard 7280: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
7281: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7282: ij++;
7283: }
7284: }
7285: else
1.225 brouard 7286: 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 7287: }
7288: fprintf(ficgp,")");
7289: }
7290: fprintf(ficgp,")");
7291: if(ng ==2)
7292: fprintf(ficgp," t \"p%d%d\" ", k2,k);
7293: else /* ng= 3 */
7294: fprintf(ficgp," t \"i%d%d\" ", k2,k);
7295: }else{ /* end ng <> 1 */
7296: if( k !=k2) /* logit p11 is hard to draw */
7297: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
7298: }
7299: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
7300: fprintf(ficgp,",");
7301: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
7302: fprintf(ficgp,",");
7303: i=i+ncovmodel;
7304: } /* end k */
7305: } /* end k2 */
7306: fprintf(ficgp,"\n set out\n");
7307: } /* end jk */
7308: } /* end ng */
7309: /* avoid: */
7310: fflush(ficgp);
1.126 brouard 7311: } /* end gnuplot */
7312:
7313:
7314: /*************** Moving average **************/
1.219 brouard 7315: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 7316: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 7317:
1.222 brouard 7318: int i, cpt, cptcod;
7319: int modcovmax =1;
7320: int mobilavrange, mob;
7321: int iage=0;
7322:
7323: double sum=0.;
7324: double age;
7325: double *sumnewp, *sumnewm;
7326: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
7327:
7328:
1.225 brouard 7329: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 7330: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
7331:
7332: sumnewp = vector(1,ncovcombmax);
7333: sumnewm = vector(1,ncovcombmax);
7334: agemingood = vector(1,ncovcombmax);
7335: agemaxgood = vector(1,ncovcombmax);
7336:
7337: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7338: sumnewm[cptcod]=0.;
7339: sumnewp[cptcod]=0.;
7340: agemingood[cptcod]=0;
7341: agemaxgood[cptcod]=0;
7342: }
7343: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
7344:
7345: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
7346: if(mobilav==1) mobilavrange=5; /* default */
7347: else mobilavrange=mobilav;
7348: for (age=bage; age<=fage; age++)
7349: for (i=1; i<=nlstate;i++)
7350: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
7351: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
7352: /* We keep the original values on the extreme ages bage, fage and for
7353: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
7354: we use a 5 terms etc. until the borders are no more concerned.
7355: */
7356: for (mob=3;mob <=mobilavrange;mob=mob+2){
7357: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
7358: for (i=1; i<=nlstate;i++){
7359: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7360: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
7361: for (cpt=1;cpt<=(mob-1)/2;cpt++){
7362: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
7363: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
7364: }
7365: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
7366: }
7367: }
7368: }/* end age */
7369: }/* end mob */
7370: }else
7371: return -1;
7372: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7373: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
7374: if(invalidvarcomb[cptcod]){
7375: printf("\nCombination (%d) ignored because no cases \n",cptcod);
7376: continue;
7377: }
1.219 brouard 7378:
1.222 brouard 7379: agemingood[cptcod]=fage-(mob-1)/2;
7380: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
7381: sumnewm[cptcod]=0.;
7382: for (i=1; i<=nlstate;i++){
7383: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7384: }
7385: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7386: agemingood[cptcod]=age;
7387: }else{ /* bad */
7388: for (i=1; i<=nlstate;i++){
7389: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7390: } /* i */
7391: } /* end bad */
7392: }/* age */
7393: sum=0.;
7394: for (i=1; i<=nlstate;i++){
7395: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7396: }
7397: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7398: 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);
7399: /* for (i=1; i<=nlstate;i++){ */
7400: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7401: /* } /\* i *\/ */
7402: } /* end bad */
7403: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
7404: /* From youngest, finding the oldest wrong */
7405: agemaxgood[cptcod]=bage+(mob-1)/2;
7406: for (age=bage+(mob-1)/2; age<=fage; age++){
7407: sumnewm[cptcod]=0.;
7408: for (i=1; i<=nlstate;i++){
7409: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7410: }
7411: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7412: agemaxgood[cptcod]=age;
7413: }else{ /* bad */
7414: for (i=1; i<=nlstate;i++){
7415: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7416: } /* i */
7417: } /* end bad */
7418: }/* age */
7419: sum=0.;
7420: for (i=1; i<=nlstate;i++){
7421: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7422: }
7423: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7424: 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);
7425: /* for (i=1; i<=nlstate;i++){ */
7426: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7427: /* } /\* i *\/ */
7428: } /* end bad */
7429:
7430: for (age=bage; age<=fage; age++){
1.235 brouard 7431: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 7432: sumnewp[cptcod]=0.;
7433: sumnewm[cptcod]=0.;
7434: for (i=1; i<=nlstate;i++){
7435: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
7436: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7437: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
7438: }
7439: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
7440: }
7441: /* printf("\n"); */
7442: /* } */
7443: /* brutal averaging */
7444: for (i=1; i<=nlstate;i++){
7445: for (age=1; age<=bage; age++){
7446: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7447: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7448: }
7449: for (age=fage; age<=AGESUP; age++){
7450: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7451: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7452: }
7453: } /* end i status */
7454: for (i=nlstate+1; i<=nlstate+ndeath;i++){
7455: for (age=1; age<=AGESUP; age++){
7456: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
7457: mobaverage[(int)age][i][cptcod]=0.;
7458: }
7459: }
7460: }/* end cptcod */
7461: free_vector(sumnewm,1, ncovcombmax);
7462: free_vector(sumnewp,1, ncovcombmax);
7463: free_vector(agemaxgood,1, ncovcombmax);
7464: free_vector(agemingood,1, ncovcombmax);
7465: return 0;
7466: }/* End movingaverage */
1.218 brouard 7467:
1.126 brouard 7468:
7469: /************** Forecasting ******************/
1.235 brouard 7470: 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 7471: /* proj1, year, month, day of starting projection
7472: agemin, agemax range of age
7473: dateprev1 dateprev2 range of dates during which prevalence is computed
7474: anproj2 year of en of projection (same day and month as proj1).
7475: */
1.235 brouard 7476: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7477: double agec; /* generic age */
7478: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7479: double *popeffectif,*popcount;
7480: double ***p3mat;
1.218 brouard 7481: /* double ***mobaverage; */
1.126 brouard 7482: char fileresf[FILENAMELENGTH];
7483:
7484: agelim=AGESUP;
1.211 brouard 7485: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7486: in each health status at the date of interview (if between dateprev1 and dateprev2).
7487: We still use firstpass and lastpass as another selection.
7488: */
1.214 brouard 7489: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7490: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7491:
1.201 brouard 7492: strcpy(fileresf,"F_");
7493: strcat(fileresf,fileresu);
1.126 brouard 7494: if((ficresf=fopen(fileresf,"w"))==NULL) {
7495: printf("Problem with forecast resultfile: %s\n", fileresf);
7496: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7497: }
1.235 brouard 7498: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7499: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7500:
1.225 brouard 7501: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7502:
7503:
7504: stepsize=(int) (stepm+YEARM-1)/YEARM;
7505: if (stepm<=12) stepsize=1;
7506: if(estepm < stepm){
7507: printf ("Problem %d lower than %d\n",estepm, stepm);
7508: }
7509: else hstepm=estepm;
7510:
7511: hstepm=hstepm/stepm;
7512: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7513: fractional in yp1 */
7514: anprojmean=yp;
7515: yp2=modf((yp1*12),&yp);
7516: mprojmean=yp;
7517: yp1=modf((yp2*30.5),&yp);
7518: jprojmean=yp;
7519: if(jprojmean==0) jprojmean=1;
7520: if(mprojmean==0) jprojmean=1;
7521:
1.227 brouard 7522: i1=pow(2,cptcoveff);
1.126 brouard 7523: if (cptcovn < 1){i1=1;}
7524:
7525: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7526:
7527: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7528:
1.126 brouard 7529: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7530: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7531: for(k=1; k<=i1;k++){
1.253 brouard 7532: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 7533: continue;
1.227 brouard 7534: if(invalidvarcomb[k]){
7535: printf("\nCombination (%d) projection ignored because no cases \n",k);
7536: continue;
7537: }
7538: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7539: for(j=1;j<=cptcoveff;j++) {
7540: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7541: }
1.235 brouard 7542: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 7543: fprintf(ficresf," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.235 brouard 7544: }
1.227 brouard 7545: fprintf(ficresf," yearproj age");
7546: for(j=1; j<=nlstate+ndeath;j++){
7547: for(i=1; i<=nlstate;i++)
7548: fprintf(ficresf," p%d%d",i,j);
7549: fprintf(ficresf," wp.%d",j);
7550: }
7551: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7552: fprintf(ficresf,"\n");
7553: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7554: for (agec=fage; agec>=(ageminpar-1); agec--){
7555: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7556: nhstepm = nhstepm/hstepm;
7557: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7558: oldm=oldms;savm=savms;
1.235 brouard 7559: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7560:
7561: for (h=0; h<=nhstepm; h++){
7562: if (h*hstepm/YEARM*stepm ==yearp) {
7563: fprintf(ficresf,"\n");
7564: for(j=1;j<=cptcoveff;j++)
7565: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7566: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7567: }
7568: for(j=1; j<=nlstate+ndeath;j++) {
7569: ppij=0.;
7570: for(i=1; i<=nlstate;i++) {
7571: if (mobilav==1)
7572: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7573: else {
7574: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7575: }
7576: if (h*hstepm/YEARM*stepm== yearp) {
7577: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7578: }
7579: } /* end i */
7580: if (h*hstepm/YEARM*stepm==yearp) {
7581: fprintf(ficresf," %.3f", ppij);
7582: }
7583: }/* end j */
7584: } /* end h */
7585: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7586: } /* end agec */
7587: } /* end yearp */
7588: } /* end k */
1.219 brouard 7589:
1.126 brouard 7590: fclose(ficresf);
1.215 brouard 7591: printf("End of Computing forecasting \n");
7592: fprintf(ficlog,"End of Computing forecasting\n");
7593:
1.126 brouard 7594: }
7595:
1.218 brouard 7596: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7597: /* 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 7598: /* /\* back1, year, month, day of starting backection */
7599: /* agemin, agemax range of age */
7600: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7601: /* anback2 year of en of backection (same day and month as back1). */
7602: /* *\/ */
7603: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7604: /* double agec; /\* generic age *\/ */
7605: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7606: /* double *popeffectif,*popcount; */
7607: /* double ***p3mat; */
7608: /* /\* double ***mobaverage; *\/ */
7609: /* char fileresfb[FILENAMELENGTH]; */
7610:
7611: /* agelim=AGESUP; */
7612: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7613: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7614: /* We still use firstpass and lastpass as another selection. */
7615: /* *\/ */
7616: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7617: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7618: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7619:
7620: /* strcpy(fileresfb,"FB_"); */
7621: /* strcat(fileresfb,fileresu); */
7622: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7623: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7624: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7625: /* } */
7626: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7627: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7628:
1.225 brouard 7629: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7630:
7631: /* /\* if (mobilav!=0) { *\/ */
7632: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7633: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7634: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7635: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7636: /* /\* } *\/ */
7637: /* /\* } *\/ */
7638:
7639: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7640: /* if (stepm<=12) stepsize=1; */
7641: /* if(estepm < stepm){ */
7642: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7643: /* } */
7644: /* else hstepm=estepm; */
7645:
7646: /* hstepm=hstepm/stepm; */
7647: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7648: /* fractional in yp1 *\/ */
7649: /* anprojmean=yp; */
7650: /* yp2=modf((yp1*12),&yp); */
7651: /* mprojmean=yp; */
7652: /* yp1=modf((yp2*30.5),&yp); */
7653: /* jprojmean=yp; */
7654: /* if(jprojmean==0) jprojmean=1; */
7655: /* if(mprojmean==0) jprojmean=1; */
7656:
1.225 brouard 7657: /* i1=cptcoveff; */
1.218 brouard 7658: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7659:
1.218 brouard 7660: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7661:
1.218 brouard 7662: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7663:
7664: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7665: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7666: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7667: /* k=k+1; */
7668: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7669: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7670: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7671: /* } */
7672: /* fprintf(ficresfb," yearbproj age"); */
7673: /* for(j=1; j<=nlstate+ndeath;j++){ */
7674: /* for(i=1; i<=nlstate;i++) */
7675: /* fprintf(ficresfb," p%d%d",i,j); */
7676: /* fprintf(ficresfb," p.%d",j); */
7677: /* } */
7678: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7679: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7680: /* fprintf(ficresfb,"\n"); */
7681: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7682: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7683: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7684: /* nhstepm = nhstepm/hstepm; */
7685: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7686: /* oldm=oldms;savm=savms; */
7687: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7688: /* for (h=0; h<=nhstepm; h++){ */
7689: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7690: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7691: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7692: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7693: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7694: /* } */
7695: /* for(j=1; j<=nlstate+ndeath;j++) { */
7696: /* ppij=0.; */
7697: /* for(i=1; i<=nlstate;i++) { */
7698: /* if (mobilav==1) */
7699: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7700: /* else { */
7701: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7702: /* } */
7703: /* if (h*hstepm/YEARM*stepm== yearp) { */
7704: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7705: /* } */
7706: /* } /\* end i *\/ */
7707: /* if (h*hstepm/YEARM*stepm==yearp) { */
7708: /* fprintf(ficresfb," %.3f", ppij); */
7709: /* } */
7710: /* }/\* end j *\/ */
7711: /* } /\* end h *\/ */
7712: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7713: /* } /\* end agec *\/ */
7714: /* } /\* end yearp *\/ */
7715: /* } /\* end cptcod *\/ */
7716: /* } /\* end cptcov *\/ */
7717:
7718: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7719:
7720: /* fclose(ficresfb); */
7721: /* printf("End of Computing Back forecasting \n"); */
7722: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7723:
1.218 brouard 7724: /* } */
1.217 brouard 7725:
1.126 brouard 7726: /************** Forecasting *****not tested NB*************/
1.227 brouard 7727: /* 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 7728:
1.227 brouard 7729: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7730: /* int *popage; */
7731: /* double calagedatem, agelim, kk1, kk2; */
7732: /* double *popeffectif,*popcount; */
7733: /* double ***p3mat,***tabpop,***tabpopprev; */
7734: /* /\* double ***mobaverage; *\/ */
7735: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7736:
1.227 brouard 7737: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7738: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7739: /* agelim=AGESUP; */
7740: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7741:
1.227 brouard 7742: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7743:
7744:
1.227 brouard 7745: /* strcpy(filerespop,"POP_"); */
7746: /* strcat(filerespop,fileresu); */
7747: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7748: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7749: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7750: /* } */
7751: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7752: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7753:
1.227 brouard 7754: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7755:
1.227 brouard 7756: /* /\* if (mobilav!=0) { *\/ */
7757: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7758: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7759: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7760: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7761: /* /\* } *\/ */
7762: /* /\* } *\/ */
1.126 brouard 7763:
1.227 brouard 7764: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7765: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7766:
1.227 brouard 7767: /* agelim=AGESUP; */
1.126 brouard 7768:
1.227 brouard 7769: /* hstepm=1; */
7770: /* hstepm=hstepm/stepm; */
1.218 brouard 7771:
1.227 brouard 7772: /* if (popforecast==1) { */
7773: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7774: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7775: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7776: /* } */
7777: /* popage=ivector(0,AGESUP); */
7778: /* popeffectif=vector(0,AGESUP); */
7779: /* popcount=vector(0,AGESUP); */
1.126 brouard 7780:
1.227 brouard 7781: /* i=1; */
7782: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7783:
1.227 brouard 7784: /* imx=i; */
7785: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7786: /* } */
1.218 brouard 7787:
1.227 brouard 7788: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7789: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7790: /* k=k+1; */
7791: /* fprintf(ficrespop,"\n#******"); */
7792: /* for(j=1;j<=cptcoveff;j++) { */
7793: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7794: /* } */
7795: /* fprintf(ficrespop,"******\n"); */
7796: /* fprintf(ficrespop,"# Age"); */
7797: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7798: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7799:
1.227 brouard 7800: /* for (cpt=0; cpt<=0;cpt++) { */
7801: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7802:
1.227 brouard 7803: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7804: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7805: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7806:
1.227 brouard 7807: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7808: /* oldm=oldms;savm=savms; */
7809: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7810:
1.227 brouard 7811: /* for (h=0; h<=nhstepm; h++){ */
7812: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7813: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7814: /* } */
7815: /* for(j=1; j<=nlstate+ndeath;j++) { */
7816: /* kk1=0.;kk2=0; */
7817: /* for(i=1; i<=nlstate;i++) { */
7818: /* if (mobilav==1) */
7819: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7820: /* else { */
7821: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7822: /* } */
7823: /* } */
7824: /* if (h==(int)(calagedatem+12*cpt)){ */
7825: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7826: /* /\*fprintf(ficrespop," %.3f", kk1); */
7827: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7828: /* } */
7829: /* } */
7830: /* for(i=1; i<=nlstate;i++){ */
7831: /* kk1=0.; */
7832: /* for(j=1; j<=nlstate;j++){ */
7833: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7834: /* } */
7835: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7836: /* } */
1.218 brouard 7837:
1.227 brouard 7838: /* if (h==(int)(calagedatem+12*cpt)) */
7839: /* for(j=1; j<=nlstate;j++) */
7840: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7841: /* } */
7842: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7843: /* } */
7844: /* } */
1.218 brouard 7845:
1.227 brouard 7846: /* /\******\/ */
1.218 brouard 7847:
1.227 brouard 7848: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7849: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7850: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7851: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7852: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7853:
1.227 brouard 7854: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7855: /* oldm=oldms;savm=savms; */
7856: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7857: /* for (h=0; h<=nhstepm; h++){ */
7858: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7859: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7860: /* } */
7861: /* for(j=1; j<=nlstate+ndeath;j++) { */
7862: /* kk1=0.;kk2=0; */
7863: /* for(i=1; i<=nlstate;i++) { */
7864: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7865: /* } */
7866: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7867: /* } */
7868: /* } */
7869: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7870: /* } */
7871: /* } */
7872: /* } */
7873: /* } */
1.218 brouard 7874:
1.227 brouard 7875: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7876:
1.227 brouard 7877: /* if (popforecast==1) { */
7878: /* free_ivector(popage,0,AGESUP); */
7879: /* free_vector(popeffectif,0,AGESUP); */
7880: /* free_vector(popcount,0,AGESUP); */
7881: /* } */
7882: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7883: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7884: /* fclose(ficrespop); */
7885: /* } /\* End of popforecast *\/ */
1.218 brouard 7886:
1.126 brouard 7887: int fileappend(FILE *fichier, char *optionfich)
7888: {
7889: if((fichier=fopen(optionfich,"a"))==NULL) {
7890: printf("Problem with file: %s\n", optionfich);
7891: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7892: return (0);
7893: }
7894: fflush(fichier);
7895: return (1);
7896: }
7897:
7898:
7899: /**************** function prwizard **********************/
7900: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7901: {
7902:
7903: /* Wizard to print covariance matrix template */
7904:
1.164 brouard 7905: char ca[32], cb[32];
7906: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7907: int numlinepar;
7908:
7909: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7910: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7911: for(i=1; i <=nlstate; i++){
7912: jj=0;
7913: for(j=1; j <=nlstate+ndeath; j++){
7914: if(j==i) continue;
7915: jj++;
7916: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7917: printf("%1d%1d",i,j);
7918: fprintf(ficparo,"%1d%1d",i,j);
7919: for(k=1; k<=ncovmodel;k++){
7920: /* printf(" %lf",param[i][j][k]); */
7921: /* fprintf(ficparo," %lf",param[i][j][k]); */
7922: printf(" 0.");
7923: fprintf(ficparo," 0.");
7924: }
7925: printf("\n");
7926: fprintf(ficparo,"\n");
7927: }
7928: }
7929: printf("# Scales (for hessian or gradient estimation)\n");
7930: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7931: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
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: fprintf(ficparo,"%1d%1d",i,j);
7938: printf("%1d%1d",i,j);
7939: fflush(stdout);
7940: for(k=1; k<=ncovmodel;k++){
7941: /* printf(" %le",delti3[i][j][k]); */
7942: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7943: printf(" 0.");
7944: fprintf(ficparo," 0.");
7945: }
7946: numlinepar++;
7947: printf("\n");
7948: fprintf(ficparo,"\n");
7949: }
7950: }
7951: printf("# Covariance matrix\n");
7952: /* # 121 Var(a12)\n\ */
7953: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7954: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7955: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7956: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7957: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7958: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7959: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7960: fflush(stdout);
7961: fprintf(ficparo,"# Covariance matrix\n");
7962: /* # 121 Var(a12)\n\ */
7963: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7964: /* # ...\n\ */
7965: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7966:
7967: for(itimes=1;itimes<=2;itimes++){
7968: jj=0;
7969: for(i=1; i <=nlstate; i++){
7970: for(j=1; j <=nlstate+ndeath; j++){
7971: if(j==i) continue;
7972: for(k=1; k<=ncovmodel;k++){
7973: jj++;
7974: ca[0]= k+'a'-1;ca[1]='\0';
7975: if(itimes==1){
7976: printf("#%1d%1d%d",i,j,k);
7977: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7978: }else{
7979: printf("%1d%1d%d",i,j,k);
7980: fprintf(ficparo,"%1d%1d%d",i,j,k);
7981: /* printf(" %.5le",matcov[i][j]); */
7982: }
7983: ll=0;
7984: for(li=1;li <=nlstate; li++){
7985: for(lj=1;lj <=nlstate+ndeath; lj++){
7986: if(lj==li) continue;
7987: for(lk=1;lk<=ncovmodel;lk++){
7988: ll++;
7989: if(ll<=jj){
7990: cb[0]= lk +'a'-1;cb[1]='\0';
7991: if(ll<jj){
7992: if(itimes==1){
7993: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7994: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7995: }else{
7996: printf(" 0.");
7997: fprintf(ficparo," 0.");
7998: }
7999: }else{
8000: if(itimes==1){
8001: printf(" Var(%s%1d%1d)",ca,i,j);
8002: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
8003: }else{
8004: printf(" 0.");
8005: fprintf(ficparo," 0.");
8006: }
8007: }
8008: }
8009: } /* end lk */
8010: } /* end lj */
8011: } /* end li */
8012: printf("\n");
8013: fprintf(ficparo,"\n");
8014: numlinepar++;
8015: } /* end k*/
8016: } /*end j */
8017: } /* end i */
8018: } /* end itimes */
8019:
8020: } /* end of prwizard */
8021: /******************* Gompertz Likelihood ******************************/
8022: double gompertz(double x[])
8023: {
8024: double A,B,L=0.0,sump=0.,num=0.;
8025: int i,n=0; /* n is the size of the sample */
8026:
1.220 brouard 8027: for (i=1;i<=imx ; i++) {
1.126 brouard 8028: sump=sump+weight[i];
8029: /* sump=sump+1;*/
8030: num=num+1;
8031: }
8032:
8033:
8034: /* for (i=0; i<=imx; i++)
8035: 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]);*/
8036:
8037: for (i=1;i<=imx ; i++)
8038: {
8039: if (cens[i] == 1 && wav[i]>1)
8040: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
8041:
8042: if (cens[i] == 0 && wav[i]>1)
8043: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
8044: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
8045:
8046: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8047: if (wav[i] > 1 ) { /* ??? */
8048: L=L+A*weight[i];
8049: /* 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]);*/
8050: }
8051: }
8052:
8053: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8054:
8055: return -2*L*num/sump;
8056: }
8057:
1.136 brouard 8058: #ifdef GSL
8059: /******************* Gompertz_f Likelihood ******************************/
8060: double gompertz_f(const gsl_vector *v, void *params)
8061: {
8062: double A,B,LL=0.0,sump=0.,num=0.;
8063: double *x= (double *) v->data;
8064: int i,n=0; /* n is the size of the sample */
8065:
8066: for (i=0;i<=imx-1 ; i++) {
8067: sump=sump+weight[i];
8068: /* sump=sump+1;*/
8069: num=num+1;
8070: }
8071:
8072:
8073: /* for (i=0; i<=imx; i++)
8074: 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]);*/
8075: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
8076: for (i=1;i<=imx ; i++)
8077: {
8078: if (cens[i] == 1 && wav[i]>1)
8079: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
8080:
8081: if (cens[i] == 0 && wav[i]>1)
8082: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
8083: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
8084:
8085: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8086: if (wav[i] > 1 ) { /* ??? */
8087: LL=LL+A*weight[i];
8088: /* 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]);*/
8089: }
8090: }
8091:
8092: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8093: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
8094:
8095: return -2*LL*num/sump;
8096: }
8097: #endif
8098:
1.126 brouard 8099: /******************* Printing html file ***********/
1.201 brouard 8100: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 8101: int lastpass, int stepm, int weightopt, char model[],\
8102: int imx, double p[],double **matcov,double agemortsup){
8103: int i,k;
8104:
8105: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
8106: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
8107: for (i=1;i<=2;i++)
8108: 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 8109: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 8110: fprintf(fichtm,"</ul>");
8111:
8112: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
8113:
8114: 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>");
8115:
8116: for (k=agegomp;k<(agemortsup-2);k++)
8117: 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]);
8118:
8119:
8120: fflush(fichtm);
8121: }
8122:
8123: /******************* Gnuplot file **************/
1.201 brouard 8124: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 8125:
8126: char dirfileres[132],optfileres[132];
1.164 brouard 8127:
1.126 brouard 8128: int ng;
8129:
8130:
8131: /*#ifdef windows */
8132: fprintf(ficgp,"cd \"%s\" \n",pathc);
8133: /*#endif */
8134:
8135:
8136: strcpy(dirfileres,optionfilefiname);
8137: strcpy(optfileres,"vpl");
1.199 brouard 8138: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 8139: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 8140: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 8141: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 8142: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
8143:
8144: }
8145:
1.136 brouard 8146: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
8147: {
1.126 brouard 8148:
1.136 brouard 8149: /*-------- data file ----------*/
8150: FILE *fic;
8151: char dummy[]=" ";
1.240 brouard 8152: int i=0, j=0, n=0, iv=0, v;
1.223 brouard 8153: int lstra;
1.136 brouard 8154: int linei, month, year,iout;
8155: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 8156: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 8157: char *stratrunc;
1.223 brouard 8158:
1.240 brouard 8159: DummyV=ivector(1,NCOVMAX); /* 1 to 3 */
8160: FixedV=ivector(1,NCOVMAX); /* 1 to 3 */
1.126 brouard 8161:
1.240 brouard 8162: for(v=1; v <=ncovcol;v++){
8163: DummyV[v]=0;
8164: FixedV[v]=0;
8165: }
8166: for(v=ncovcol+1; v <=ncovcol+nqv;v++){
8167: DummyV[v]=1;
8168: FixedV[v]=0;
8169: }
8170: for(v=ncovcol+nqv+1; v <=ncovcol+nqv+ntv;v++){
8171: DummyV[v]=0;
8172: FixedV[v]=1;
8173: }
8174: for(v=ncovcol+nqv+ntv+1; v <=ncovcol+nqv+ntv+nqtv;v++){
8175: DummyV[v]=1;
8176: FixedV[v]=1;
8177: }
8178: for(v=1; v <=ncovcol+nqv+ntv+nqtv;v++){
8179: printf("Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
8180: 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]);
8181: }
1.126 brouard 8182:
1.136 brouard 8183: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 8184: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
8185: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 8186: }
1.126 brouard 8187:
1.136 brouard 8188: i=1;
8189: linei=0;
8190: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
8191: linei=linei+1;
8192: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
8193: if(line[j] == '\t')
8194: line[j] = ' ';
8195: }
8196: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
8197: ;
8198: };
8199: line[j+1]=0; /* Trims blanks at end of line */
8200: if(line[0]=='#'){
8201: fprintf(ficlog,"Comment line\n%s\n",line);
8202: printf("Comment line\n%s\n",line);
8203: continue;
8204: }
8205: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 8206: strcpy(line, linetmp);
1.223 brouard 8207:
8208: /* Loops on waves */
8209: for (j=maxwav;j>=1;j--){
8210: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.238 brouard 8211: cutv(stra, strb, line, ' ');
8212: if(strb[0]=='.') { /* Missing value */
8213: lval=-1;
8214: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
8215: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
8216: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
8217: 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);
8218: 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);
8219: return 1;
8220: }
8221: }else{
8222: errno=0;
8223: /* what_kind_of_number(strb); */
8224: dval=strtod(strb,&endptr);
8225: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
8226: /* if(strb != endptr && *endptr == '\0') */
8227: /* dval=dlval; */
8228: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8229: if( strb[0]=='\0' || (*endptr != '\0')){
8230: 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);
8231: 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);
8232: return 1;
8233: }
8234: cotqvar[j][iv][i]=dval;
8235: cotvar[j][ntv+iv][i]=dval;
8236: }
8237: strcpy(line,stra);
1.223 brouard 8238: }/* end loop ntqv */
1.225 brouard 8239:
1.223 brouard 8240: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.238 brouard 8241: cutv(stra, strb, line, ' ');
8242: if(strb[0]=='.') { /* Missing value */
8243: lval=-1;
8244: }else{
8245: errno=0;
8246: lval=strtol(strb,&endptr,10);
8247: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8248: if( strb[0]=='\0' || (*endptr != '\0')){
8249: 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);
8250: 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);
8251: return 1;
8252: }
8253: }
8254: if(lval <-1 || lval >1){
8255: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8256: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8257: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8258: For example, for multinomial values like 1, 2 and 3,\n \
8259: build V1=0 V2=0 for the reference value (1),\n \
8260: V1=1 V2=0 for (2) \n \
1.223 brouard 8261: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8262: output of IMaCh is often meaningless.\n \
1.223 brouard 8263: Exiting.\n",lval,linei, i,line,j);
1.238 brouard 8264: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8265: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8266: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8267: For example, for multinomial values like 1, 2 and 3,\n \
8268: build V1=0 V2=0 for the reference value (1),\n \
8269: V1=1 V2=0 for (2) \n \
1.223 brouard 8270: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8271: output of IMaCh is often meaningless.\n \
1.223 brouard 8272: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.238 brouard 8273: return 1;
8274: }
8275: cotvar[j][iv][i]=(double)(lval);
8276: strcpy(line,stra);
1.223 brouard 8277: }/* end loop ntv */
1.225 brouard 8278:
1.223 brouard 8279: /* Statuses at wave */
1.137 brouard 8280: cutv(stra, strb, line, ' ');
1.223 brouard 8281: if(strb[0]=='.') { /* Missing value */
1.238 brouard 8282: lval=-1;
1.136 brouard 8283: }else{
1.238 brouard 8284: errno=0;
8285: lval=strtol(strb,&endptr,10);
8286: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8287: if( strb[0]=='\0' || (*endptr != '\0')){
8288: 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);
8289: 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);
8290: return 1;
8291: }
1.136 brouard 8292: }
1.225 brouard 8293:
1.136 brouard 8294: s[j][i]=lval;
1.225 brouard 8295:
1.223 brouard 8296: /* Date of Interview */
1.136 brouard 8297: strcpy(line,stra);
8298: cutv(stra, strb,line,' ');
1.169 brouard 8299: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8300: }
1.169 brouard 8301: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 8302: month=99;
8303: year=9999;
1.136 brouard 8304: }else{
1.225 brouard 8305: 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);
8306: 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);
8307: return 1;
1.136 brouard 8308: }
8309: anint[j][i]= (double) year;
8310: mint[j][i]= (double)month;
8311: strcpy(line,stra);
1.223 brouard 8312: } /* End loop on waves */
1.225 brouard 8313:
1.223 brouard 8314: /* Date of death */
1.136 brouard 8315: cutv(stra, strb,line,' ');
1.169 brouard 8316: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8317: }
1.169 brouard 8318: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 8319: month=99;
8320: year=9999;
8321: }else{
1.141 brouard 8322: 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 8323: 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);
8324: return 1;
1.136 brouard 8325: }
8326: andc[i]=(double) year;
8327: moisdc[i]=(double) month;
8328: strcpy(line,stra);
8329:
1.223 brouard 8330: /* Date of birth */
1.136 brouard 8331: cutv(stra, strb,line,' ');
1.169 brouard 8332: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8333: }
1.169 brouard 8334: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 8335: month=99;
8336: year=9999;
8337: }else{
1.141 brouard 8338: 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);
8339: 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 8340: return 1;
1.136 brouard 8341: }
8342: if (year==9999) {
1.141 brouard 8343: 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);
8344: 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 8345: return 1;
8346:
1.136 brouard 8347: }
8348: annais[i]=(double)(year);
8349: moisnais[i]=(double)(month);
8350: strcpy(line,stra);
1.225 brouard 8351:
1.223 brouard 8352: /* Sample weight */
1.136 brouard 8353: cutv(stra, strb,line,' ');
8354: errno=0;
8355: dval=strtod(strb,&endptr);
8356: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 8357: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
8358: 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 8359: fflush(ficlog);
8360: return 1;
8361: }
8362: weight[i]=dval;
8363: strcpy(line,stra);
1.225 brouard 8364:
1.223 brouard 8365: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
8366: cutv(stra, strb, line, ' ');
8367: if(strb[0]=='.') { /* Missing value */
1.225 brouard 8368: lval=-1;
1.223 brouard 8369: }else{
1.225 brouard 8370: errno=0;
8371: /* what_kind_of_number(strb); */
8372: dval=strtod(strb,&endptr);
8373: /* if(strb != endptr && *endptr == '\0') */
8374: /* dval=dlval; */
8375: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8376: if( strb[0]=='\0' || (*endptr != '\0')){
8377: 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);
8378: 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);
8379: return 1;
8380: }
8381: coqvar[iv][i]=dval;
1.226 brouard 8382: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 8383: }
8384: strcpy(line,stra);
8385: }/* end loop nqv */
1.136 brouard 8386:
1.223 brouard 8387: /* Covariate values */
1.136 brouard 8388: for (j=ncovcol;j>=1;j--){
8389: cutv(stra, strb,line,' ');
1.223 brouard 8390: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 8391: lval=-1;
1.136 brouard 8392: }else{
1.225 brouard 8393: errno=0;
8394: lval=strtol(strb,&endptr,10);
8395: if( strb[0]=='\0' || (*endptr != '\0')){
8396: 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);
8397: 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);
8398: return 1;
8399: }
1.136 brouard 8400: }
8401: if(lval <-1 || lval >1){
1.225 brouard 8402: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8403: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8404: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8405: For example, for multinomial values like 1, 2 and 3,\n \
8406: build V1=0 V2=0 for the reference value (1),\n \
8407: V1=1 V2=0 for (2) \n \
1.136 brouard 8408: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8409: output of IMaCh is often meaningless.\n \
1.136 brouard 8410: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 8411: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8412: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8413: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8414: For example, for multinomial values like 1, 2 and 3,\n \
8415: build V1=0 V2=0 for the reference value (1),\n \
8416: V1=1 V2=0 for (2) \n \
1.136 brouard 8417: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8418: output of IMaCh is often meaningless.\n \
1.136 brouard 8419: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 8420: return 1;
1.136 brouard 8421: }
8422: covar[j][i]=(double)(lval);
8423: strcpy(line,stra);
8424: }
8425: lstra=strlen(stra);
1.225 brouard 8426:
1.136 brouard 8427: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
8428: stratrunc = &(stra[lstra-9]);
8429: num[i]=atol(stratrunc);
8430: }
8431: else
8432: num[i]=atol(stra);
8433: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
8434: 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;}*/
8435:
8436: i=i+1;
8437: } /* End loop reading data */
1.225 brouard 8438:
1.136 brouard 8439: *imax=i-1; /* Number of individuals */
8440: fclose(fic);
1.225 brouard 8441:
1.136 brouard 8442: return (0);
1.164 brouard 8443: /* endread: */
1.225 brouard 8444: printf("Exiting readdata: ");
8445: fclose(fic);
8446: return (1);
1.223 brouard 8447: }
1.126 brouard 8448:
1.234 brouard 8449: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 8450: char *p1 = *stri, *p2 = *stri;
1.235 brouard 8451: while (*p2 == ' ')
1.234 brouard 8452: p2++;
8453: /* while ((*p1++ = *p2++) !=0) */
8454: /* ; */
8455: /* do */
8456: /* while (*p2 == ' ') */
8457: /* p2++; */
8458: /* while (*p1++ == *p2++); */
8459: *stri=p2;
1.145 brouard 8460: }
8461:
1.235 brouard 8462: int decoderesult ( char resultline[], int nres)
1.230 brouard 8463: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
8464: {
1.235 brouard 8465: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 8466: char resultsav[MAXLINE];
1.234 brouard 8467: int resultmodel[MAXLINE];
8468: int modelresult[MAXLINE];
1.230 brouard 8469: char stra[80], strb[80], strc[80], strd[80],stre[80];
8470:
1.234 brouard 8471: removefirstspace(&resultline);
1.233 brouard 8472: printf("decoderesult:%s\n",resultline);
1.230 brouard 8473:
8474: if (strstr(resultline,"v") !=0){
8475: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
8476: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
8477: return 1;
8478: }
8479: trimbb(resultsav, resultline);
8480: if (strlen(resultsav) >1){
8481: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
8482: }
1.253 brouard 8483: if(j == 0){ /* Resultline but no = */
8484: TKresult[nres]=0; /* Combination for the nresult and the model */
8485: return (0);
8486: }
8487:
1.234 brouard 8488: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
8489: 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);
8490: 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);
8491: }
8492: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8493: if(nbocc(resultsav,'=') >1){
8494: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8495: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8496: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8497: }else
8498: cutl(strc,strd,resultsav,'=');
1.230 brouard 8499: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8500:
1.230 brouard 8501: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8502: Tvarsel[k]=atoi(strc);
8503: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8504: /* cptcovsel++; */
8505: if (nbocc(stra,'=') >0)
8506: strcpy(resultsav,stra); /* and analyzes it */
8507: }
1.235 brouard 8508: /* Checking for missing or useless values in comparison of current model needs */
1.236 brouard 8509: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8510: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8511: match=0;
1.236 brouard 8512: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.237 brouard 8513: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */
1.236 brouard 8514: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8515: match=1;
8516: break;
8517: }
8518: }
8519: if(match == 0){
8520: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8521: }
8522: }
8523: }
1.235 brouard 8524: /* Checking for missing or useless values in comparison of current model needs */
8525: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8526: match=0;
1.235 brouard 8527: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8528: if(Typevar[k1]==0){ /* Single */
1.237 brouard 8529: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */
1.235 brouard 8530: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8531: ++match;
8532: }
8533: }
8534: }
8535: if(match == 0){
8536: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8537: }else if(match > 1){
8538: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8539: }
8540: }
1.235 brouard 8541:
1.234 brouard 8542: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8543: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8544: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8545: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8546: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8547: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8548: /* 1 0 0 0 */
8549: /* 2 1 0 0 */
8550: /* 3 0 1 0 */
8551: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8552: /* 5 0 0 1 */
8553: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8554: /* 7 0 1 1 */
8555: /* 8 1 1 1 */
1.237 brouard 8556: /* V(Tvresult)=Tresult V4=1 V3=0 V1=1 Tresult[nres=1][2]=0 */
8557: /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */
8558: /* V5*age V5 known which value for nres? */
8559: /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */
1.235 brouard 8560: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8561: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
1.237 brouard 8562: k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */
1.235 brouard 8563: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8564: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
1.237 brouard 8565: Tresult[nres][k4+1]=Tvalsel[k3];/* Tresult[nres][1]=1(V4=1) Tresult[nres][2]=0(V3=0) */
8566: Tvresult[nres][k4+1]=(int)Tvarsel[k3];/* Tvresult[nres][1]=4 Tvresult[nres][3]=1 */
8567: Tinvresult[nres][(int)Tvarsel[k3]]=Tvalsel[k3]; /* Tinvresult[nres][4]=1 */
1.235 brouard 8568: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8569: k4++;;
8570: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8571: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8572: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
1.237 brouard 8573: Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */
8574: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */
8575: Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */
1.235 brouard 8576: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8577: k4q++;;
8578: }
8579: }
1.234 brouard 8580:
1.235 brouard 8581: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8582: return (0);
8583: }
1.235 brouard 8584:
1.230 brouard 8585: int decodemodel( char model[], int lastobs)
8586: /**< This routine decodes the model and returns:
1.224 brouard 8587: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8588: * - nagesqr = 1 if age*age in the model, otherwise 0.
8589: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8590: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8591: * - cptcovage number of covariates with age*products =2
8592: * - cptcovs number of simple covariates
8593: * - 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
8594: * which is a new column after the 9 (ncovcol) variables.
8595: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8596: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8597: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8598: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8599: */
1.136 brouard 8600: {
1.238 brouard 8601: int i, j, k, ks, v;
1.227 brouard 8602: int j1, k1, k2, k3, k4;
1.136 brouard 8603: char modelsav[80];
1.145 brouard 8604: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8605: char *strpt;
1.136 brouard 8606:
1.145 brouard 8607: /*removespace(model);*/
1.136 brouard 8608: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8609: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8610: if (strstr(model,"AGE") !=0){
1.192 brouard 8611: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8612: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8613: return 1;
8614: }
1.141 brouard 8615: if (strstr(model,"v") !=0){
8616: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8617: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8618: return 1;
8619: }
1.187 brouard 8620: strcpy(modelsav,model);
8621: if ((strpt=strstr(model,"age*age")) !=0){
8622: printf(" strpt=%s, model=%s\n",strpt, model);
8623: if(strpt != model){
1.234 brouard 8624: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8625: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8626: corresponding column of parameters.\n",model);
1.234 brouard 8627: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8628: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8629: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8630: return 1;
1.225 brouard 8631: }
1.187 brouard 8632: nagesqr=1;
8633: if (strstr(model,"+age*age") !=0)
1.234 brouard 8634: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8635: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8636: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8637: else
1.234 brouard 8638: substrchaine(modelsav, model, "age*age");
1.187 brouard 8639: }else
8640: nagesqr=0;
8641: if (strlen(modelsav) >1){
8642: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8643: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8644: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8645: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8646: * cst, age and age*age
8647: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8648: /* including age products which are counted in cptcovage.
8649: * but the covariates which are products must be treated
8650: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8651: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8652: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8653:
8654:
1.187 brouard 8655: /* Design
8656: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8657: * < ncovcol=8 >
8658: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8659: * k= 1 2 3 4 5 6 7 8
8660: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8661: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8662: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8663: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8664: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8665: * Tage[++cptcovage]=k
8666: * if products, new covar are created after ncovcol with k1
8667: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8668: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8669: * 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
8670: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8671: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8672: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8673: * < ncovcol=8 >
8674: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8675: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8676: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8677: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8678: * p Tprod[1]@2={ 6, 5}
8679: *p Tvard[1][1]@4= {7, 8, 5, 6}
8680: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8681: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8682: *How to reorganize?
8683: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8684: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8685: * {2, 1, 4, 8, 5, 6, 3, 7}
8686: * Struct []
8687: */
1.225 brouard 8688:
1.187 brouard 8689: /* This loop fills the array Tvar from the string 'model'.*/
8690: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8691: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8692: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8693: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8694: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8695: /* k=1 Tvar[1]=2 (from V2) */
8696: /* k=5 Tvar[5] */
8697: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8698: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8699: /* } */
1.198 brouard 8700: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8701: /*
8702: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8703: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8704: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8705: }
1.187 brouard 8706: cptcovage=0;
8707: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8708: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8709: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8710: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8711: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8712: /*scanf("%d",i);*/
8713: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8714: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8715: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8716: /* covar is not filled and then is empty */
8717: cptcovprod--;
8718: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8719: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8720: Typevar[k]=1; /* 1 for age product */
8721: cptcovage++; /* Sums the number of covariates which include age as a product */
8722: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8723: /*printf("stre=%s ", stre);*/
8724: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8725: cptcovprod--;
8726: cutl(stre,strb,strc,'V');
8727: Tvar[k]=atoi(stre);
8728: Typevar[k]=1; /* 1 for age product */
8729: cptcovage++;
8730: Tage[cptcovage]=k;
8731: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8732: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8733: cptcovn++;
8734: cptcovprodnoage++;k1++;
8735: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8736: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8737: because this model-covariate is a construction we invent a new column
8738: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8739: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8740: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8741: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8742: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8743: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8744: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8745: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8746: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8747: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8748: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8749: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8750: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8751: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8752: for (i=1; i<=lastobs;i++){
8753: /* Computes the new covariate which is a product of
8754: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8755: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8756: }
8757: } /* End age is not in the model */
8758: } /* End if model includes a product */
8759: else { /* no more sum */
8760: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8761: /* scanf("%d",i);*/
8762: cutl(strd,strc,strb,'V');
8763: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8764: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8765: Tvar[k]=atoi(strd);
8766: Typevar[k]=0; /* 0 for simple covariates */
8767: }
8768: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8769: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8770: scanf("%d",i);*/
1.187 brouard 8771: } /* end of loop + on total covariates */
8772: } /* end if strlen(modelsave == 0) age*age might exist */
8773: } /* end if strlen(model == 0) */
1.136 brouard 8774:
8775: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8776: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8777:
1.136 brouard 8778: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8779: printf("cptcovprod=%d ", cptcovprod);
8780: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8781: scanf("%d ",i);*/
8782:
8783:
1.230 brouard 8784: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8785: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8786: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8787: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8788: k = 1 2 3 4 5 6 7 8 9
8789: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8790: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8791: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8792: Dummy[k] 1 0 0 0 3 1 1 2 3
8793: Tmodelind[combination of covar]=k;
1.225 brouard 8794: */
8795: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8796: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8797: /* 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 8798: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8799: printf("Model=%s\n\
8800: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8801: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8802: 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);
8803: fprintf(ficlog,"Model=%s\n\
8804: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8805: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8806: 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 8807: for(k=1;k<=cptcovt; k++){ Fixed[k]=0; Dummy[k]=0;}
1.234 brouard 8808: 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 */
8809: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 8810: Fixed[k]= 0;
8811: Dummy[k]= 0;
1.225 brouard 8812: ncoveff++;
1.232 brouard 8813: ncovf++;
1.234 brouard 8814: nsd++;
8815: modell[k].maintype= FTYPE;
8816: TvarsD[nsd]=Tvar[k];
8817: TvarsDind[nsd]=k;
8818: TvarF[ncovf]=Tvar[k];
8819: TvarFind[ncovf]=k;
8820: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8821: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8822: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
8823: Fixed[k]= 0;
8824: Dummy[k]= 0;
8825: ncoveff++;
8826: ncovf++;
8827: modell[k].maintype= FTYPE;
8828: TvarF[ncovf]=Tvar[k];
8829: TvarFind[ncovf]=k;
1.230 brouard 8830: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8831: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.240 brouard 8832: }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 8833: Fixed[k]= 0;
8834: Dummy[k]= 1;
1.230 brouard 8835: nqfveff++;
1.234 brouard 8836: modell[k].maintype= FTYPE;
8837: modell[k].subtype= FQ;
8838: nsq++;
8839: TvarsQ[nsq]=Tvar[k];
8840: TvarsQind[nsq]=k;
1.232 brouard 8841: ncovf++;
1.234 brouard 8842: TvarF[ncovf]=Tvar[k];
8843: TvarFind[ncovf]=k;
1.231 brouard 8844: 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 8845: 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 8846: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying dummy variables */
1.227 brouard 8847: Fixed[k]= 1;
8848: Dummy[k]= 0;
1.225 brouard 8849: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 8850: modell[k].maintype= VTYPE;
8851: modell[k].subtype= VD;
8852: nsd++;
8853: TvarsD[nsd]=Tvar[k];
8854: TvarsDind[nsd]=k;
8855: ncovv++; /* Only simple time varying variables */
8856: TvarV[ncovv]=Tvar[k];
1.242 brouard 8857: 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 8858: 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 */
8859: 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 8860: 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);
8861: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8862: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 8863: Fixed[k]= 1;
8864: Dummy[k]= 1;
8865: nqtveff++;
8866: modell[k].maintype= VTYPE;
8867: modell[k].subtype= VQ;
8868: ncovv++; /* Only simple time varying variables */
8869: nsq++;
8870: TvarsQ[nsq]=Tvar[k];
8871: TvarsQind[nsq]=k;
8872: TvarV[ncovv]=Tvar[k];
1.242 brouard 8873: 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 8874: 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 */
8875: 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 8876: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8877: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8878: 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 8879: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8880: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 8881: ncova++;
8882: TvarA[ncova]=Tvar[k];
8883: TvarAind[ncova]=k;
1.231 brouard 8884: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.240 brouard 8885: Fixed[k]= 2;
8886: Dummy[k]= 2;
8887: modell[k].maintype= ATYPE;
8888: modell[k].subtype= APFD;
8889: /* ncoveff++; */
1.227 brouard 8890: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.240 brouard 8891: Fixed[k]= 2;
8892: Dummy[k]= 3;
8893: modell[k].maintype= ATYPE;
8894: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8895: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8896: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.240 brouard 8897: Fixed[k]= 3;
8898: Dummy[k]= 2;
8899: modell[k].maintype= ATYPE;
8900: modell[k].subtype= APVD; /* Product age * varying dummy */
8901: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8902: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8903: Fixed[k]= 3;
8904: Dummy[k]= 3;
8905: modell[k].maintype= ATYPE;
8906: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8907: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8908: }
8909: }else if (Typevar[k] == 2) { /* product without age */
8910: k1=Tposprod[k];
8911: if(Tvard[k1][1] <=ncovcol){
1.240 brouard 8912: if(Tvard[k1][2] <=ncovcol){
8913: Fixed[k]= 1;
8914: Dummy[k]= 0;
8915: modell[k].maintype= FTYPE;
8916: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
8917: ncovf++; /* Fixed variables without age */
8918: TvarF[ncovf]=Tvar[k];
8919: TvarFind[ncovf]=k;
8920: }else if(Tvard[k1][2] <=ncovcol+nqv){
8921: Fixed[k]= 0; /* or 2 ?*/
8922: Dummy[k]= 1;
8923: modell[k].maintype= FTYPE;
8924: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
8925: ncovf++; /* Varying variables without age */
8926: TvarF[ncovf]=Tvar[k];
8927: TvarFind[ncovf]=k;
8928: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8929: Fixed[k]= 1;
8930: Dummy[k]= 0;
8931: modell[k].maintype= VTYPE;
8932: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
8933: ncovv++; /* Varying variables without age */
8934: TvarV[ncovv]=Tvar[k];
8935: TvarVind[ncovv]=k;
8936: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8937: Fixed[k]= 1;
8938: Dummy[k]= 1;
8939: modell[k].maintype= VTYPE;
8940: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
8941: ncovv++; /* Varying variables without age */
8942: TvarV[ncovv]=Tvar[k];
8943: TvarVind[ncovv]=k;
8944: }
1.227 brouard 8945: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.240 brouard 8946: if(Tvard[k1][2] <=ncovcol){
8947: Fixed[k]= 0; /* or 2 ?*/
8948: Dummy[k]= 1;
8949: modell[k].maintype= FTYPE;
8950: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
8951: ncovf++; /* Fixed variables without age */
8952: TvarF[ncovf]=Tvar[k];
8953: TvarFind[ncovf]=k;
8954: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8955: Fixed[k]= 1;
8956: Dummy[k]= 1;
8957: modell[k].maintype= VTYPE;
8958: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
8959: ncovv++; /* Varying variables without age */
8960: TvarV[ncovv]=Tvar[k];
8961: TvarVind[ncovv]=k;
8962: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8963: Fixed[k]= 1;
8964: Dummy[k]= 1;
8965: modell[k].maintype= VTYPE;
8966: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
8967: ncovv++; /* Varying variables without age */
8968: TvarV[ncovv]=Tvar[k];
8969: TvarVind[ncovv]=k;
8970: ncovv++; /* Varying variables without age */
8971: TvarV[ncovv]=Tvar[k];
8972: TvarVind[ncovv]=k;
8973: }
1.227 brouard 8974: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.240 brouard 8975: if(Tvard[k1][2] <=ncovcol){
8976: Fixed[k]= 1;
8977: Dummy[k]= 1;
8978: modell[k].maintype= VTYPE;
8979: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
8980: ncovv++; /* Varying variables without age */
8981: TvarV[ncovv]=Tvar[k];
8982: TvarVind[ncovv]=k;
8983: }else if(Tvard[k1][2] <=ncovcol+nqv){
8984: Fixed[k]= 1;
8985: Dummy[k]= 1;
8986: modell[k].maintype= VTYPE;
8987: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
8988: ncovv++; /* Varying variables without age */
8989: TvarV[ncovv]=Tvar[k];
8990: TvarVind[ncovv]=k;
8991: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8992: Fixed[k]= 1;
8993: Dummy[k]= 0;
8994: modell[k].maintype= VTYPE;
8995: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
8996: ncovv++; /* Varying variables without age */
8997: TvarV[ncovv]=Tvar[k];
8998: TvarVind[ncovv]=k;
8999: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9000: Fixed[k]= 1;
9001: Dummy[k]= 1;
9002: modell[k].maintype= VTYPE;
9003: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
9004: ncovv++; /* Varying variables without age */
9005: TvarV[ncovv]=Tvar[k];
9006: TvarVind[ncovv]=k;
9007: }
1.227 brouard 9008: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 9009: if(Tvard[k1][2] <=ncovcol){
9010: Fixed[k]= 1;
9011: Dummy[k]= 1;
9012: modell[k].maintype= VTYPE;
9013: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
9014: ncovv++; /* Varying variables without age */
9015: TvarV[ncovv]=Tvar[k];
9016: TvarVind[ncovv]=k;
9017: }else if(Tvard[k1][2] <=ncovcol+nqv){
9018: Fixed[k]= 1;
9019: Dummy[k]= 1;
9020: modell[k].maintype= VTYPE;
9021: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
9022: ncovv++; /* Varying variables without age */
9023: TvarV[ncovv]=Tvar[k];
9024: TvarVind[ncovv]=k;
9025: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9026: Fixed[k]= 1;
9027: Dummy[k]= 1;
9028: modell[k].maintype= VTYPE;
9029: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
9030: ncovv++; /* Varying variables without age */
9031: TvarV[ncovv]=Tvar[k];
9032: TvarVind[ncovv]=k;
9033: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9034: Fixed[k]= 1;
9035: Dummy[k]= 1;
9036: modell[k].maintype= VTYPE;
9037: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
9038: ncovv++; /* Varying variables without age */
9039: TvarV[ncovv]=Tvar[k];
9040: TvarVind[ncovv]=k;
9041: }
1.227 brouard 9042: }else{
1.240 brouard 9043: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9044: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9045: } /*end k1*/
1.225 brouard 9046: }else{
1.226 brouard 9047: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
9048: 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 9049: }
1.227 brouard 9050: 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 9051: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 9052: 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]);
9053: }
9054: /* Searching for doublons in the model */
9055: for(k1=1; k1<= cptcovt;k1++){
9056: for(k2=1; k2 <k1;k2++){
9057: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 9058: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
9059: if(Tvar[k1]==Tvar[k2]){
9060: 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]]);
9061: 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);
9062: return(1);
9063: }
9064: }else if (Typevar[k1] ==2){
9065: k3=Tposprod[k1];
9066: k4=Tposprod[k2];
9067: 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])) ){
9068: 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]]);
9069: 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);
9070: return(1);
9071: }
9072: }
1.227 brouard 9073: }
9074: }
1.225 brouard 9075: }
9076: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
9077: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 9078: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
9079: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 9080: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 9081: /*endread:*/
1.225 brouard 9082: printf("Exiting decodemodel: ");
9083: return (1);
1.136 brouard 9084: }
9085:
1.169 brouard 9086: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.248 brouard 9087: {/* Check ages at death */
1.136 brouard 9088: int i, m;
1.218 brouard 9089: int firstone=0;
9090:
1.136 brouard 9091: for (i=1; i<=imx; i++) {
9092: for(m=2; (m<= maxwav); m++) {
9093: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
9094: anint[m][i]=9999;
1.216 brouard 9095: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
9096: s[m][i]=-1;
1.136 brouard 9097: }
9098: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.169 brouard 9099: *nberr = *nberr + 1;
1.218 brouard 9100: if(firstone == 0){
9101: firstone=1;
9102: printf("Error! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown, you must set an arbitrary year of death or he/she is skipped and results can be biased (%d) because status is a death state %d at wave %d. Wave dropped.\nOther similar cases in log file\n",(int)moisdc[i],(int)andc[i],num[i],i, *nberr,s[m][i],m);
9103: }
9104: fprintf(ficlog,"Error! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown, you must set an arbitrary year of death or he/she is skipped and results can be biased (%d) because status is a death state %d at wave %d. Wave dropped.\n",(int)moisdc[i],(int)andc[i],num[i],i, *nberr,s[m][i],m);
1.136 brouard 9105: s[m][i]=-1;
9106: }
9107: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 9108: (*nberr)++;
1.136 brouard 9109: printf("Error! Month of death of individual %ld on line %d was unknown %2d, you should set it otherwise the information on the death is skipped and results are biased.\n",num[i],i,(int)moisdc[i]);
9110: fprintf(ficlog,"Error! Month of death of individual %ld on line %d was unknown %f, you should set it otherwise the information on the death is skipped and results are biased.\n",num[i],i,moisdc[i]);
9111: s[m][i]=-1; /* We prefer to skip it (and to skip it in version 0.8a1 too */
9112: }
9113: }
9114: }
9115:
9116: for (i=1; i<=imx; i++) {
9117: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
9118: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 9119: 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 9120: if (s[m][i] >= nlstate+1) {
1.169 brouard 9121: if(agedc[i]>0){
9122: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 9123: agev[m][i]=agedc[i];
1.214 brouard 9124: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 9125: }else {
1.136 brouard 9126: if ((int)andc[i]!=9999){
9127: nbwarn++;
9128: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
9129: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
9130: agev[m][i]=-1;
9131: }
9132: }
1.169 brouard 9133: } /* agedc > 0 */
1.214 brouard 9134: } /* end if */
1.136 brouard 9135: else if(s[m][i] !=9){ /* Standard case, age in fractional
9136: years but with the precision of a month */
9137: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
9138: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
9139: agev[m][i]=1;
9140: else if(agev[m][i] < *agemin){
9141: *agemin=agev[m][i];
9142: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
9143: }
9144: else if(agev[m][i] >*agemax){
9145: *agemax=agev[m][i];
1.156 brouard 9146: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 9147: }
9148: /*agev[m][i]=anint[m][i]-annais[i];*/
9149: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 9150: } /* en if 9*/
1.136 brouard 9151: else { /* =9 */
1.214 brouard 9152: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 9153: agev[m][i]=1;
9154: s[m][i]=-1;
9155: }
9156: }
1.214 brouard 9157: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 9158: agev[m][i]=1;
1.214 brouard 9159: else{
9160: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9161: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9162: agev[m][i]=0;
9163: }
9164: } /* End for lastpass */
9165: }
1.136 brouard 9166:
9167: for (i=1; i<=imx; i++) {
9168: for(m=firstpass; (m<=lastpass); m++){
9169: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 9170: (*nberr)++;
1.136 brouard 9171: 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);
9172: 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);
9173: return 1;
9174: }
9175: }
9176: }
9177:
9178: /*for (i=1; i<=imx; i++){
9179: for (m=firstpass; (m<lastpass); m++){
9180: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
9181: }
9182:
9183: }*/
9184:
9185:
1.139 brouard 9186: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
9187: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 9188:
9189: return (0);
1.164 brouard 9190: /* endread:*/
1.136 brouard 9191: printf("Exiting calandcheckages: ");
9192: return (1);
9193: }
9194:
1.172 brouard 9195: #if defined(_MSC_VER)
9196: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9197: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9198: //#include "stdafx.h"
9199: //#include <stdio.h>
9200: //#include <tchar.h>
9201: //#include <windows.h>
9202: //#include <iostream>
9203: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
9204:
9205: LPFN_ISWOW64PROCESS fnIsWow64Process;
9206:
9207: BOOL IsWow64()
9208: {
9209: BOOL bIsWow64 = FALSE;
9210:
9211: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
9212: // (HANDLE, PBOOL);
9213:
9214: //LPFN_ISWOW64PROCESS fnIsWow64Process;
9215:
9216: HMODULE module = GetModuleHandle(_T("kernel32"));
9217: const char funcName[] = "IsWow64Process";
9218: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
9219: GetProcAddress(module, funcName);
9220:
9221: if (NULL != fnIsWow64Process)
9222: {
9223: if (!fnIsWow64Process(GetCurrentProcess(),
9224: &bIsWow64))
9225: //throw std::exception("Unknown error");
9226: printf("Unknown error\n");
9227: }
9228: return bIsWow64 != FALSE;
9229: }
9230: #endif
1.177 brouard 9231:
1.191 brouard 9232: void syscompilerinfo(int logged)
1.167 brouard 9233: {
9234: /* #include "syscompilerinfo.h"*/
1.185 brouard 9235: /* command line Intel compiler 32bit windows, XP compatible:*/
9236: /* /GS /W3 /Gy
9237: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
9238: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
9239: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 9240: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
9241: */
9242: /* 64 bits */
1.185 brouard 9243: /*
9244: /GS /W3 /Gy
9245: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
9246: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
9247: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
9248: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
9249: /* Optimization are useless and O3 is slower than O2 */
9250: /*
9251: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
9252: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
9253: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
9254: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
9255: */
1.186 brouard 9256: /* Link is */ /* /OUT:"visual studio
1.185 brouard 9257: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
9258: /PDB:"visual studio
9259: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
9260: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
9261: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
9262: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
9263: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
9264: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
9265: uiAccess='false'"
9266: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
9267: /NOLOGO /TLBID:1
9268: */
1.177 brouard 9269: #if defined __INTEL_COMPILER
1.178 brouard 9270: #if defined(__GNUC__)
9271: struct utsname sysInfo; /* For Intel on Linux and OS/X */
9272: #endif
1.177 brouard 9273: #elif defined(__GNUC__)
1.179 brouard 9274: #ifndef __APPLE__
1.174 brouard 9275: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 9276: #endif
1.177 brouard 9277: struct utsname sysInfo;
1.178 brouard 9278: int cross = CROSS;
9279: if (cross){
9280: printf("Cross-");
1.191 brouard 9281: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 9282: }
1.174 brouard 9283: #endif
9284:
1.171 brouard 9285: #include <stdint.h>
1.178 brouard 9286:
1.191 brouard 9287: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 9288: #if defined(__clang__)
1.191 brouard 9289: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 9290: #endif
9291: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 9292: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 9293: #endif
9294: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 9295: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 9296: #endif
9297: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 9298: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 9299: #endif
9300: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 9301: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 9302: #endif
9303: #if defined(_MSC_VER)
1.191 brouard 9304: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 9305: #endif
9306: #if defined(__PGI)
1.191 brouard 9307: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 9308: #endif
9309: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 9310: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 9311: #endif
1.191 brouard 9312: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 9313:
1.167 brouard 9314: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
9315: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
9316: // Windows (x64 and x86)
1.191 brouard 9317: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 9318: #elif __unix__ // all unices, not all compilers
9319: // Unix
1.191 brouard 9320: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 9321: #elif __linux__
9322: // linux
1.191 brouard 9323: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 9324: #elif __APPLE__
1.174 brouard 9325: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 9326: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 9327: #endif
9328:
9329: /* __MINGW32__ */
9330: /* __CYGWIN__ */
9331: /* __MINGW64__ */
9332: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
9333: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
9334: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
9335: /* _WIN64 // Defined for applications for Win64. */
9336: /* _M_X64 // Defined for compilations that target x64 processors. */
9337: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 9338:
1.167 brouard 9339: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 9340: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 9341: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 9342: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 9343: #else
1.191 brouard 9344: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 9345: #endif
9346:
1.169 brouard 9347: #if defined(__GNUC__)
9348: # if defined(__GNUC_PATCHLEVEL__)
9349: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9350: + __GNUC_MINOR__ * 100 \
9351: + __GNUC_PATCHLEVEL__)
9352: # else
9353: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9354: + __GNUC_MINOR__ * 100)
9355: # endif
1.174 brouard 9356: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 9357: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 9358:
9359: if (uname(&sysInfo) != -1) {
9360: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 9361: 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 9362: }
9363: else
9364: perror("uname() error");
1.179 brouard 9365: //#ifndef __INTEL_COMPILER
9366: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 9367: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 9368: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 9369: #endif
1.169 brouard 9370: #endif
1.172 brouard 9371:
9372: // void main()
9373: // {
1.169 brouard 9374: #if defined(_MSC_VER)
1.174 brouard 9375: if (IsWow64()){
1.191 brouard 9376: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
9377: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 9378: }
9379: else{
1.191 brouard 9380: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
9381: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 9382: }
1.172 brouard 9383: // printf("\nPress Enter to continue...");
9384: // getchar();
9385: // }
9386:
1.169 brouard 9387: #endif
9388:
1.167 brouard 9389:
1.219 brouard 9390: }
1.136 brouard 9391:
1.219 brouard 9392: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 9393: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 9394: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 9395: /* double ftolpl = 1.e-10; */
1.180 brouard 9396: double age, agebase, agelim;
1.203 brouard 9397: double tot;
1.180 brouard 9398:
1.202 brouard 9399: strcpy(filerespl,"PL_");
9400: strcat(filerespl,fileresu);
9401: if((ficrespl=fopen(filerespl,"w"))==NULL) {
9402: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9403: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9404: }
1.227 brouard 9405: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
9406: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 9407: pstamp(ficrespl);
1.203 brouard 9408: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 9409: fprintf(ficrespl,"#Age ");
9410: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
9411: fprintf(ficrespl,"\n");
1.180 brouard 9412:
1.219 brouard 9413: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 9414:
1.219 brouard 9415: agebase=ageminpar;
9416: agelim=agemaxpar;
1.180 brouard 9417:
1.227 brouard 9418: /* i1=pow(2,ncoveff); */
1.234 brouard 9419: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 9420: if (cptcovn < 1){i1=1;}
1.180 brouard 9421:
1.238 brouard 9422: for(k=1; k<=i1;k++){ /* For each combination k of dummy covariates in the model */
9423: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 9424: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9425: continue;
1.235 brouard 9426:
1.238 brouard 9427: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9428: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
9429: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
9430: /* k=k+1; */
9431: /* to clean */
9432: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9433: fprintf(ficrespl,"#******");
9434: printf("#******");
9435: fprintf(ficlog,"#******");
9436: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9437: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
9438: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9439: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9440: }
9441: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9442: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9443: fprintf(ficrespl," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9444: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9445: }
9446: fprintf(ficrespl,"******\n");
9447: printf("******\n");
9448: fprintf(ficlog,"******\n");
9449: if(invalidvarcomb[k]){
9450: printf("\nCombination (%d) ignored because no case \n",k);
9451: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
9452: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
9453: continue;
9454: }
1.219 brouard 9455:
1.238 brouard 9456: fprintf(ficrespl,"#Age ");
9457: for(j=1;j<=cptcoveff;j++) {
9458: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9459: }
9460: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
9461: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 9462:
1.238 brouard 9463: for (age=agebase; age<=agelim; age++){
9464: /* for (age=agebase; age<=agebase; age++){ */
9465: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
9466: fprintf(ficrespl,"%.0f ",age );
9467: for(j=1;j<=cptcoveff;j++)
9468: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9469: tot=0.;
9470: for(i=1; i<=nlstate;i++){
9471: tot += prlim[i][i];
9472: fprintf(ficrespl," %.5f", prlim[i][i]);
9473: }
9474: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
9475: } /* Age */
9476: /* was end of cptcod */
9477: } /* cptcov */
9478: } /* nres */
1.219 brouard 9479: return 0;
1.180 brouard 9480: }
9481:
1.218 brouard 9482: 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){
9483: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
9484:
9485: /* Computes the back prevalence limit for any combination of covariate values
9486: * at any age between ageminpar and agemaxpar
9487: */
1.235 brouard 9488: int i, j, k, i1, nres=0 ;
1.217 brouard 9489: /* double ftolpl = 1.e-10; */
9490: double age, agebase, agelim;
9491: double tot;
1.218 brouard 9492: /* double ***mobaverage; */
9493: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 9494:
9495: strcpy(fileresplb,"PLB_");
9496: strcat(fileresplb,fileresu);
9497: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
9498: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9499: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9500: }
9501: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9502: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9503: pstamp(ficresplb);
9504: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9505: fprintf(ficresplb,"#Age ");
9506: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9507: fprintf(ficresplb,"\n");
9508:
1.218 brouard 9509:
9510: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9511:
9512: agebase=ageminpar;
9513: agelim=agemaxpar;
9514:
9515:
1.227 brouard 9516: i1=pow(2,cptcoveff);
1.218 brouard 9517: if (cptcovn < 1){i1=1;}
1.227 brouard 9518:
1.238 brouard 9519: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9520: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9521: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9522: continue;
9523: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9524: fprintf(ficresplb,"#******");
9525: printf("#******");
9526: fprintf(ficlog,"#******");
9527: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9528: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9529: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9530: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9531: }
9532: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9533: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9534: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9535: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9536: }
9537: fprintf(ficresplb,"******\n");
9538: printf("******\n");
9539: fprintf(ficlog,"******\n");
9540: if(invalidvarcomb[k]){
9541: printf("\nCombination (%d) ignored because no cases \n",k);
9542: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9543: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9544: continue;
9545: }
1.218 brouard 9546:
1.238 brouard 9547: fprintf(ficresplb,"#Age ");
9548: for(j=1;j<=cptcoveff;j++) {
9549: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9550: }
9551: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9552: fprintf(ficresplb,"Total Years_to_converge\n");
1.218 brouard 9553:
9554:
1.238 brouard 9555: for (age=agebase; age<=agelim; age++){
9556: /* for (age=agebase; age<=agebase; age++){ */
9557: if(mobilavproj > 0){
9558: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9559: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9560: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k, nres);
1.238 brouard 9561: }else if (mobilavproj == 0){
9562: 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);
9563: 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);
9564: exit(1);
9565: }else{
9566: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9567: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k,nres);
1.238 brouard 9568: }
9569: fprintf(ficresplb,"%.0f ",age );
9570: for(j=1;j<=cptcoveff;j++)
9571: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9572: tot=0.;
9573: for(i=1; i<=nlstate;i++){
9574: tot += bprlim[i][i];
9575: fprintf(ficresplb," %.5f", bprlim[i][i]);
9576: }
9577: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9578: } /* Age */
9579: /* was end of cptcod */
1.255 brouard 9580: /*fprintf(ficresplb,"\n");*/ /* Seems to be necessary for gnuplot only if two result lines and no covariate. */
1.238 brouard 9581: } /* end of any combination */
9582: } /* end of nres */
1.218 brouard 9583: /* hBijx(p, bage, fage); */
9584: /* fclose(ficrespijb); */
9585:
9586: return 0;
1.217 brouard 9587: }
1.218 brouard 9588:
1.180 brouard 9589: int hPijx(double *p, int bage, int fage){
9590: /*------------- h Pij x at various ages ------------*/
9591:
9592: int stepsize;
9593: int agelim;
9594: int hstepm;
9595: int nhstepm;
1.235 brouard 9596: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9597:
9598: double agedeb;
9599: double ***p3mat;
9600:
1.201 brouard 9601: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9602: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9603: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9604: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9605: }
9606: printf("Computing pij: result on file '%s' \n", filerespij);
9607: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9608:
9609: stepsize=(int) (stepm+YEARM-1)/YEARM;
9610: /*if (stepm<=24) stepsize=2;*/
9611:
9612: agelim=AGESUP;
9613: hstepm=stepsize*YEARM; /* Every year of age */
9614: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9615:
1.180 brouard 9616: /* hstepm=1; aff par mois*/
9617: pstamp(ficrespij);
9618: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9619: i1= pow(2,cptcoveff);
1.218 brouard 9620: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9621: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9622: /* k=k+1; */
1.235 brouard 9623: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9624: for(k=1; k<=i1;k++){
1.253 brouard 9625: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 9626: continue;
1.183 brouard 9627: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9628: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9629: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9630: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9631: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9632: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9633: }
1.183 brouard 9634: fprintf(ficrespij,"******\n");
9635:
9636: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9637: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9638: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9639:
9640: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9641:
1.183 brouard 9642: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9643: oldm=oldms;savm=savms;
1.235 brouard 9644: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9645: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9646: for(i=1; i<=nlstate;i++)
9647: for(j=1; j<=nlstate+ndeath;j++)
9648: fprintf(ficrespij," %1d-%1d",i,j);
9649: fprintf(ficrespij,"\n");
9650: for (h=0; h<=nhstepm; h++){
9651: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9652: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9653: for(i=1; i<=nlstate;i++)
9654: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9655: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9656: fprintf(ficrespij,"\n");
9657: }
1.183 brouard 9658: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9659: fprintf(ficrespij,"\n");
9660: }
1.180 brouard 9661: /*}*/
9662: }
1.218 brouard 9663: return 0;
1.180 brouard 9664: }
1.218 brouard 9665:
9666: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9667: /*------------- h Bij x at various ages ------------*/
9668:
9669: int stepsize;
1.218 brouard 9670: /* int agelim; */
9671: int ageminl;
1.217 brouard 9672: int hstepm;
9673: int nhstepm;
1.238 brouard 9674: int h, i, i1, j, k, nres;
1.218 brouard 9675:
1.217 brouard 9676: double agedeb;
9677: double ***p3mat;
1.218 brouard 9678:
9679: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9680: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9681: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9682: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9683: }
9684: printf("Computing pij back: result on file '%s' \n", filerespijb);
9685: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9686:
9687: stepsize=(int) (stepm+YEARM-1)/YEARM;
9688: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9689:
1.218 brouard 9690: /* agelim=AGESUP; */
9691: ageminl=30;
9692: hstepm=stepsize*YEARM; /* Every year of age */
9693: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9694:
9695: /* hstepm=1; aff par mois*/
9696: pstamp(ficrespijb);
1.255 brouard 9697: 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 9698: i1= pow(2,cptcoveff);
1.218 brouard 9699: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9700: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9701: /* k=k+1; */
1.238 brouard 9702: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9703: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9704: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9705: continue;
9706: fprintf(ficrespijb,"\n#****** ");
9707: for(j=1;j<=cptcoveff;j++)
9708: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9709: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9710: fprintf(ficrespijb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9711: }
9712: fprintf(ficrespijb,"******\n");
9713: if(invalidvarcomb[k]){
9714: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9715: continue;
9716: }
9717:
9718: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9719: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9720: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9721: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9722: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9723:
9724: /* nhstepm=nhstepm*YEARM; aff par mois*/
9725:
9726: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9727: /* oldm=oldms;savm=savms; */
9728: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9729: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9730: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
1.255 brouard 9731: fprintf(ficrespijb,"# Cov Agex agex-h hbijx with i,j=");
1.217 brouard 9732: for(i=1; i<=nlstate;i++)
9733: for(j=1; j<=nlstate+ndeath;j++)
1.238 brouard 9734: fprintf(ficrespijb," %1d-%1d",i,j);
1.217 brouard 9735: fprintf(ficrespijb,"\n");
1.238 brouard 9736: for (h=0; h<=nhstepm; h++){
9737: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9738: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9739: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
9740: for(i=1; i<=nlstate;i++)
9741: for(j=1; j<=nlstate+ndeath;j++)
9742: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
9743: fprintf(ficrespijb,"\n");
9744: }
9745: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9746: fprintf(ficrespijb,"\n");
9747: } /* end age deb */
9748: } /* end combination */
9749: } /* end nres */
1.218 brouard 9750: return 0;
9751: } /* hBijx */
1.217 brouard 9752:
1.180 brouard 9753:
1.136 brouard 9754: /***********************************************/
9755: /**************** Main Program *****************/
9756: /***********************************************/
9757:
9758: int main(int argc, char *argv[])
9759: {
9760: #ifdef GSL
9761: const gsl_multimin_fminimizer_type *T;
9762: size_t iteri = 0, it;
9763: int rval = GSL_CONTINUE;
9764: int status = GSL_SUCCESS;
9765: double ssval;
9766: #endif
9767: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9768: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9769: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9770: int jj, ll, li, lj, lk;
1.136 brouard 9771: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9772: int num_filled;
1.136 brouard 9773: int itimes;
9774: int NDIM=2;
9775: int vpopbased=0;
1.235 brouard 9776: int nres=0;
1.136 brouard 9777:
1.164 brouard 9778: char ca[32], cb[32];
1.136 brouard 9779: /* FILE *fichtm; *//* Html File */
9780: /* FILE *ficgp;*/ /*Gnuplot File */
9781: struct stat info;
1.191 brouard 9782: double agedeb=0.;
1.194 brouard 9783:
9784: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9785: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9786:
1.165 brouard 9787: double fret;
1.191 brouard 9788: double dum=0.; /* Dummy variable */
1.136 brouard 9789: double ***p3mat;
1.218 brouard 9790: /* double ***mobaverage; */
1.164 brouard 9791:
9792: char line[MAXLINE];
1.197 brouard 9793: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9794:
1.234 brouard 9795: char modeltemp[MAXLINE];
1.230 brouard 9796: char resultline[MAXLINE];
9797:
1.136 brouard 9798: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9799: char *tok, *val; /* pathtot */
1.136 brouard 9800: int firstobs=1, lastobs=10;
1.195 brouard 9801: int c, h , cpt, c2;
1.191 brouard 9802: int jl=0;
9803: int i1, j1, jk, stepsize=0;
1.194 brouard 9804: int count=0;
9805:
1.164 brouard 9806: int *tab;
1.136 brouard 9807: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9808: int backcast=0;
1.136 brouard 9809: int mobilav=0,popforecast=0;
1.191 brouard 9810: int hstepm=0, nhstepm=0;
1.136 brouard 9811: int agemortsup;
9812: float sumlpop=0.;
9813: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9814: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9815:
1.191 brouard 9816: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9817: double ftolpl=FTOL;
9818: double **prlim;
1.217 brouard 9819: double **bprlim;
1.136 brouard 9820: double ***param; /* Matrix of parameters */
1.251 brouard 9821: double ***paramstart; /* Matrix of starting parameter values */
9822: double *p, *pstart; /* p=param[1][1] pstart is for starting values guessed by freqsummary */
1.136 brouard 9823: double **matcov; /* Matrix of covariance */
1.203 brouard 9824: double **hess; /* Hessian matrix */
1.136 brouard 9825: double ***delti3; /* Scale */
9826: double *delti; /* Scale */
9827: double ***eij, ***vareij;
9828: double **varpl; /* Variances of prevalence limits by age */
9829: double *epj, vepp;
1.164 brouard 9830:
1.136 brouard 9831: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9832: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9833:
1.136 brouard 9834: double **ximort;
1.145 brouard 9835: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9836: int *dcwave;
9837:
1.164 brouard 9838: char z[1]="c";
1.136 brouard 9839:
9840: /*char *strt;*/
9841: char strtend[80];
1.126 brouard 9842:
1.164 brouard 9843:
1.126 brouard 9844: /* setlocale (LC_ALL, ""); */
9845: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9846: /* textdomain (PACKAGE); */
9847: /* setlocale (LC_CTYPE, ""); */
9848: /* setlocale (LC_MESSAGES, ""); */
9849:
9850: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9851: rstart_time = time(NULL);
9852: /* (void) gettimeofday(&start_time,&tzp);*/
9853: start_time = *localtime(&rstart_time);
1.126 brouard 9854: curr_time=start_time;
1.157 brouard 9855: /*tml = *localtime(&start_time.tm_sec);*/
9856: /* strcpy(strstart,asctime(&tml)); */
9857: strcpy(strstart,asctime(&start_time));
1.126 brouard 9858:
9859: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9860: /* tp.tm_sec = tp.tm_sec +86400; */
9861: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9862: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9863: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9864: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9865: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9866: /* strt=asctime(&tmg); */
9867: /* printf("Time(after) =%s",strstart); */
9868: /* (void) time (&time_value);
9869: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9870: * tm = *localtime(&time_value);
9871: * strstart=asctime(&tm);
9872: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9873: */
9874:
9875: nberr=0; /* Number of errors and warnings */
9876: nbwarn=0;
1.184 brouard 9877: #ifdef WIN32
9878: _getcwd(pathcd, size);
9879: #else
1.126 brouard 9880: getcwd(pathcd, size);
1.184 brouard 9881: #endif
1.191 brouard 9882: syscompilerinfo(0);
1.196 brouard 9883: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9884: if(argc <=1){
9885: printf("\nEnter the parameter file name: ");
1.205 brouard 9886: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9887: printf("ERROR Empty parameter file name\n");
9888: goto end;
9889: }
1.126 brouard 9890: i=strlen(pathr);
9891: if(pathr[i-1]=='\n')
9892: pathr[i-1]='\0';
1.156 brouard 9893: i=strlen(pathr);
1.205 brouard 9894: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9895: pathr[i-1]='\0';
1.205 brouard 9896: }
9897: i=strlen(pathr);
9898: if( i==0 ){
9899: printf("ERROR Empty parameter file name\n");
9900: goto end;
9901: }
9902: for (tok = pathr; tok != NULL; ){
1.126 brouard 9903: printf("Pathr |%s|\n",pathr);
9904: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9905: printf("val= |%s| pathr=%s\n",val,pathr);
9906: strcpy (pathtot, val);
9907: if(pathr[0] == '\0') break; /* Dirty */
9908: }
9909: }
9910: else{
9911: strcpy(pathtot,argv[1]);
9912: }
9913: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9914: /*cygwin_split_path(pathtot,path,optionfile);
9915: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9916: /* cutv(path,optionfile,pathtot,'\\');*/
9917:
9918: /* Split argv[0], imach program to get pathimach */
9919: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9920: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9921: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9922: /* strcpy(pathimach,argv[0]); */
9923: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9924: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9925: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9926: #ifdef WIN32
9927: _chdir(path); /* Can be a relative path */
9928: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9929: #else
1.126 brouard 9930: chdir(path); /* Can be a relative path */
1.184 brouard 9931: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9932: #endif
9933: printf("Current directory %s!\n",pathcd);
1.126 brouard 9934: strcpy(command,"mkdir ");
9935: strcat(command,optionfilefiname);
9936: if((outcmd=system(command)) != 0){
1.169 brouard 9937: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9938: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9939: /* fclose(ficlog); */
9940: /* exit(1); */
9941: }
9942: /* if((imk=mkdir(optionfilefiname))<0){ */
9943: /* perror("mkdir"); */
9944: /* } */
9945:
9946: /*-------- arguments in the command line --------*/
9947:
1.186 brouard 9948: /* Main Log file */
1.126 brouard 9949: strcat(filelog, optionfilefiname);
9950: strcat(filelog,".log"); /* */
9951: if((ficlog=fopen(filelog,"w"))==NULL) {
9952: printf("Problem with logfile %s\n",filelog);
9953: goto end;
9954: }
9955: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9956: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9957: fprintf(ficlog,"\nEnter the parameter file name: \n");
9958: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9959: path=%s \n\
9960: optionfile=%s\n\
9961: optionfilext=%s\n\
1.156 brouard 9962: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9963:
1.197 brouard 9964: syscompilerinfo(1);
1.167 brouard 9965:
1.126 brouard 9966: printf("Local time (at start):%s",strstart);
9967: fprintf(ficlog,"Local time (at start): %s",strstart);
9968: fflush(ficlog);
9969: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9970: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9971:
9972: /* */
9973: strcpy(fileres,"r");
9974: strcat(fileres, optionfilefiname);
1.201 brouard 9975: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9976: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9977: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 9978:
1.186 brouard 9979: /* Main ---------arguments file --------*/
1.126 brouard 9980:
9981: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 9982: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
9983: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 9984: fflush(ficlog);
1.149 brouard 9985: /* goto end; */
9986: exit(70);
1.126 brouard 9987: }
9988:
9989:
9990:
9991: strcpy(filereso,"o");
1.201 brouard 9992: strcat(filereso,fileresu);
1.126 brouard 9993: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
9994: printf("Problem with Output resultfile: %s\n", filereso);
9995: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
9996: fflush(ficlog);
9997: goto end;
9998: }
9999:
10000: /* Reads comments: lines beginning with '#' */
10001: numlinepar=0;
1.197 brouard 10002:
10003: /* First parameter line */
10004: while(fgets(line, MAXLINE, ficpar)) {
10005: /* If line starts with a # it is a comment */
10006: if (line[0] == '#') {
10007: numlinepar++;
10008: fputs(line,stdout);
10009: fputs(line,ficparo);
10010: fputs(line,ficlog);
10011: continue;
10012: }else
10013: break;
10014: }
10015: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
10016: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
10017: if (num_filled != 5) {
10018: printf("Should be 5 parameters\n");
10019: }
1.126 brouard 10020: numlinepar++;
1.197 brouard 10021: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
10022: }
10023: /* Second parameter line */
10024: while(fgets(line, MAXLINE, ficpar)) {
10025: /* If line starts with a # it is a comment */
10026: if (line[0] == '#') {
10027: numlinepar++;
10028: fputs(line,stdout);
10029: fputs(line,ficparo);
10030: fputs(line,ficlog);
10031: continue;
10032: }else
10033: break;
10034: }
1.223 brouard 10035: 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", \
10036: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
10037: if (num_filled != 11) {
10038: 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 10039: printf("but line=%s\n",line);
1.197 brouard 10040: }
1.223 brouard 10041: 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 10042: }
1.203 brouard 10043: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 10044: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 10045: /* Third 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: }
1.201 brouard 10057: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
10058: if (num_filled == 0)
10059: model[0]='\0';
10060: else if (num_filled != 1){
1.197 brouard 10061: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10062: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10063: model[0]='\0';
10064: goto end;
10065: }
10066: else{
10067: if (model[0]=='+'){
10068: for(i=1; i<=strlen(model);i++)
10069: modeltemp[i-1]=model[i];
1.201 brouard 10070: strcpy(model,modeltemp);
1.197 brouard 10071: }
10072: }
1.199 brouard 10073: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 10074: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 10075: }
10076: /* 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); */
10077: /* numlinepar=numlinepar+3; /\* In general *\/ */
10078: /* 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 10079: 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);
10080: 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 10081: fflush(ficlog);
1.190 brouard 10082: /* if(model[0]=='#'|| model[0]== '\0'){ */
10083: if(model[0]=='#'){
1.187 brouard 10084: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
10085: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
10086: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
10087: if(mle != -1){
10088: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
10089: exit(1);
10090: }
10091: }
1.126 brouard 10092: while((c=getc(ficpar))=='#' && c!= EOF){
10093: ungetc(c,ficpar);
10094: fgets(line, MAXLINE, ficpar);
10095: numlinepar++;
1.195 brouard 10096: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
10097: z[0]=line[1];
10098: }
10099: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 10100: fputs(line, stdout);
10101: //puts(line);
1.126 brouard 10102: fputs(line,ficparo);
10103: fputs(line,ficlog);
10104: }
10105: ungetc(c,ficpar);
10106:
10107:
1.145 brouard 10108: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 10109: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 10110: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 10111: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 10112: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
10113: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
10114: v1+v2*age+v2*v3 makes cptcovn = 3
10115: */
10116: if (strlen(model)>1)
1.187 brouard 10117: 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 10118: else
1.187 brouard 10119: ncovmodel=2; /* Constant and age */
1.133 brouard 10120: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
10121: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 10122: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
10123: 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);
10124: 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);
10125: fflush(stdout);
10126: fclose (ficlog);
10127: goto end;
10128: }
1.126 brouard 10129: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10130: delti=delti3[1][1];
10131: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
10132: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
1.247 brouard 10133: /* We could also provide initial parameters values giving by simple logistic regression
10134: * only one way, that is without matrix product. We will have nlstate maximizations */
10135: /* for(i=1;i<nlstate;i++){ */
10136: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10137: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10138: /* } */
1.126 brouard 10139: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 10140: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
10141: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10142: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
10143: fclose (ficparo);
10144: fclose (ficlog);
10145: goto end;
10146: exit(0);
1.220 brouard 10147: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 10148: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 10149: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
10150: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10151: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10152: matcov=matrix(1,npar,1,npar);
1.203 brouard 10153: hess=matrix(1,npar,1,npar);
1.220 brouard 10154: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 10155: /* Read guessed parameters */
1.126 brouard 10156: /* Reads comments: lines beginning with '#' */
10157: while((c=getc(ficpar))=='#' && c!= EOF){
10158: ungetc(c,ficpar);
10159: fgets(line, MAXLINE, ficpar);
10160: numlinepar++;
1.141 brouard 10161: fputs(line,stdout);
1.126 brouard 10162: fputs(line,ficparo);
10163: fputs(line,ficlog);
10164: }
10165: ungetc(c,ficpar);
10166:
10167: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.251 brouard 10168: paramstart= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.126 brouard 10169: for(i=1; i <=nlstate; i++){
1.234 brouard 10170: j=0;
1.126 brouard 10171: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 10172: if(jj==i) continue;
10173: j++;
10174: fscanf(ficpar,"%1d%1d",&i1,&j1);
10175: if ((i1 != i) || (j1 != jj)){
10176: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 10177: It might be a problem of design; if ncovcol and the model are correct\n \
10178: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 10179: exit(1);
10180: }
10181: fprintf(ficparo,"%1d%1d",i1,j1);
10182: if(mle==1)
10183: printf("%1d%1d",i,jj);
10184: fprintf(ficlog,"%1d%1d",i,jj);
10185: for(k=1; k<=ncovmodel;k++){
10186: fscanf(ficpar," %lf",¶m[i][j][k]);
10187: if(mle==1){
10188: printf(" %lf",param[i][j][k]);
10189: fprintf(ficlog," %lf",param[i][j][k]);
10190: }
10191: else
10192: fprintf(ficlog," %lf",param[i][j][k]);
10193: fprintf(ficparo," %lf",param[i][j][k]);
10194: }
10195: fscanf(ficpar,"\n");
10196: numlinepar++;
10197: if(mle==1)
10198: printf("\n");
10199: fprintf(ficlog,"\n");
10200: fprintf(ficparo,"\n");
1.126 brouard 10201: }
10202: }
10203: fflush(ficlog);
1.234 brouard 10204:
1.251 brouard 10205: /* Reads parameters values */
1.126 brouard 10206: p=param[1][1];
1.251 brouard 10207: pstart=paramstart[1][1];
1.126 brouard 10208:
10209: /* Reads comments: lines beginning with '#' */
10210: while((c=getc(ficpar))=='#' && c!= EOF){
10211: ungetc(c,ficpar);
10212: fgets(line, MAXLINE, ficpar);
10213: numlinepar++;
1.141 brouard 10214: fputs(line,stdout);
1.126 brouard 10215: fputs(line,ficparo);
10216: fputs(line,ficlog);
10217: }
10218: ungetc(c,ficpar);
10219:
10220: for(i=1; i <=nlstate; i++){
10221: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 10222: fscanf(ficpar,"%1d%1d",&i1,&j1);
10223: if ( (i1-i) * (j1-j) != 0){
10224: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
10225: exit(1);
10226: }
10227: printf("%1d%1d",i,j);
10228: fprintf(ficparo,"%1d%1d",i1,j1);
10229: fprintf(ficlog,"%1d%1d",i1,j1);
10230: for(k=1; k<=ncovmodel;k++){
10231: fscanf(ficpar,"%le",&delti3[i][j][k]);
10232: printf(" %le",delti3[i][j][k]);
10233: fprintf(ficparo," %le",delti3[i][j][k]);
10234: fprintf(ficlog," %le",delti3[i][j][k]);
10235: }
10236: fscanf(ficpar,"\n");
10237: numlinepar++;
10238: printf("\n");
10239: fprintf(ficparo,"\n");
10240: fprintf(ficlog,"\n");
1.126 brouard 10241: }
10242: }
10243: fflush(ficlog);
1.234 brouard 10244:
1.145 brouard 10245: /* Reads covariance matrix */
1.126 brouard 10246: delti=delti3[1][1];
1.220 brouard 10247:
10248:
1.126 brouard 10249: /* 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 10250:
1.126 brouard 10251: /* Reads comments: lines beginning with '#' */
10252: while((c=getc(ficpar))=='#' && c!= EOF){
10253: ungetc(c,ficpar);
10254: fgets(line, MAXLINE, ficpar);
10255: numlinepar++;
1.141 brouard 10256: fputs(line,stdout);
1.126 brouard 10257: fputs(line,ficparo);
10258: fputs(line,ficlog);
10259: }
10260: ungetc(c,ficpar);
1.220 brouard 10261:
1.126 brouard 10262: matcov=matrix(1,npar,1,npar);
1.203 brouard 10263: hess=matrix(1,npar,1,npar);
1.131 brouard 10264: for(i=1; i <=npar; i++)
10265: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 10266:
1.194 brouard 10267: /* Scans npar lines */
1.126 brouard 10268: for(i=1; i <=npar; i++){
1.226 brouard 10269: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 10270: if(count != 3){
1.226 brouard 10271: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10272: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10273: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10274: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10275: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10276: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10277: exit(1);
1.220 brouard 10278: }else{
1.226 brouard 10279: if(mle==1)
10280: printf("%1d%1d%d",i1,j1,jk);
10281: }
10282: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
10283: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 10284: for(j=1; j <=i; j++){
1.226 brouard 10285: fscanf(ficpar," %le",&matcov[i][j]);
10286: if(mle==1){
10287: printf(" %.5le",matcov[i][j]);
10288: }
10289: fprintf(ficlog," %.5le",matcov[i][j]);
10290: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 10291: }
10292: fscanf(ficpar,"\n");
10293: numlinepar++;
10294: if(mle==1)
1.220 brouard 10295: printf("\n");
1.126 brouard 10296: fprintf(ficlog,"\n");
10297: fprintf(ficparo,"\n");
10298: }
1.194 brouard 10299: /* End of read covariance matrix npar lines */
1.126 brouard 10300: for(i=1; i <=npar; i++)
10301: for(j=i+1;j<=npar;j++)
1.226 brouard 10302: matcov[i][j]=matcov[j][i];
1.126 brouard 10303:
10304: if(mle==1)
10305: printf("\n");
10306: fprintf(ficlog,"\n");
10307:
10308: fflush(ficlog);
10309:
10310: /*-------- Rewriting parameter file ----------*/
10311: strcpy(rfileres,"r"); /* "Rparameterfile */
10312: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
10313: strcat(rfileres,"."); /* */
10314: strcat(rfileres,optionfilext); /* Other files have txt extension */
10315: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 10316: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
10317: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 10318: }
10319: fprintf(ficres,"#%s\n",version);
10320: } /* End of mle != -3 */
1.218 brouard 10321:
1.186 brouard 10322: /* Main data
10323: */
1.126 brouard 10324: n= lastobs;
10325: num=lvector(1,n);
10326: moisnais=vector(1,n);
10327: annais=vector(1,n);
10328: moisdc=vector(1,n);
10329: andc=vector(1,n);
1.220 brouard 10330: weight=vector(1,n);
1.126 brouard 10331: agedc=vector(1,n);
10332: cod=ivector(1,n);
1.220 brouard 10333: for(i=1;i<=n;i++){
1.234 brouard 10334: num[i]=0;
10335: moisnais[i]=0;
10336: annais[i]=0;
10337: moisdc[i]=0;
10338: andc[i]=0;
10339: agedc[i]=0;
10340: cod[i]=0;
10341: weight[i]=1.0; /* Equal weights, 1 by default */
10342: }
1.126 brouard 10343: mint=matrix(1,maxwav,1,n);
10344: anint=matrix(1,maxwav,1,n);
1.131 brouard 10345: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 10346: tab=ivector(1,NCOVMAX);
1.144 brouard 10347: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 10348: 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 10349:
1.136 brouard 10350: /* Reads data from file datafile */
10351: if (readdata(datafile, firstobs, lastobs, &imx)==1)
10352: goto end;
10353:
10354: /* Calculation of the number of parameters from char model */
1.234 brouard 10355: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 10356: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
10357: k=3 V4 Tvar[k=3]= 4 (from V4)
10358: k=2 V1 Tvar[k=2]= 1 (from V1)
10359: k=1 Tvar[1]=2 (from V2)
1.234 brouard 10360: */
10361:
10362: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
10363: TvarsDind=ivector(1,NCOVMAX); /* */
10364: TvarsD=ivector(1,NCOVMAX); /* */
10365: TvarsQind=ivector(1,NCOVMAX); /* */
10366: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 10367: TvarF=ivector(1,NCOVMAX); /* */
10368: TvarFind=ivector(1,NCOVMAX); /* */
10369: TvarV=ivector(1,NCOVMAX); /* */
10370: TvarVind=ivector(1,NCOVMAX); /* */
10371: TvarA=ivector(1,NCOVMAX); /* */
10372: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 10373: TvarFD=ivector(1,NCOVMAX); /* */
10374: TvarFDind=ivector(1,NCOVMAX); /* */
10375: TvarFQ=ivector(1,NCOVMAX); /* */
10376: TvarFQind=ivector(1,NCOVMAX); /* */
10377: TvarVD=ivector(1,NCOVMAX); /* */
10378: TvarVDind=ivector(1,NCOVMAX); /* */
10379: TvarVQ=ivector(1,NCOVMAX); /* */
10380: TvarVQind=ivector(1,NCOVMAX); /* */
10381:
1.230 brouard 10382: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 10383: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 10384: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
10385: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
10386: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 10387: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
10388: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
10389: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
10390: */
10391: /* For model-covariate k tells which data-covariate to use but
10392: because this model-covariate is a construction we invent a new column
10393: ncovcol + k1
10394: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
10395: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 10396: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
10397: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 10398: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
10399: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 10400: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 10401: */
1.145 brouard 10402: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
10403: 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 10404: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
10405: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 10406: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 10407: 4 covariates (3 plus signs)
10408: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
10409: */
1.230 brouard 10410: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 10411: * individual dummy, fixed or varying:
10412: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
10413: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 10414: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
10415: * V1 df, V2 qf, V3 & V4 dv, V5 qv
10416: * Tmodelind[1]@9={9,0,3,2,}*/
10417: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
10418: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 10419: * individual quantitative, fixed or varying:
10420: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
10421: * 3, 1, 0, 0, 0, 0, 0, 0},
10422: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 10423: /* Main decodemodel */
10424:
1.187 brouard 10425:
1.223 brouard 10426: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 10427: goto end;
10428:
1.137 brouard 10429: if((double)(lastobs-imx)/(double)imx > 1.10){
10430: nbwarn++;
10431: 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);
10432: 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);
10433: }
1.136 brouard 10434: /* if(mle==1){*/
1.137 brouard 10435: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
10436: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 10437: }
10438:
10439: /*-calculation of age at interview from date of interview and age at death -*/
10440: agev=matrix(1,maxwav,1,imx);
10441:
10442: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
10443: goto end;
10444:
1.126 brouard 10445:
1.136 brouard 10446: agegomp=(int)agemin;
10447: free_vector(moisnais,1,n);
10448: free_vector(annais,1,n);
1.126 brouard 10449: /* free_matrix(mint,1,maxwav,1,n);
10450: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 10451: /* free_vector(moisdc,1,n); */
10452: /* free_vector(andc,1,n); */
1.145 brouard 10453: /* */
10454:
1.126 brouard 10455: wav=ivector(1,imx);
1.214 brouard 10456: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
10457: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
10458: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
10459: 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.*/
10460: bh=imatrix(1,lastpass-firstpass+2,1,imx);
10461: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 10462:
10463: /* Concatenates waves */
1.214 brouard 10464: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
10465: Death is a valid wave (if date is known).
10466: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
10467: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
10468: and mw[mi+1][i]. dh depends on stepm.
10469: */
10470:
1.126 brouard 10471: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.248 brouard 10472: /* Concatenates waves */
1.145 brouard 10473:
1.215 brouard 10474: free_vector(moisdc,1,n);
10475: free_vector(andc,1,n);
10476:
1.126 brouard 10477: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
10478: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
10479: ncodemax[1]=1;
1.145 brouard 10480: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 10481: cptcoveff=0;
1.220 brouard 10482: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
10483: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 10484: }
10485:
10486: ncovcombmax=pow(2,cptcoveff);
10487: invalidvarcomb=ivector(1, ncovcombmax);
10488: for(i=1;i<ncovcombmax;i++)
10489: invalidvarcomb[i]=0;
10490:
1.211 brouard 10491: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 10492: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 10493: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 10494:
1.200 brouard 10495: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 10496: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 10497: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 10498: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
10499: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
10500: * (currently 0 or 1) in the data.
10501: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
10502: * corresponding modality (h,j).
10503: */
10504:
1.145 brouard 10505: h=0;
10506: /*if (cptcovn > 0) */
1.126 brouard 10507: m=pow(2,cptcoveff);
10508:
1.144 brouard 10509: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 10510: * For k=4 covariates, h goes from 1 to m=2**k
10511: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
10512: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 10513: * h\k 1 2 3 4
1.143 brouard 10514: *______________________________
10515: * 1 i=1 1 i=1 1 i=1 1 i=1 1
10516: * 2 2 1 1 1
10517: * 3 i=2 1 2 1 1
10518: * 4 2 2 1 1
10519: * 5 i=3 1 i=2 1 2 1
10520: * 6 2 1 2 1
10521: * 7 i=4 1 2 2 1
10522: * 8 2 2 2 1
1.197 brouard 10523: * 9 i=5 1 i=3 1 i=2 1 2
10524: * 10 2 1 1 2
10525: * 11 i=6 1 2 1 2
10526: * 12 2 2 1 2
10527: * 13 i=7 1 i=4 1 2 2
10528: * 14 2 1 2 2
10529: * 15 i=8 1 2 2 2
10530: * 16 2 2 2 2
1.143 brouard 10531: */
1.212 brouard 10532: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10533: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10534: * and the value of each covariate?
10535: * V1=1, V2=1, V3=2, V4=1 ?
10536: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10537: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10538: * In order to get the real value in the data, we use nbcode
10539: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10540: * We are keeping this crazy system in order to be able (in the future?)
10541: * to have more than 2 values (0 or 1) for a covariate.
10542: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10543: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10544: * bbbbbbbb
10545: * 76543210
10546: * h-1 00000101 (6-1=5)
1.219 brouard 10547: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10548: * &
10549: * 1 00000001 (1)
1.219 brouard 10550: * 00000000 = 1 & ((h-1) >> (k-1))
10551: * +1= 00000001 =1
1.211 brouard 10552: *
10553: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10554: * h' 1101 =2^3+2^2+0x2^1+2^0
10555: * >>k' 11
10556: * & 00000001
10557: * = 00000001
10558: * +1 = 00000010=2 = codtabm(14,3)
10559: * Reverse h=6 and m=16?
10560: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10561: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10562: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10563: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10564: * V3=decodtabm(14,3,2**4)=2
10565: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10566: *(h-1) >> (j-1) 0011 =13 >> 2
10567: * &1 000000001
10568: * = 000000001
10569: * +1= 000000010 =2
10570: * 2211
10571: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10572: * V3=2
1.220 brouard 10573: * codtabm and decodtabm are identical
1.211 brouard 10574: */
10575:
1.145 brouard 10576:
10577: free_ivector(Ndum,-1,NCOVMAX);
10578:
10579:
1.126 brouard 10580:
1.186 brouard 10581: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10582: strcpy(optionfilegnuplot,optionfilefiname);
10583: if(mle==-3)
1.201 brouard 10584: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10585: strcat(optionfilegnuplot,".gp");
10586:
10587: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10588: printf("Problem with file %s",optionfilegnuplot);
10589: }
10590: else{
1.204 brouard 10591: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10592: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10593: //fprintf(ficgp,"set missing 'NaNq'\n");
10594: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10595: }
10596: /* fclose(ficgp);*/
1.186 brouard 10597:
10598:
10599: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10600:
10601: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10602: if(mle==-3)
1.201 brouard 10603: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10604: strcat(optionfilehtm,".htm");
10605: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10606: printf("Problem with %s \n",optionfilehtm);
10607: exit(0);
1.126 brouard 10608: }
10609:
10610: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10611: strcat(optionfilehtmcov,"-cov.htm");
10612: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10613: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10614: }
10615: else{
10616: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10617: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10618: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10619: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10620: }
10621:
1.213 brouard 10622: 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 10623: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10624: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10625: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10626: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10627: \n\
10628: <hr size=\"2\" color=\"#EC5E5E\">\
10629: <ul><li><h4>Parameter files</h4>\n\
10630: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10631: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10632: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10633: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10634: - Date and time at start: %s</ul>\n",\
10635: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10636: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10637: fileres,fileres,\
10638: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10639: fflush(fichtm);
10640:
10641: strcpy(pathr,path);
10642: strcat(pathr,optionfilefiname);
1.184 brouard 10643: #ifdef WIN32
10644: _chdir(optionfilefiname); /* Move to directory named optionfile */
10645: #else
1.126 brouard 10646: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10647: #endif
10648:
1.126 brouard 10649:
1.220 brouard 10650: /* Calculates basic frequencies. Computes observed prevalence at single age
10651: and for any valid combination of covariates
1.126 brouard 10652: and prints on file fileres'p'. */
1.251 brouard 10653: freqsummary(fileres, p, pstart, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
1.227 brouard 10654: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10655:
10656: fprintf(fichtm,"\n");
10657: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10658: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10659: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10660: imx,agemin,agemax,jmin,jmax,jmean);
10661: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10662: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10663: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10664: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10665: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10666:
1.126 brouard 10667: /* For Powell, parameters are in a vector p[] starting at p[1]
10668: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10669: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10670:
10671: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10672: /* For mortality only */
1.126 brouard 10673: if (mle==-3){
1.136 brouard 10674: ximort=matrix(1,NDIM,1,NDIM);
1.248 brouard 10675: for(i=1;i<=NDIM;i++)
10676: for(j=1;j<=NDIM;j++)
10677: ximort[i][j]=0.;
1.186 brouard 10678: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10679: cens=ivector(1,n);
10680: ageexmed=vector(1,n);
10681: agecens=vector(1,n);
10682: dcwave=ivector(1,n);
1.223 brouard 10683:
1.126 brouard 10684: for (i=1; i<=imx; i++){
10685: dcwave[i]=-1;
10686: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10687: if (s[m][i]>nlstate) {
10688: dcwave[i]=m;
10689: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10690: break;
10691: }
1.126 brouard 10692: }
1.226 brouard 10693:
1.126 brouard 10694: for (i=1; i<=imx; i++) {
10695: if (wav[i]>0){
1.226 brouard 10696: ageexmed[i]=agev[mw[1][i]][i];
10697: j=wav[i];
10698: agecens[i]=1.;
10699:
10700: if (ageexmed[i]> 1 && wav[i] > 0){
10701: agecens[i]=agev[mw[j][i]][i];
10702: cens[i]= 1;
10703: }else if (ageexmed[i]< 1)
10704: cens[i]= -1;
10705: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10706: cens[i]=0 ;
1.126 brouard 10707: }
10708: else cens[i]=-1;
10709: }
10710:
10711: for (i=1;i<=NDIM;i++) {
10712: for (j=1;j<=NDIM;j++)
1.226 brouard 10713: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10714: }
10715:
1.145 brouard 10716: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10717: /*printf("%lf %lf", p[1], p[2]);*/
10718:
10719:
1.136 brouard 10720: #ifdef GSL
10721: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10722: #else
1.126 brouard 10723: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10724: #endif
1.201 brouard 10725: strcpy(filerespow,"POW-MORT_");
10726: strcat(filerespow,fileresu);
1.126 brouard 10727: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10728: printf("Problem with resultfile: %s\n", filerespow);
10729: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10730: }
1.136 brouard 10731: #ifdef GSL
10732: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10733: #else
1.126 brouard 10734: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10735: #endif
1.126 brouard 10736: /* for (i=1;i<=nlstate;i++)
10737: for(j=1;j<=nlstate+ndeath;j++)
10738: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10739: */
10740: fprintf(ficrespow,"\n");
1.136 brouard 10741: #ifdef GSL
10742: /* gsl starts here */
10743: T = gsl_multimin_fminimizer_nmsimplex;
10744: gsl_multimin_fminimizer *sfm = NULL;
10745: gsl_vector *ss, *x;
10746: gsl_multimin_function minex_func;
10747:
10748: /* Initial vertex size vector */
10749: ss = gsl_vector_alloc (NDIM);
10750:
10751: if (ss == NULL){
10752: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10753: }
10754: /* Set all step sizes to 1 */
10755: gsl_vector_set_all (ss, 0.001);
10756:
10757: /* Starting point */
1.126 brouard 10758:
1.136 brouard 10759: x = gsl_vector_alloc (NDIM);
10760:
10761: if (x == NULL){
10762: gsl_vector_free(ss);
10763: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10764: }
10765:
10766: /* Initialize method and iterate */
10767: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10768: /* gsl_vector_set(x, 0, 0.0268); */
10769: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10770: gsl_vector_set(x, 0, p[1]);
10771: gsl_vector_set(x, 1, p[2]);
10772:
10773: minex_func.f = &gompertz_f;
10774: minex_func.n = NDIM;
10775: minex_func.params = (void *)&p; /* ??? */
10776:
10777: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10778: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10779:
10780: printf("Iterations beginning .....\n\n");
10781: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10782:
10783: iteri=0;
10784: while (rval == GSL_CONTINUE){
10785: iteri++;
10786: status = gsl_multimin_fminimizer_iterate(sfm);
10787:
10788: if (status) printf("error: %s\n", gsl_strerror (status));
10789: fflush(0);
10790:
10791: if (status)
10792: break;
10793:
10794: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10795: ssval = gsl_multimin_fminimizer_size (sfm);
10796:
10797: if (rval == GSL_SUCCESS)
10798: printf ("converged to a local maximum at\n");
10799:
10800: printf("%5d ", iteri);
10801: for (it = 0; it < NDIM; it++){
10802: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10803: }
10804: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10805: }
10806:
10807: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10808:
10809: gsl_vector_free(x); /* initial values */
10810: gsl_vector_free(ss); /* inital step size */
10811: for (it=0; it<NDIM; it++){
10812: p[it+1]=gsl_vector_get(sfm->x,it);
10813: fprintf(ficrespow," %.12lf", p[it]);
10814: }
10815: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10816: #endif
10817: #ifdef POWELL
10818: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10819: #endif
1.126 brouard 10820: fclose(ficrespow);
10821:
1.203 brouard 10822: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10823:
10824: for(i=1; i <=NDIM; i++)
10825: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10826: matcov[i][j]=matcov[j][i];
1.126 brouard 10827:
10828: printf("\nCovariance matrix\n ");
1.203 brouard 10829: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10830: for(i=1; i <=NDIM; i++) {
10831: for(j=1;j<=NDIM;j++){
1.220 brouard 10832: printf("%f ",matcov[i][j]);
10833: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10834: }
1.203 brouard 10835: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10836: }
10837:
10838: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10839: for (i=1;i<=NDIM;i++) {
1.126 brouard 10840: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10841: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10842: }
1.126 brouard 10843: lsurv=vector(1,AGESUP);
10844: lpop=vector(1,AGESUP);
10845: tpop=vector(1,AGESUP);
10846: lsurv[agegomp]=100000;
10847:
10848: for (k=agegomp;k<=AGESUP;k++) {
10849: agemortsup=k;
10850: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10851: }
10852:
10853: for (k=agegomp;k<agemortsup;k++)
10854: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10855:
10856: for (k=agegomp;k<agemortsup;k++){
10857: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10858: sumlpop=sumlpop+lpop[k];
10859: }
10860:
10861: tpop[agegomp]=sumlpop;
10862: for (k=agegomp;k<(agemortsup-3);k++){
10863: /* tpop[k+1]=2;*/
10864: tpop[k+1]=tpop[k]-lpop[k];
10865: }
10866:
10867:
10868: printf("\nAge lx qx dx Lx Tx e(x)\n");
10869: for (k=agegomp;k<(agemortsup-2);k++)
10870: 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]);
10871:
10872:
10873: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10874: ageminpar=50;
10875: agemaxpar=100;
1.194 brouard 10876: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10877: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10878: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10879: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10880: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10881: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10882: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10883: }else{
10884: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10885: 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 10886: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10887: }
1.201 brouard 10888: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10889: stepm, weightopt,\
10890: model,imx,p,matcov,agemortsup);
10891:
10892: free_vector(lsurv,1,AGESUP);
10893: free_vector(lpop,1,AGESUP);
10894: free_vector(tpop,1,AGESUP);
1.220 brouard 10895: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10896: free_ivector(cens,1,n);
10897: free_vector(agecens,1,n);
10898: free_ivector(dcwave,1,n);
1.220 brouard 10899: #ifdef GSL
1.136 brouard 10900: #endif
1.186 brouard 10901: } /* Endof if mle==-3 mortality only */
1.205 brouard 10902: /* Standard */
10903: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10904: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10905: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10906: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10907: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10908: for (k=1; k<=npar;k++)
10909: printf(" %d %8.5f",k,p[k]);
10910: printf("\n");
1.205 brouard 10911: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10912: /* mlikeli uses func not funcone */
1.247 brouard 10913: /* for(i=1;i<nlstate;i++){ */
10914: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10915: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10916: /* } */
1.205 brouard 10917: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10918: }
10919: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10920: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10921: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10922: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10923: }
10924: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10925: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10926: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10927: for (k=1; k<=npar;k++)
10928: printf(" %d %8.5f",k,p[k]);
10929: printf("\n");
10930:
10931: /*--------- results files --------------*/
1.224 brouard 10932: 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 10933:
10934:
10935: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10936: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10937: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10938: for(i=1,jk=1; i <=nlstate; i++){
10939: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10940: if (k != i) {
10941: printf("%d%d ",i,k);
10942: fprintf(ficlog,"%d%d ",i,k);
10943: fprintf(ficres,"%1d%1d ",i,k);
10944: for(j=1; j <=ncovmodel; j++){
10945: printf("%12.7f ",p[jk]);
10946: fprintf(ficlog,"%12.7f ",p[jk]);
10947: fprintf(ficres,"%12.7f ",p[jk]);
10948: jk++;
10949: }
10950: printf("\n");
10951: fprintf(ficlog,"\n");
10952: fprintf(ficres,"\n");
10953: }
1.126 brouard 10954: }
10955: }
1.203 brouard 10956: if(mle != 0){
10957: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10958: ftolhess=ftol; /* Usually correct */
1.203 brouard 10959: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10960: 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");
10961: 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");
10962: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10963: for(k=1; k <=(nlstate+ndeath); k++){
10964: if (k != i) {
10965: printf("%d%d ",i,k);
10966: fprintf(ficlog,"%d%d ",i,k);
10967: for(j=1; j <=ncovmodel; j++){
10968: 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]));
10969: 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]));
10970: jk++;
10971: }
10972: printf("\n");
10973: fprintf(ficlog,"\n");
10974: }
10975: }
1.193 brouard 10976: }
1.203 brouard 10977: } /* end of hesscov and Wald tests */
1.225 brouard 10978:
1.203 brouard 10979: /* */
1.126 brouard 10980: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
10981: printf("# Scales (for hessian or gradient estimation)\n");
10982: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
10983: for(i=1,jk=1; i <=nlstate; i++){
10984: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 10985: if (j!=i) {
10986: fprintf(ficres,"%1d%1d",i,j);
10987: printf("%1d%1d",i,j);
10988: fprintf(ficlog,"%1d%1d",i,j);
10989: for(k=1; k<=ncovmodel;k++){
10990: printf(" %.5e",delti[jk]);
10991: fprintf(ficlog," %.5e",delti[jk]);
10992: fprintf(ficres," %.5e",delti[jk]);
10993: jk++;
10994: }
10995: printf("\n");
10996: fprintf(ficlog,"\n");
10997: fprintf(ficres,"\n");
10998: }
1.126 brouard 10999: }
11000: }
11001:
11002: 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 11003: if(mle >= 1) /* To big for the screen */
1.126 brouard 11004: 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");
11005: 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");
11006: /* # 121 Var(a12)\n\ */
11007: /* # 122 Cov(b12,a12) Var(b12)\n\ */
11008: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
11009: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
11010: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
11011: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
11012: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
11013: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
11014:
11015:
11016: /* Just to have a covariance matrix which will be more understandable
11017: even is we still don't want to manage dictionary of variables
11018: */
11019: for(itimes=1;itimes<=2;itimes++){
11020: jj=0;
11021: for(i=1; i <=nlstate; i++){
1.225 brouard 11022: for(j=1; j <=nlstate+ndeath; j++){
11023: if(j==i) continue;
11024: for(k=1; k<=ncovmodel;k++){
11025: jj++;
11026: ca[0]= k+'a'-1;ca[1]='\0';
11027: if(itimes==1){
11028: if(mle>=1)
11029: printf("#%1d%1d%d",i,j,k);
11030: fprintf(ficlog,"#%1d%1d%d",i,j,k);
11031: fprintf(ficres,"#%1d%1d%d",i,j,k);
11032: }else{
11033: if(mle>=1)
11034: printf("%1d%1d%d",i,j,k);
11035: fprintf(ficlog,"%1d%1d%d",i,j,k);
11036: fprintf(ficres,"%1d%1d%d",i,j,k);
11037: }
11038: ll=0;
11039: for(li=1;li <=nlstate; li++){
11040: for(lj=1;lj <=nlstate+ndeath; lj++){
11041: if(lj==li) continue;
11042: for(lk=1;lk<=ncovmodel;lk++){
11043: ll++;
11044: if(ll<=jj){
11045: cb[0]= lk +'a'-1;cb[1]='\0';
11046: if(ll<jj){
11047: if(itimes==1){
11048: if(mle>=1)
11049: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11050: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11051: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11052: }else{
11053: if(mle>=1)
11054: printf(" %.5e",matcov[jj][ll]);
11055: fprintf(ficlog," %.5e",matcov[jj][ll]);
11056: fprintf(ficres," %.5e",matcov[jj][ll]);
11057: }
11058: }else{
11059: if(itimes==1){
11060: if(mle>=1)
11061: printf(" Var(%s%1d%1d)",ca,i,j);
11062: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
11063: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
11064: }else{
11065: if(mle>=1)
11066: printf(" %.7e",matcov[jj][ll]);
11067: fprintf(ficlog," %.7e",matcov[jj][ll]);
11068: fprintf(ficres," %.7e",matcov[jj][ll]);
11069: }
11070: }
11071: }
11072: } /* end lk */
11073: } /* end lj */
11074: } /* end li */
11075: if(mle>=1)
11076: printf("\n");
11077: fprintf(ficlog,"\n");
11078: fprintf(ficres,"\n");
11079: numlinepar++;
11080: } /* end k*/
11081: } /*end j */
1.126 brouard 11082: } /* end i */
11083: } /* end itimes */
11084:
11085: fflush(ficlog);
11086: fflush(ficres);
1.225 brouard 11087: while(fgets(line, MAXLINE, ficpar)) {
11088: /* If line starts with a # it is a comment */
11089: if (line[0] == '#') {
11090: numlinepar++;
11091: fputs(line,stdout);
11092: fputs(line,ficparo);
11093: fputs(line,ficlog);
11094: continue;
11095: }else
11096: break;
11097: }
11098:
1.209 brouard 11099: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
11100: /* ungetc(c,ficpar); */
11101: /* fgets(line, MAXLINE, ficpar); */
11102: /* fputs(line,stdout); */
11103: /* fputs(line,ficparo); */
11104: /* } */
11105: /* ungetc(c,ficpar); */
1.126 brouard 11106:
11107: estepm=0;
1.209 brouard 11108: 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 11109:
11110: if (num_filled != 6) {
11111: 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);
11112: 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);
11113: goto end;
11114: }
11115: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
11116: }
11117: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
11118: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
11119:
1.209 brouard 11120: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 11121: if (estepm==0 || estepm < stepm) estepm=stepm;
11122: if (fage <= 2) {
11123: bage = ageminpar;
11124: fage = agemaxpar;
11125: }
11126:
11127: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 11128: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
11129: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 11130:
1.186 brouard 11131: /* Other stuffs, more or less useful */
1.254 brouard 11132: while(fgets(line, MAXLINE, ficpar)) {
11133: /* If line starts with a # it is a comment */
11134: if (line[0] == '#') {
11135: numlinepar++;
11136: fputs(line,stdout);
11137: fputs(line,ficparo);
11138: fputs(line,ficlog);
11139: continue;
11140: }else
11141: break;
11142: }
11143:
11144: 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){
11145:
11146: if (num_filled != 7) {
11147: 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);
11148: 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);
11149: goto end;
11150: }
11151: /* fscanf(ficpar,"begin-prev-date=%lf/%lf/%lf end-prev-date=%lf/%lf/%lf mov_average=%d\n",&jprev1, &mprev1,&anprev1,&jprev2, &mprev2,&anprev2,&mobilav); */
11152: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11153: 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);
11154: 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);
11155: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11156: 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 11157: }
1.254 brouard 11158:
11159: while(fgets(line, MAXLINE, ficpar)) {
11160: /* If line starts with a # it is a comment */
11161: if (line[0] == '#') {
11162: numlinepar++;
11163: fputs(line,stdout);
11164: fputs(line,ficparo);
11165: fputs(line,ficlog);
11166: continue;
11167: }else
11168: break;
1.126 brouard 11169: }
11170:
11171:
11172: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
11173: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
11174:
1.254 brouard 11175: if((num_filled=sscanf(line,"pop_based=%d\n",&popbased)) !=EOF){
11176: if (num_filled != 1) {
11177: 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);
11178: 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);
11179: goto end;
11180: }
11181: printf("pop_based=%d\n",popbased);
11182: fprintf(ficlog,"pop_based=%d\n",popbased);
11183: fprintf(ficparo,"pop_based=%d\n",popbased);
11184: fprintf(ficres,"pop_based=%d\n",popbased);
11185: }
11186:
11187: while(fgets(line, MAXLINE, ficpar)) {
11188: /* If line starts with a # it is a comment */
11189: if (line[0] == '#') {
11190: numlinepar++;
11191: fputs(line,stdout);
11192: fputs(line,ficparo);
11193: fputs(line,ficlog);
11194: continue;
11195: }else
11196: break;
1.126 brouard 11197: }
1.254 brouard 11198: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
11199: /* ungetc(c,ficpar); */
11200: /* fgets(line, MAXLINE, ficpar); */
11201: /* fputs(line,stdout); */
11202: /* fputs(line,ficres); */
11203: /* fputs(line,ficparo); */
11204: /* } */
11205: /* ungetc(c,ficpar); */
11206:
11207: /* fscanf(ficpar,"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); */
11208: 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){
11209: if (num_filled != 8) {
11210: 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);
11211: 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);
11212: goto end;
11213: }
11214: 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);
11215: 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);
11216: 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);
11217: 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);
1.126 brouard 11218: /* day and month of proj2 are not used but only year anproj2.*/
1.217 brouard 11219: }
1.254 brouard 11220: while(fgets(line, MAXLINE, ficpar)) {
11221: /* If line starts with a # it is a comment */
11222: if (line[0] == '#') {
11223: numlinepar++;
11224: fputs(line,stdout);
11225: fputs(line,ficparo);
11226: fputs(line,ficlog);
11227: continue;
11228: }else
11229: break;
11230: }
11231: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
11232: /* ungetc(c,ficpar); */
11233: /* fgets(line, MAXLINE, ficpar); */
11234: /* fputs(line,stdout); */
11235: /* fputs(line,ficparo); */
11236: /* fputs(line,ficres); */
11237: /* } */
11238: /* ungetc(c,ficpar); */
1.217 brouard 11239:
11240: 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);
1.254 brouard 11241: 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){
11242: if (num_filled != 8) {
11243: printf("Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 finloal-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11244: fprintf(ficlog,"Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 finloal-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11245: goto end;
11246: }
11247: 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);
11248: 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);
11249: 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);
11250: 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);
1.217 brouard 11251: /* day and month of proj2 are not used but only year anproj2.*/
1.254 brouard 11252: }
1.230 brouard 11253: /* Results */
1.235 brouard 11254: nresult=0;
1.230 brouard 11255: while(fgets(line, MAXLINE, ficpar)) {
11256: /* If line starts with a # it is a comment */
11257: if (line[0] == '#') {
11258: numlinepar++;
11259: fputs(line,stdout);
11260: fputs(line,ficparo);
11261: fputs(line,ficlog);
1.238 brouard 11262: fputs(line,ficres);
1.230 brouard 11263: continue;
11264: }else
11265: break;
11266: }
1.240 brouard 11267: if (!feof(ficpar))
1.230 brouard 11268: while((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
1.240 brouard 11269: if (num_filled == 0){
1.230 brouard 11270: resultline[0]='\0';
1.253 brouard 11271: printf("Warning %d: no result line should be at minimum 'result: V2=0 V1=1 or result:.\n%s\n", num_filled, line);
1.240 brouard 11272: break;
11273: } else if (num_filled != 1){
1.253 brouard 11274: printf("ERROR %d: result line should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",num_filled, line);
1.230 brouard 11275: }
1.235 brouard 11276: nresult++; /* Sum of resultlines */
11277: printf("Result %d: result=%s\n",nresult, resultline);
11278: if(nresult > MAXRESULTLINES){
11279: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11280: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11281: goto end;
11282: }
11283: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
1.238 brouard 11284: fprintf(ficparo,"result: %s\n",resultline);
11285: fprintf(ficres,"result: %s\n",resultline);
11286: fprintf(ficlog,"result: %s\n",resultline);
1.230 brouard 11287: while(fgets(line, MAXLINE, ficpar)) {
11288: /* If line starts with a # it is a comment */
11289: if (line[0] == '#') {
11290: numlinepar++;
11291: fputs(line,stdout);
11292: fputs(line,ficparo);
1.238 brouard 11293: fputs(line,ficres);
1.230 brouard 11294: fputs(line,ficlog);
11295: continue;
11296: }else
11297: break;
11298: }
11299: if (feof(ficpar))
11300: break;
11301: else{ /* Processess output results for this combination of covariate values */
11302: }
1.240 brouard 11303: } /* end while */
1.230 brouard 11304:
11305:
1.126 brouard 11306:
1.230 brouard 11307: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 11308: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 11309:
11310: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 11311: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 11312: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11313: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11314: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 11315: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11316: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11317: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11318: }else{
1.218 brouard 11319: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 11320: }
11321: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.225 brouard 11322: model,imx,jmin,jmax,jmean,rfileres,popforecast,prevfcast,backcast, estepm, \
11323: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 11324:
1.225 brouard 11325: /*------------ free_vector -------------*/
11326: /* chdir(path); */
1.220 brouard 11327:
1.215 brouard 11328: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
11329: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
11330: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
11331: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 11332: free_lvector(num,1,n);
11333: free_vector(agedc,1,n);
11334: /*free_matrix(covar,0,NCOVMAX,1,n);*/
11335: /*free_matrix(covar,1,NCOVMAX,1,n);*/
11336: fclose(ficparo);
11337: fclose(ficres);
1.220 brouard 11338:
11339:
1.186 brouard 11340: /* Other results (useful)*/
1.220 brouard 11341:
11342:
1.126 brouard 11343: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 11344: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
11345: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 11346: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 11347: fclose(ficrespl);
11348:
11349: /*------------- h Pij x at various ages ------------*/
1.180 brouard 11350: /*#include "hpijx.h"*/
11351: hPijx(p, bage, fage);
1.145 brouard 11352: fclose(ficrespij);
1.227 brouard 11353:
1.220 brouard 11354: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 11355: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 11356: k=1;
1.126 brouard 11357: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 11358:
1.219 brouard 11359: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 11360: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 11361: for(i=1;i<=AGESUP;i++)
1.219 brouard 11362: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 11363: for(k=1;k<=ncovcombmax;k++)
11364: probs[i][j][k]=0.;
1.219 brouard 11365: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
11366: if (mobilav!=0 ||mobilavproj !=0 ) {
11367: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 11368: for(i=1;i<=AGESUP;i++)
11369: for(j=1;j<=nlstate;j++)
11370: for(k=1;k<=ncovcombmax;k++)
11371: mobaverages[i][j][k]=0.;
1.219 brouard 11372: mobaverage=mobaverages;
11373: if (mobilav!=0) {
1.235 brouard 11374: printf("Movingaveraging observed prevalence\n");
1.227 brouard 11375: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
11376: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
11377: printf(" Error in movingaverage mobilav=%d\n",mobilav);
11378: }
1.219 brouard 11379: }
11380: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
11381: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
11382: else if (mobilavproj !=0) {
1.235 brouard 11383: printf("Movingaveraging projected observed prevalence\n");
1.227 brouard 11384: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
11385: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
11386: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
11387: }
1.219 brouard 11388: }
11389: }/* end if moving average */
1.227 brouard 11390:
1.126 brouard 11391: /*---------- Forecasting ------------------*/
11392: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
11393: if(prevfcast==1){
11394: /* if(stepm ==1){*/
1.225 brouard 11395: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 11396: }
1.217 brouard 11397: if(backcast==1){
1.219 brouard 11398: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11399: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11400: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11401:
11402: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
11403:
11404: bprlim=matrix(1,nlstate,1,nlstate);
11405: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
11406: fclose(ficresplb);
11407:
1.222 brouard 11408: hBijx(p, bage, fage, mobaverage);
11409: fclose(ficrespijb);
1.219 brouard 11410: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
11411:
11412: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 11413: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 11414: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11415: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11416: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11417: }
1.217 brouard 11418:
1.186 brouard 11419:
11420: /* ------ Other prevalence ratios------------ */
1.126 brouard 11421:
1.215 brouard 11422: free_ivector(wav,1,imx);
11423: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
11424: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
11425: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 11426:
11427:
1.127 brouard 11428: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 11429:
1.201 brouard 11430: strcpy(filerese,"E_");
11431: strcat(filerese,fileresu);
1.126 brouard 11432: if((ficreseij=fopen(filerese,"w"))==NULL) {
11433: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11434: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11435: }
1.208 brouard 11436: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
11437: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.238 brouard 11438:
11439: pstamp(ficreseij);
1.219 brouard 11440:
1.235 brouard 11441: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11442: if (cptcovn < 1){i1=1;}
11443:
11444: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11445: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11446: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11447: continue;
1.219 brouard 11448: fprintf(ficreseij,"\n#****** ");
1.235 brouard 11449: printf("\n#****** ");
1.225 brouard 11450: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11451: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11452: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11453: }
11454: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11455: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11456: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 11457: }
11458: fprintf(ficreseij,"******\n");
1.235 brouard 11459: printf("******\n");
1.219 brouard 11460:
11461: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11462: oldm=oldms;savm=savms;
1.235 brouard 11463: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 11464:
1.219 brouard 11465: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 11466: }
11467: fclose(ficreseij);
1.208 brouard 11468: printf("done evsij\n");fflush(stdout);
11469: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 11470:
1.227 brouard 11471: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 11472:
11473:
1.201 brouard 11474: strcpy(filerest,"T_");
11475: strcat(filerest,fileresu);
1.127 brouard 11476: if((ficrest=fopen(filerest,"w"))==NULL) {
11477: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
11478: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
11479: }
1.208 brouard 11480: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
11481: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 11482:
1.126 brouard 11483:
1.201 brouard 11484: strcpy(fileresstde,"STDE_");
11485: strcat(fileresstde,fileresu);
1.126 brouard 11486: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 11487: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
11488: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 11489: }
1.227 brouard 11490: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
11491: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 11492:
1.201 brouard 11493: strcpy(filerescve,"CVE_");
11494: strcat(filerescve,fileresu);
1.126 brouard 11495: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 11496: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
11497: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 11498: }
1.227 brouard 11499: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
11500: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 11501:
1.201 brouard 11502: strcpy(fileresv,"V_");
11503: strcat(fileresv,fileresu);
1.126 brouard 11504: if((ficresvij=fopen(fileresv,"w"))==NULL) {
11505: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
11506: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
11507: }
1.227 brouard 11508: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
11509: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 11510:
1.145 brouard 11511: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11512: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
11513:
1.235 brouard 11514: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11515: if (cptcovn < 1){i1=1;}
11516:
11517: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11518: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11519: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11520: continue;
1.242 brouard 11521: printf("\n#****** Result for:");
11522: fprintf(ficrest,"\n#****** Result for:");
11523: fprintf(ficlog,"\n#****** Result for:");
1.227 brouard 11524: for(j=1;j<=cptcoveff;j++){
11525: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11526: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11527: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11528: }
1.235 brouard 11529: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11530: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11531: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11532: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11533: }
1.208 brouard 11534: fprintf(ficrest,"******\n");
1.227 brouard 11535: fprintf(ficlog,"******\n");
11536: printf("******\n");
1.208 brouard 11537:
11538: fprintf(ficresstdeij,"\n#****** ");
11539: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 11540: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11541: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11542: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 11543: }
1.235 brouard 11544: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11545: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11546: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11547: }
1.208 brouard 11548: fprintf(ficresstdeij,"******\n");
11549: fprintf(ficrescveij,"******\n");
11550:
11551: fprintf(ficresvij,"\n#****** ");
1.238 brouard 11552: /* pstamp(ficresvij); */
1.225 brouard 11553: for(j=1;j<=cptcoveff;j++)
1.227 brouard 11554: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11555: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11556: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11557: }
1.208 brouard 11558: fprintf(ficresvij,"******\n");
11559:
11560: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11561: oldm=oldms;savm=savms;
1.235 brouard 11562: printf(" cvevsij ");
11563: fprintf(ficlog, " cvevsij ");
11564: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 11565: printf(" end cvevsij \n ");
11566: fprintf(ficlog, " end cvevsij \n ");
11567:
11568: /*
11569: */
11570: /* goto endfree; */
11571:
11572: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11573: pstamp(ficrest);
11574:
11575:
11576: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 11577: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
11578: cptcod= 0; /* To be deleted */
11579: printf("varevsij vpopbased=%d \n",vpopbased);
11580: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 11581: 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 11582: 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 ");
11583: if(vpopbased==1)
11584: 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);
11585: else
11586: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
11587: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
11588: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
11589: fprintf(ficrest,"\n");
11590: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
11591: epj=vector(1,nlstate+1);
11592: printf("Computing age specific period (stable) prevalences in each health state \n");
11593: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11594: for(age=bage; age <=fage ;age++){
1.235 brouard 11595: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11596: if (vpopbased==1) {
11597: if(mobilav ==0){
11598: for(i=1; i<=nlstate;i++)
11599: prlim[i][i]=probs[(int)age][i][k];
11600: }else{ /* mobilav */
11601: for(i=1; i<=nlstate;i++)
11602: prlim[i][i]=mobaverage[(int)age][i][k];
11603: }
11604: }
1.219 brouard 11605:
1.227 brouard 11606: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11607: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11608: /* printf(" age %4.0f ",age); */
11609: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11610: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11611: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11612: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11613: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11614: }
11615: epj[nlstate+1] +=epj[j];
11616: }
11617: /* printf(" age %4.0f \n",age); */
1.219 brouard 11618:
1.227 brouard 11619: for(i=1, vepp=0.;i <=nlstate;i++)
11620: for(j=1;j <=nlstate;j++)
11621: vepp += vareij[i][j][(int)age];
11622: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11623: for(j=1;j <=nlstate;j++){
11624: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11625: }
11626: fprintf(ficrest,"\n");
11627: }
1.208 brouard 11628: } /* End vpopbased */
11629: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11630: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11631: free_vector(epj,1,nlstate+1);
1.235 brouard 11632: printf("done selection\n");fflush(stdout);
11633: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11634:
1.145 brouard 11635: /*}*/
1.235 brouard 11636: } /* End k selection */
1.227 brouard 11637:
11638: printf("done State-specific expectancies\n");fflush(stdout);
11639: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11640:
1.126 brouard 11641: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11642:
1.201 brouard 11643: strcpy(fileresvpl,"VPL_");
11644: strcat(fileresvpl,fileresu);
1.126 brouard 11645: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11646: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11647: exit(0);
11648: }
1.208 brouard 11649: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11650: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11651:
1.145 brouard 11652: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11653: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11654:
1.235 brouard 11655: i1=pow(2,cptcoveff);
11656: if (cptcovn < 1){i1=1;}
11657:
11658: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11659: for(k=1; k<=i1;k++){
1.253 brouard 11660: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11661: continue;
1.227 brouard 11662: fprintf(ficresvpl,"\n#****** ");
11663: printf("\n#****** ");
11664: fprintf(ficlog,"\n#****** ");
11665: for(j=1;j<=cptcoveff;j++) {
11666: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11667: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11668: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11669: }
1.235 brouard 11670: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11671: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11672: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11673: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11674: }
1.227 brouard 11675: fprintf(ficresvpl,"******\n");
11676: printf("******\n");
11677: fprintf(ficlog,"******\n");
11678:
11679: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11680: oldm=oldms;savm=savms;
1.235 brouard 11681: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11682: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11683: /*}*/
1.126 brouard 11684: }
1.227 brouard 11685:
1.126 brouard 11686: fclose(ficresvpl);
1.208 brouard 11687: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11688: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11689:
11690: free_vector(weight,1,n);
11691: free_imatrix(Tvard,1,NCOVMAX,1,2);
11692: free_imatrix(s,1,maxwav+1,1,n);
11693: free_matrix(anint,1,maxwav,1,n);
11694: free_matrix(mint,1,maxwav,1,n);
11695: free_ivector(cod,1,n);
11696: free_ivector(tab,1,NCOVMAX);
11697: fclose(ficresstdeij);
11698: fclose(ficrescveij);
11699: fclose(ficresvij);
11700: fclose(ficrest);
11701: fclose(ficpar);
11702:
11703:
1.126 brouard 11704: /*---------- End : free ----------------*/
1.219 brouard 11705: if (mobilav!=0 ||mobilavproj !=0)
11706: 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 11707: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11708: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11709: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11710: } /* mle==-3 arrives here for freeing */
1.227 brouard 11711: /* endfree:*/
11712: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11713: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11714: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11715: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11716: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11717: free_matrix(coqvar,1,maxwav,1,n);
11718: free_matrix(covar,0,NCOVMAX,1,n);
11719: free_matrix(matcov,1,npar,1,npar);
11720: free_matrix(hess,1,npar,1,npar);
11721: /*free_vector(delti,1,npar);*/
11722: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11723: free_matrix(agev,1,maxwav,1,imx);
11724: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11725:
11726: free_ivector(ncodemax,1,NCOVMAX);
11727: free_ivector(ncodemaxwundef,1,NCOVMAX);
11728: free_ivector(Dummy,-1,NCOVMAX);
11729: free_ivector(Fixed,-1,NCOVMAX);
1.238 brouard 11730: free_ivector(DummyV,1,NCOVMAX);
11731: free_ivector(FixedV,1,NCOVMAX);
1.227 brouard 11732: free_ivector(Typevar,-1,NCOVMAX);
11733: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11734: free_ivector(TvarsQ,1,NCOVMAX);
11735: free_ivector(TvarsQind,1,NCOVMAX);
11736: free_ivector(TvarsD,1,NCOVMAX);
11737: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11738: free_ivector(TvarFD,1,NCOVMAX);
11739: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11740: free_ivector(TvarF,1,NCOVMAX);
11741: free_ivector(TvarFind,1,NCOVMAX);
11742: free_ivector(TvarV,1,NCOVMAX);
11743: free_ivector(TvarVind,1,NCOVMAX);
11744: free_ivector(TvarA,1,NCOVMAX);
11745: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11746: free_ivector(TvarFQ,1,NCOVMAX);
11747: free_ivector(TvarFQind,1,NCOVMAX);
11748: free_ivector(TvarVD,1,NCOVMAX);
11749: free_ivector(TvarVDind,1,NCOVMAX);
11750: free_ivector(TvarVQ,1,NCOVMAX);
11751: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11752: free_ivector(Tvarsel,1,NCOVMAX);
11753: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11754: free_ivector(Tposprod,1,NCOVMAX);
11755: free_ivector(Tprod,1,NCOVMAX);
11756: free_ivector(Tvaraff,1,NCOVMAX);
11757: free_ivector(invalidvarcomb,1,ncovcombmax);
11758: free_ivector(Tage,1,NCOVMAX);
11759: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11760: free_ivector(TmodelInvind,1,NCOVMAX);
11761: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11762:
11763: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11764: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11765: fflush(fichtm);
11766: fflush(ficgp);
11767:
1.227 brouard 11768:
1.126 brouard 11769: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11770: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11771: 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 11772: }else{
11773: printf("End of Imach\n");
11774: fprintf(ficlog,"End of Imach\n");
11775: }
11776: printf("See log file on %s\n",filelog);
11777: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11778: /*(void) gettimeofday(&end_time,&tzp);*/
11779: rend_time = time(NULL);
11780: end_time = *localtime(&rend_time);
11781: /* tml = *localtime(&end_time.tm_sec); */
11782: strcpy(strtend,asctime(&end_time));
1.126 brouard 11783: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11784: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11785: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11786:
1.157 brouard 11787: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11788: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11789: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11790: /* printf("Total time was %d uSec.\n", total_usecs);*/
11791: /* if(fileappend(fichtm,optionfilehtm)){ */
11792: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11793: fclose(fichtm);
11794: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11795: fclose(fichtmcov);
11796: fclose(ficgp);
11797: fclose(ficlog);
11798: /*------ End -----------*/
1.227 brouard 11799:
11800:
11801: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11802: #ifdef WIN32
1.227 brouard 11803: if (_chdir(pathcd) != 0)
11804: printf("Can't move to directory %s!\n",path);
11805: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11806: #else
1.227 brouard 11807: if(chdir(pathcd) != 0)
11808: printf("Can't move to directory %s!\n", path);
11809: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11810: #endif
1.126 brouard 11811: printf("Current directory %s!\n",pathcd);
11812: /*strcat(plotcmd,CHARSEPARATOR);*/
11813: sprintf(plotcmd,"gnuplot");
1.157 brouard 11814: #ifdef _WIN32
1.126 brouard 11815: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11816: #endif
11817: if(!stat(plotcmd,&info)){
1.158 brouard 11818: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11819: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11820: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11821: }else
11822: strcpy(pplotcmd,plotcmd);
1.157 brouard 11823: #ifdef __unix
1.126 brouard 11824: strcpy(plotcmd,GNUPLOTPROGRAM);
11825: if(!stat(plotcmd,&info)){
1.158 brouard 11826: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11827: }else
11828: strcpy(pplotcmd,plotcmd);
11829: #endif
11830: }else
11831: strcpy(pplotcmd,plotcmd);
11832:
11833: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 11834: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 11835:
1.126 brouard 11836: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 11837: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 11838: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 11839: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 11840: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 11841: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 11842: }
1.158 brouard 11843: printf(" Successful, please wait...");
1.126 brouard 11844: while (z[0] != 'q') {
11845: /* chdir(path); */
1.154 brouard 11846: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 11847: scanf("%s",z);
11848: /* if (z[0] == 'c') system("./imach"); */
11849: if (z[0] == 'e') {
1.158 brouard 11850: #ifdef __APPLE__
1.152 brouard 11851: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 11852: #elif __linux
11853: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 11854: #else
1.152 brouard 11855: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 11856: #endif
11857: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
11858: system(pplotcmd);
1.126 brouard 11859: }
11860: else if (z[0] == 'g') system(plotcmd);
11861: else if (z[0] == 'q') exit(0);
11862: }
1.227 brouard 11863: end:
1.126 brouard 11864: while (z[0] != 'q') {
1.195 brouard 11865: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 11866: scanf("%s",z);
11867: }
11868: }
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