Annotation of imach/src/imach.c, revision 1.249
1.249 ! brouard 1: /* $Id: imach.c,v 1.248 2016/09/07 14:10:18 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.249 ! brouard 4: Revision 1.248 2016/09/07 14:10:18 brouard
! 5: *** empty log message ***
! 6:
1.248 brouard 7: Revision 1.247 2016/09/02 11:11:21 brouard
8: *** empty log message ***
9:
1.247 brouard 10: Revision 1.246 2016/09/02 08:49:22 brouard
11: *** empty log message ***
12:
1.246 brouard 13: Revision 1.245 2016/09/02 07:25:01 brouard
14: *** empty log message ***
15:
1.245 brouard 16: Revision 1.244 2016/09/02 07:17:34 brouard
17: *** empty log message ***
18:
1.244 brouard 19: Revision 1.243 2016/09/02 06:45:35 brouard
20: *** empty log message ***
21:
1.243 brouard 22: Revision 1.242 2016/08/30 15:01:20 brouard
23: Summary: Fixing a lots
24:
1.242 brouard 25: Revision 1.241 2016/08/29 17:17:25 brouard
26: Summary: gnuplot problem in Back projection to fix
27:
1.241 brouard 28: Revision 1.240 2016/08/29 07:53:18 brouard
29: Summary: Better
30:
1.240 brouard 31: Revision 1.239 2016/08/26 15:51:03 brouard
32: Summary: Improvement in Powell output in order to copy and paste
33:
34: Author:
35:
1.239 brouard 36: Revision 1.238 2016/08/26 14:23:35 brouard
37: Summary: Starting tests of 0.99
38:
1.238 brouard 39: Revision 1.237 2016/08/26 09:20:19 brouard
40: Summary: to valgrind
41:
1.237 brouard 42: Revision 1.236 2016/08/25 10:50:18 brouard
43: *** empty log message ***
44:
1.236 brouard 45: Revision 1.235 2016/08/25 06:59:23 brouard
46: *** empty log message ***
47:
1.235 brouard 48: Revision 1.234 2016/08/23 16:51:20 brouard
49: *** empty log message ***
50:
1.234 brouard 51: Revision 1.233 2016/08/23 07:40:50 brouard
52: Summary: not working
53:
1.233 brouard 54: Revision 1.232 2016/08/22 14:20:21 brouard
55: Summary: not working
56:
1.232 brouard 57: Revision 1.231 2016/08/22 07:17:15 brouard
58: Summary: not working
59:
1.231 brouard 60: Revision 1.230 2016/08/22 06:55:53 brouard
61: Summary: Not working
62:
1.230 brouard 63: Revision 1.229 2016/07/23 09:45:53 brouard
64: Summary: Completing for func too
65:
1.229 brouard 66: Revision 1.228 2016/07/22 17:45:30 brouard
67: Summary: Fixing some arrays, still debugging
68:
1.227 brouard 69: Revision 1.226 2016/07/12 18:42:34 brouard
70: Summary: temp
71:
1.226 brouard 72: Revision 1.225 2016/07/12 08:40:03 brouard
73: Summary: saving but not running
74:
1.225 brouard 75: Revision 1.224 2016/07/01 13:16:01 brouard
76: Summary: Fixes
77:
1.224 brouard 78: Revision 1.223 2016/02/19 09:23:35 brouard
79: Summary: temporary
80:
1.223 brouard 81: Revision 1.222 2016/02/17 08:14:50 brouard
82: Summary: Probably last 0.98 stable version 0.98r6
83:
1.222 brouard 84: Revision 1.221 2016/02/15 23:35:36 brouard
85: Summary: minor bug
86:
1.220 brouard 87: Revision 1.219 2016/02/15 00:48:12 brouard
88: *** empty log message ***
89:
1.219 brouard 90: Revision 1.218 2016/02/12 11:29:23 brouard
91: Summary: 0.99 Back projections
92:
1.218 brouard 93: Revision 1.217 2015/12/23 17:18:31 brouard
94: Summary: Experimental backcast
95:
1.217 brouard 96: Revision 1.216 2015/12/18 17:32:11 brouard
97: Summary: 0.98r4 Warning and status=-2
98:
99: Version 0.98r4 is now:
100: - displaying an error when status is -1, date of interview unknown and date of death known;
101: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
102: Older changes concerning s=-2, dating from 2005 have been supersed.
103:
1.216 brouard 104: Revision 1.215 2015/12/16 08:52:24 brouard
105: Summary: 0.98r4 working
106:
1.215 brouard 107: Revision 1.214 2015/12/16 06:57:54 brouard
108: Summary: temporary not working
109:
1.214 brouard 110: Revision 1.213 2015/12/11 18:22:17 brouard
111: Summary: 0.98r4
112:
1.213 brouard 113: Revision 1.212 2015/11/21 12:47:24 brouard
114: Summary: minor typo
115:
1.212 brouard 116: Revision 1.211 2015/11/21 12:41:11 brouard
117: Summary: 0.98r3 with some graph of projected cross-sectional
118:
119: Author: Nicolas Brouard
120:
1.211 brouard 121: Revision 1.210 2015/11/18 17:41:20 brouard
122: Summary: Start working on projected prevalences
123:
1.210 brouard 124: Revision 1.209 2015/11/17 22:12:03 brouard
125: Summary: Adding ftolpl parameter
126: Author: N Brouard
127:
128: We had difficulties to get smoothed confidence intervals. It was due
129: to the period prevalence which wasn't computed accurately. The inner
130: parameter ftolpl is now an outer parameter of the .imach parameter
131: file after estepm. If ftolpl is small 1.e-4 and estepm too,
132: computation are long.
133:
1.209 brouard 134: Revision 1.208 2015/11/17 14:31:57 brouard
135: Summary: temporary
136:
1.208 brouard 137: Revision 1.207 2015/10/27 17:36:57 brouard
138: *** empty log message ***
139:
1.207 brouard 140: Revision 1.206 2015/10/24 07:14:11 brouard
141: *** empty log message ***
142:
1.206 brouard 143: Revision 1.205 2015/10/23 15:50:53 brouard
144: Summary: 0.98r3 some clarification for graphs on likelihood contributions
145:
1.205 brouard 146: Revision 1.204 2015/10/01 16:20:26 brouard
147: Summary: Some new graphs of contribution to likelihood
148:
1.204 brouard 149: Revision 1.203 2015/09/30 17:45:14 brouard
150: Summary: looking at better estimation of the hessian
151:
152: Also a better criteria for convergence to the period prevalence And
153: therefore adding the number of years needed to converge. (The
154: prevalence in any alive state shold sum to one
155:
1.203 brouard 156: Revision 1.202 2015/09/22 19:45:16 brouard
157: Summary: Adding some overall graph on contribution to likelihood. Might change
158:
1.202 brouard 159: Revision 1.201 2015/09/15 17:34:58 brouard
160: Summary: 0.98r0
161:
162: - Some new graphs like suvival functions
163: - Some bugs fixed like model=1+age+V2.
164:
1.201 brouard 165: Revision 1.200 2015/09/09 16:53:55 brouard
166: Summary: Big bug thanks to Flavia
167:
168: Even model=1+age+V2. did not work anymore
169:
1.200 brouard 170: Revision 1.199 2015/09/07 14:09:23 brouard
171: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
172:
1.199 brouard 173: Revision 1.198 2015/09/03 07:14:39 brouard
174: Summary: 0.98q5 Flavia
175:
1.198 brouard 176: Revision 1.197 2015/09/01 18:24:39 brouard
177: *** empty log message ***
178:
1.197 brouard 179: Revision 1.196 2015/08/18 23:17:52 brouard
180: Summary: 0.98q5
181:
1.196 brouard 182: Revision 1.195 2015/08/18 16:28:39 brouard
183: Summary: Adding a hack for testing purpose
184:
185: After reading the title, ftol and model lines, if the comment line has
186: a q, starting with #q, the answer at the end of the run is quit. It
187: permits to run test files in batch with ctest. The former workaround was
188: $ echo q | imach foo.imach
189:
1.195 brouard 190: Revision 1.194 2015/08/18 13:32:00 brouard
191: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
192:
1.194 brouard 193: Revision 1.193 2015/08/04 07:17:42 brouard
194: Summary: 0.98q4
195:
1.193 brouard 196: Revision 1.192 2015/07/16 16:49:02 brouard
197: Summary: Fixing some outputs
198:
1.192 brouard 199: Revision 1.191 2015/07/14 10:00:33 brouard
200: Summary: Some fixes
201:
1.191 brouard 202: Revision 1.190 2015/05/05 08:51:13 brouard
203: Summary: Adding digits in output parameters (7 digits instead of 6)
204:
205: Fix 1+age+.
206:
1.190 brouard 207: Revision 1.189 2015/04/30 14:45:16 brouard
208: Summary: 0.98q2
209:
1.189 brouard 210: Revision 1.188 2015/04/30 08:27:53 brouard
211: *** empty log message ***
212:
1.188 brouard 213: Revision 1.187 2015/04/29 09:11:15 brouard
214: *** empty log message ***
215:
1.187 brouard 216: Revision 1.186 2015/04/23 12:01:52 brouard
217: Summary: V1*age is working now, version 0.98q1
218:
219: Some codes had been disabled in order to simplify and Vn*age was
220: working in the optimization phase, ie, giving correct MLE parameters,
221: but, as usual, outputs were not correct and program core dumped.
222:
1.186 brouard 223: Revision 1.185 2015/03/11 13:26:42 brouard
224: Summary: Inclusion of compile and links command line for Intel Compiler
225:
1.185 brouard 226: Revision 1.184 2015/03/11 11:52:39 brouard
227: Summary: Back from Windows 8. Intel Compiler
228:
1.184 brouard 229: Revision 1.183 2015/03/10 20:34:32 brouard
230: Summary: 0.98q0, trying with directest, mnbrak fixed
231:
232: We use directest instead of original Powell test; probably no
233: incidence on the results, but better justifications;
234: We fixed Numerical Recipes mnbrak routine which was wrong and gave
235: wrong results.
236:
1.183 brouard 237: Revision 1.182 2015/02/12 08:19:57 brouard
238: Summary: Trying to keep directest which seems simpler and more general
239: Author: Nicolas Brouard
240:
1.182 brouard 241: Revision 1.181 2015/02/11 23:22:24 brouard
242: Summary: Comments on Powell added
243:
244: Author:
245:
1.181 brouard 246: Revision 1.180 2015/02/11 17:33:45 brouard
247: Summary: Finishing move from main to function (hpijx and prevalence_limit)
248:
1.180 brouard 249: Revision 1.179 2015/01/04 09:57:06 brouard
250: Summary: back to OS/X
251:
1.179 brouard 252: Revision 1.178 2015/01/04 09:35:48 brouard
253: *** empty log message ***
254:
1.178 brouard 255: Revision 1.177 2015/01/03 18:40:56 brouard
256: Summary: Still testing ilc32 on OSX
257:
1.177 brouard 258: Revision 1.176 2015/01/03 16:45:04 brouard
259: *** empty log message ***
260:
1.176 brouard 261: Revision 1.175 2015/01/03 16:33:42 brouard
262: *** empty log message ***
263:
1.175 brouard 264: Revision 1.174 2015/01/03 16:15:49 brouard
265: Summary: Still in cross-compilation
266:
1.174 brouard 267: Revision 1.173 2015/01/03 12:06:26 brouard
268: Summary: trying to detect cross-compilation
269:
1.173 brouard 270: Revision 1.172 2014/12/27 12:07:47 brouard
271: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
272:
1.172 brouard 273: Revision 1.171 2014/12/23 13:26:59 brouard
274: Summary: Back from Visual C
275:
276: Still problem with utsname.h on Windows
277:
1.171 brouard 278: Revision 1.170 2014/12/23 11:17:12 brouard
279: Summary: Cleaning some \%% back to %%
280:
281: The escape was mandatory for a specific compiler (which one?), but too many warnings.
282:
1.170 brouard 283: Revision 1.169 2014/12/22 23:08:31 brouard
284: Summary: 0.98p
285:
286: Outputs some informations on compiler used, OS etc. Testing on different platforms.
287:
1.169 brouard 288: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 289: Summary: update
1.169 brouard 290:
1.168 brouard 291: Revision 1.167 2014/12/22 13:50:56 brouard
292: Summary: Testing uname and compiler version and if compiled 32 or 64
293:
294: Testing on Linux 64
295:
1.167 brouard 296: Revision 1.166 2014/12/22 11:40:47 brouard
297: *** empty log message ***
298:
1.166 brouard 299: Revision 1.165 2014/12/16 11:20:36 brouard
300: Summary: After compiling on Visual C
301:
302: * imach.c (Module): Merging 1.61 to 1.162
303:
1.165 brouard 304: Revision 1.164 2014/12/16 10:52:11 brouard
305: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
306:
307: * imach.c (Module): Merging 1.61 to 1.162
308:
1.164 brouard 309: Revision 1.163 2014/12/16 10:30:11 brouard
310: * imach.c (Module): Merging 1.61 to 1.162
311:
1.163 brouard 312: Revision 1.162 2014/09/25 11:43:39 brouard
313: Summary: temporary backup 0.99!
314:
1.162 brouard 315: Revision 1.1 2014/09/16 11:06:58 brouard
316: Summary: With some code (wrong) for nlopt
317:
318: Author:
319:
320: Revision 1.161 2014/09/15 20:41:41 brouard
321: Summary: Problem with macro SQR on Intel compiler
322:
1.161 brouard 323: Revision 1.160 2014/09/02 09:24:05 brouard
324: *** empty log message ***
325:
1.160 brouard 326: Revision 1.159 2014/09/01 10:34:10 brouard
327: Summary: WIN32
328: Author: Brouard
329:
1.159 brouard 330: Revision 1.158 2014/08/27 17:11:51 brouard
331: *** empty log message ***
332:
1.158 brouard 333: Revision 1.157 2014/08/27 16:26:55 brouard
334: Summary: Preparing windows Visual studio version
335: Author: Brouard
336:
337: In order to compile on Visual studio, time.h is now correct and time_t
338: and tm struct should be used. difftime should be used but sometimes I
339: just make the differences in raw time format (time(&now).
340: Trying to suppress #ifdef LINUX
341: Add xdg-open for __linux in order to open default browser.
342:
1.157 brouard 343: Revision 1.156 2014/08/25 20:10:10 brouard
344: *** empty log message ***
345:
1.156 brouard 346: Revision 1.155 2014/08/25 18:32:34 brouard
347: Summary: New compile, minor changes
348: Author: Brouard
349:
1.155 brouard 350: Revision 1.154 2014/06/20 17:32:08 brouard
351: Summary: Outputs now all graphs of convergence to period prevalence
352:
1.154 brouard 353: Revision 1.153 2014/06/20 16:45:46 brouard
354: Summary: If 3 live state, convergence to period prevalence on same graph
355: Author: Brouard
356:
1.153 brouard 357: Revision 1.152 2014/06/18 17:54:09 brouard
358: Summary: open browser, use gnuplot on same dir than imach if not found in the path
359:
1.152 brouard 360: Revision 1.151 2014/06/18 16:43:30 brouard
361: *** empty log message ***
362:
1.151 brouard 363: Revision 1.150 2014/06/18 16:42:35 brouard
364: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
365: Author: brouard
366:
1.150 brouard 367: Revision 1.149 2014/06/18 15:51:14 brouard
368: Summary: Some fixes in parameter files errors
369: Author: Nicolas Brouard
370:
1.149 brouard 371: Revision 1.148 2014/06/17 17:38:48 brouard
372: Summary: Nothing new
373: Author: Brouard
374:
375: Just a new packaging for OS/X version 0.98nS
376:
1.148 brouard 377: Revision 1.147 2014/06/16 10:33:11 brouard
378: *** empty log message ***
379:
1.147 brouard 380: Revision 1.146 2014/06/16 10:20:28 brouard
381: Summary: Merge
382: Author: Brouard
383:
384: Merge, before building revised version.
385:
1.146 brouard 386: Revision 1.145 2014/06/10 21:23:15 brouard
387: Summary: Debugging with valgrind
388: Author: Nicolas Brouard
389:
390: Lot of changes in order to output the results with some covariates
391: After the Edimburgh REVES conference 2014, it seems mandatory to
392: improve the code.
393: No more memory valgrind error but a lot has to be done in order to
394: continue the work of splitting the code into subroutines.
395: Also, decodemodel has been improved. Tricode is still not
396: optimal. nbcode should be improved. Documentation has been added in
397: the source code.
398:
1.144 brouard 399: Revision 1.143 2014/01/26 09:45:38 brouard
400: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
401:
402: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
403: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
404:
1.143 brouard 405: Revision 1.142 2014/01/26 03:57:36 brouard
406: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
407:
408: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
409:
1.142 brouard 410: Revision 1.141 2014/01/26 02:42:01 brouard
411: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
412:
1.141 brouard 413: Revision 1.140 2011/09/02 10:37:54 brouard
414: Summary: times.h is ok with mingw32 now.
415:
1.140 brouard 416: Revision 1.139 2010/06/14 07:50:17 brouard
417: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
418: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
419:
1.139 brouard 420: Revision 1.138 2010/04/30 18:19:40 brouard
421: *** empty log message ***
422:
1.138 brouard 423: Revision 1.137 2010/04/29 18:11:38 brouard
424: (Module): Checking covariates for more complex models
425: than V1+V2. A lot of change to be done. Unstable.
426:
1.137 brouard 427: Revision 1.136 2010/04/26 20:30:53 brouard
428: (Module): merging some libgsl code. Fixing computation
429: of likelione (using inter/intrapolation if mle = 0) in order to
430: get same likelihood as if mle=1.
431: Some cleaning of code and comments added.
432:
1.136 brouard 433: Revision 1.135 2009/10/29 15:33:14 brouard
434: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
435:
1.135 brouard 436: Revision 1.134 2009/10/29 13:18:53 brouard
437: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
438:
1.134 brouard 439: Revision 1.133 2009/07/06 10:21:25 brouard
440: just nforces
441:
1.133 brouard 442: Revision 1.132 2009/07/06 08:22:05 brouard
443: Many tings
444:
1.132 brouard 445: Revision 1.131 2009/06/20 16:22:47 brouard
446: Some dimensions resccaled
447:
1.131 brouard 448: Revision 1.130 2009/05/26 06:44:34 brouard
449: (Module): Max Covariate is now set to 20 instead of 8. A
450: lot of cleaning with variables initialized to 0. Trying to make
451: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
452:
1.130 brouard 453: Revision 1.129 2007/08/31 13:49:27 lievre
454: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
455:
1.129 lievre 456: Revision 1.128 2006/06/30 13:02:05 brouard
457: (Module): Clarifications on computing e.j
458:
1.128 brouard 459: Revision 1.127 2006/04/28 18:11:50 brouard
460: (Module): Yes the sum of survivors was wrong since
461: imach-114 because nhstepm was no more computed in the age
462: loop. Now we define nhstepma in the age loop.
463: (Module): In order to speed up (in case of numerous covariates) we
464: compute health expectancies (without variances) in a first step
465: and then all the health expectancies with variances or standard
466: deviation (needs data from the Hessian matrices) which slows the
467: computation.
468: In the future we should be able to stop the program is only health
469: expectancies and graph are needed without standard deviations.
470:
1.127 brouard 471: Revision 1.126 2006/04/28 17:23:28 brouard
472: (Module): Yes the sum of survivors was wrong since
473: imach-114 because nhstepm was no more computed in the age
474: loop. Now we define nhstepma in the age loop.
475: Version 0.98h
476:
1.126 brouard 477: Revision 1.125 2006/04/04 15:20:31 lievre
478: Errors in calculation of health expectancies. Age was not initialized.
479: Forecasting file added.
480:
481: Revision 1.124 2006/03/22 17:13:53 lievre
482: Parameters are printed with %lf instead of %f (more numbers after the comma).
483: The log-likelihood is printed in the log file
484:
485: Revision 1.123 2006/03/20 10:52:43 brouard
486: * imach.c (Module): <title> changed, corresponds to .htm file
487: name. <head> headers where missing.
488:
489: * imach.c (Module): Weights can have a decimal point as for
490: English (a comma might work with a correct LC_NUMERIC environment,
491: otherwise the weight is truncated).
492: Modification of warning when the covariates values are not 0 or
493: 1.
494: Version 0.98g
495:
496: Revision 1.122 2006/03/20 09:45:41 brouard
497: (Module): Weights can have a decimal point as for
498: English (a comma might work with a correct LC_NUMERIC environment,
499: otherwise the weight is truncated).
500: Modification of warning when the covariates values are not 0 or
501: 1.
502: Version 0.98g
503:
504: Revision 1.121 2006/03/16 17:45:01 lievre
505: * imach.c (Module): Comments concerning covariates added
506:
507: * imach.c (Module): refinements in the computation of lli if
508: status=-2 in order to have more reliable computation if stepm is
509: not 1 month. Version 0.98f
510:
511: Revision 1.120 2006/03/16 15:10:38 lievre
512: (Module): refinements in the computation of lli if
513: status=-2 in order to have more reliable computation if stepm is
514: not 1 month. Version 0.98f
515:
516: Revision 1.119 2006/03/15 17:42:26 brouard
517: (Module): Bug if status = -2, the loglikelihood was
518: computed as likelihood omitting the logarithm. Version O.98e
519:
520: Revision 1.118 2006/03/14 18:20:07 brouard
521: (Module): varevsij Comments added explaining the second
522: table of variances if popbased=1 .
523: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
524: (Module): Function pstamp added
525: (Module): Version 0.98d
526:
527: Revision 1.117 2006/03/14 17:16:22 brouard
528: (Module): varevsij Comments added explaining the second
529: table of variances if popbased=1 .
530: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
531: (Module): Function pstamp added
532: (Module): Version 0.98d
533:
534: Revision 1.116 2006/03/06 10:29:27 brouard
535: (Module): Variance-covariance wrong links and
536: varian-covariance of ej. is needed (Saito).
537:
538: Revision 1.115 2006/02/27 12:17:45 brouard
539: (Module): One freematrix added in mlikeli! 0.98c
540:
541: Revision 1.114 2006/02/26 12:57:58 brouard
542: (Module): Some improvements in processing parameter
543: filename with strsep.
544:
545: Revision 1.113 2006/02/24 14:20:24 brouard
546: (Module): Memory leaks checks with valgrind and:
547: datafile was not closed, some imatrix were not freed and on matrix
548: allocation too.
549:
550: Revision 1.112 2006/01/30 09:55:26 brouard
551: (Module): Back to gnuplot.exe instead of wgnuplot.exe
552:
553: Revision 1.111 2006/01/25 20:38:18 brouard
554: (Module): Lots of cleaning and bugs added (Gompertz)
555: (Module): Comments can be added in data file. Missing date values
556: can be a simple dot '.'.
557:
558: Revision 1.110 2006/01/25 00:51:50 brouard
559: (Module): Lots of cleaning and bugs added (Gompertz)
560:
561: Revision 1.109 2006/01/24 19:37:15 brouard
562: (Module): Comments (lines starting with a #) are allowed in data.
563:
564: Revision 1.108 2006/01/19 18:05:42 lievre
565: Gnuplot problem appeared...
566: To be fixed
567:
568: Revision 1.107 2006/01/19 16:20:37 brouard
569: Test existence of gnuplot in imach path
570:
571: Revision 1.106 2006/01/19 13:24:36 brouard
572: Some cleaning and links added in html output
573:
574: Revision 1.105 2006/01/05 20:23:19 lievre
575: *** empty log message ***
576:
577: Revision 1.104 2005/09/30 16:11:43 lievre
578: (Module): sump fixed, loop imx fixed, and simplifications.
579: (Module): If the status is missing at the last wave but we know
580: that the person is alive, then we can code his/her status as -2
581: (instead of missing=-1 in earlier versions) and his/her
582: contributions to the likelihood is 1 - Prob of dying from last
583: health status (= 1-p13= p11+p12 in the easiest case of somebody in
584: the healthy state at last known wave). Version is 0.98
585:
586: Revision 1.103 2005/09/30 15:54:49 lievre
587: (Module): sump fixed, loop imx fixed, and simplifications.
588:
589: Revision 1.102 2004/09/15 17:31:30 brouard
590: Add the possibility to read data file including tab characters.
591:
592: Revision 1.101 2004/09/15 10:38:38 brouard
593: Fix on curr_time
594:
595: Revision 1.100 2004/07/12 18:29:06 brouard
596: Add version for Mac OS X. Just define UNIX in Makefile
597:
598: Revision 1.99 2004/06/05 08:57:40 brouard
599: *** empty log message ***
600:
601: Revision 1.98 2004/05/16 15:05:56 brouard
602: New version 0.97 . First attempt to estimate force of mortality
603: directly from the data i.e. without the need of knowing the health
604: state at each age, but using a Gompertz model: log u =a + b*age .
605: This is the basic analysis of mortality and should be done before any
606: other analysis, in order to test if the mortality estimated from the
607: cross-longitudinal survey is different from the mortality estimated
608: from other sources like vital statistic data.
609:
610: The same imach parameter file can be used but the option for mle should be -3.
611:
1.133 brouard 612: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 613: former routines in order to include the new code within the former code.
614:
615: The output is very simple: only an estimate of the intercept and of
616: the slope with 95% confident intervals.
617:
618: Current limitations:
619: A) Even if you enter covariates, i.e. with the
620: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
621: B) There is no computation of Life Expectancy nor Life Table.
622:
623: Revision 1.97 2004/02/20 13:25:42 lievre
624: Version 0.96d. Population forecasting command line is (temporarily)
625: suppressed.
626:
627: Revision 1.96 2003/07/15 15:38:55 brouard
628: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
629: rewritten within the same printf. Workaround: many printfs.
630:
631: Revision 1.95 2003/07/08 07:54:34 brouard
632: * imach.c (Repository):
633: (Repository): Using imachwizard code to output a more meaningful covariance
634: matrix (cov(a12,c31) instead of numbers.
635:
636: Revision 1.94 2003/06/27 13:00:02 brouard
637: Just cleaning
638:
639: Revision 1.93 2003/06/25 16:33:55 brouard
640: (Module): On windows (cygwin) function asctime_r doesn't
641: exist so I changed back to asctime which exists.
642: (Module): Version 0.96b
643:
644: Revision 1.92 2003/06/25 16:30:45 brouard
645: (Module): On windows (cygwin) function asctime_r doesn't
646: exist so I changed back to asctime which exists.
647:
648: Revision 1.91 2003/06/25 15:30:29 brouard
649: * imach.c (Repository): Duplicated warning errors corrected.
650: (Repository): Elapsed time after each iteration is now output. It
651: helps to forecast when convergence will be reached. Elapsed time
652: is stamped in powell. We created a new html file for the graphs
653: concerning matrix of covariance. It has extension -cov.htm.
654:
655: Revision 1.90 2003/06/24 12:34:15 brouard
656: (Module): Some bugs corrected for windows. Also, when
657: mle=-1 a template is output in file "or"mypar.txt with the design
658: of the covariance matrix to be input.
659:
660: Revision 1.89 2003/06/24 12:30:52 brouard
661: (Module): Some bugs corrected for windows. Also, when
662: mle=-1 a template is output in file "or"mypar.txt with the design
663: of the covariance matrix to be input.
664:
665: Revision 1.88 2003/06/23 17:54:56 brouard
666: * 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.
667:
668: Revision 1.87 2003/06/18 12:26:01 brouard
669: Version 0.96
670:
671: Revision 1.86 2003/06/17 20:04:08 brouard
672: (Module): Change position of html and gnuplot routines and added
673: routine fileappend.
674:
675: Revision 1.85 2003/06/17 13:12:43 brouard
676: * imach.c (Repository): Check when date of death was earlier that
677: current date of interview. It may happen when the death was just
678: prior to the death. In this case, dh was negative and likelihood
679: was wrong (infinity). We still send an "Error" but patch by
680: assuming that the date of death was just one stepm after the
681: interview.
682: (Repository): Because some people have very long ID (first column)
683: we changed int to long in num[] and we added a new lvector for
684: memory allocation. But we also truncated to 8 characters (left
685: truncation)
686: (Repository): No more line truncation errors.
687:
688: Revision 1.84 2003/06/13 21:44:43 brouard
689: * imach.c (Repository): Replace "freqsummary" at a correct
690: place. It differs from routine "prevalence" which may be called
691: many times. Probs is memory consuming and must be used with
692: parcimony.
693: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
694:
695: Revision 1.83 2003/06/10 13:39:11 lievre
696: *** empty log message ***
697:
698: Revision 1.82 2003/06/05 15:57:20 brouard
699: Add log in imach.c and fullversion number is now printed.
700:
701: */
702: /*
703: Interpolated Markov Chain
704:
705: Short summary of the programme:
706:
1.227 brouard 707: This program computes Healthy Life Expectancies or State-specific
708: (if states aren't health statuses) Expectancies from
709: cross-longitudinal data. Cross-longitudinal data consist in:
710:
711: -1- a first survey ("cross") where individuals from different ages
712: are interviewed on their health status or degree of disability (in
713: the case of a health survey which is our main interest)
714:
715: -2- at least a second wave of interviews ("longitudinal") which
716: measure each change (if any) in individual health status. Health
717: expectancies are computed from the time spent in each health state
718: according to a model. More health states you consider, more time is
719: necessary to reach the Maximum Likelihood of the parameters involved
720: in the model. The simplest model is the multinomial logistic model
721: where pij is the probability to be observed in state j at the second
722: wave conditional to be observed in state i at the first
723: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
724: etc , where 'age' is age and 'sex' is a covariate. If you want to
725: have a more complex model than "constant and age", you should modify
726: the program where the markup *Covariates have to be included here
727: again* invites you to do it. More covariates you add, slower the
1.126 brouard 728: convergence.
729:
730: The advantage of this computer programme, compared to a simple
731: multinomial logistic model, is clear when the delay between waves is not
732: identical for each individual. Also, if a individual missed an
733: intermediate interview, the information is lost, but taken into
734: account using an interpolation or extrapolation.
735:
736: hPijx is the probability to be observed in state i at age x+h
737: conditional to the observed state i at age x. The delay 'h' can be
738: split into an exact number (nh*stepm) of unobserved intermediate
739: states. This elementary transition (by month, quarter,
740: semester or year) is modelled as a multinomial logistic. The hPx
741: matrix is simply the matrix product of nh*stepm elementary matrices
742: and the contribution of each individual to the likelihood is simply
743: hPijx.
744:
745: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 746: of the life expectancies. It also computes the period (stable) prevalence.
747:
748: Back prevalence and projections:
1.227 brouard 749:
750: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
751: double agemaxpar, double ftolpl, int *ncvyearp, double
752: dateprev1,double dateprev2, int firstpass, int lastpass, int
753: mobilavproj)
754:
755: Computes the back prevalence limit for any combination of
756: covariate values k at any age between ageminpar and agemaxpar and
757: returns it in **bprlim. In the loops,
758:
759: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
760: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
761:
762: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 763: Computes for any combination of covariates k and any age between bage and fage
764: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
765: oldm=oldms;savm=savms;
1.227 brouard 766:
767: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 768: Computes the transition matrix starting at age 'age' over
769: 'nhstepm*hstepm*stepm' months (i.e. until
770: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 771: nhstepm*hstepm matrices.
772:
773: Returns p3mat[i][j][h] after calling
774: p3mat[i][j][h]=matprod2(newm,
775: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
776: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
777: oldm);
1.226 brouard 778:
779: Important routines
780:
781: - func (or funcone), computes logit (pij) distinguishing
782: o fixed variables (single or product dummies or quantitative);
783: o varying variables by:
784: (1) wave (single, product dummies, quantitative),
785: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
786: % fixed dummy (treated) or quantitative (not done because time-consuming);
787: % varying dummy (not done) or quantitative (not done);
788: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
789: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
790: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
791: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
792: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 793:
1.226 brouard 794:
795:
1.133 brouard 796: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
797: Institut national d'études démographiques, Paris.
1.126 brouard 798: This software have been partly granted by Euro-REVES, a concerted action
799: from the European Union.
800: It is copyrighted identically to a GNU software product, ie programme and
801: software can be distributed freely for non commercial use. Latest version
802: can be accessed at http://euroreves.ined.fr/imach .
803:
804: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
805: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
806:
807: **********************************************************************/
808: /*
809: main
810: read parameterfile
811: read datafile
812: concatwav
813: freqsummary
814: if (mle >= 1)
815: mlikeli
816: print results files
817: if mle==1
818: computes hessian
819: read end of parameter file: agemin, agemax, bage, fage, estepm
820: begin-prev-date,...
821: open gnuplot file
822: open html file
1.145 brouard 823: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
824: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
825: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
826: freexexit2 possible for memory heap.
827:
828: h Pij x | pij_nom ficrestpij
829: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
830: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
831: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
832:
833: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
834: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
835: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
836: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
837: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
838:
1.126 brouard 839: forecasting if prevfcast==1 prevforecast call prevalence()
840: health expectancies
841: Variance-covariance of DFLE
842: prevalence()
843: movingaverage()
844: varevsij()
845: if popbased==1 varevsij(,popbased)
846: total life expectancies
847: Variance of period (stable) prevalence
848: end
849: */
850:
1.187 brouard 851: /* #define DEBUG */
852: /* #define DEBUGBRENT */
1.203 brouard 853: /* #define DEBUGLINMIN */
854: /* #define DEBUGHESS */
855: #define DEBUGHESSIJ
1.224 brouard 856: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 857: #define POWELL /* Instead of NLOPT */
1.224 brouard 858: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 859: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
860: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 861:
862: #include <math.h>
863: #include <stdio.h>
864: #include <stdlib.h>
865: #include <string.h>
1.226 brouard 866: #include <ctype.h>
1.159 brouard 867:
868: #ifdef _WIN32
869: #include <io.h>
1.172 brouard 870: #include <windows.h>
871: #include <tchar.h>
1.159 brouard 872: #else
1.126 brouard 873: #include <unistd.h>
1.159 brouard 874: #endif
1.126 brouard 875:
876: #include <limits.h>
877: #include <sys/types.h>
1.171 brouard 878:
879: #if defined(__GNUC__)
880: #include <sys/utsname.h> /* Doesn't work on Windows */
881: #endif
882:
1.126 brouard 883: #include <sys/stat.h>
884: #include <errno.h>
1.159 brouard 885: /* extern int errno; */
1.126 brouard 886:
1.157 brouard 887: /* #ifdef LINUX */
888: /* #include <time.h> */
889: /* #include "timeval.h" */
890: /* #else */
891: /* #include <sys/time.h> */
892: /* #endif */
893:
1.126 brouard 894: #include <time.h>
895:
1.136 brouard 896: #ifdef GSL
897: #include <gsl/gsl_errno.h>
898: #include <gsl/gsl_multimin.h>
899: #endif
900:
1.167 brouard 901:
1.162 brouard 902: #ifdef NLOPT
903: #include <nlopt.h>
904: typedef struct {
905: double (* function)(double [] );
906: } myfunc_data ;
907: #endif
908:
1.126 brouard 909: /* #include <libintl.h> */
910: /* #define _(String) gettext (String) */
911:
1.141 brouard 912: #define MAXLINE 1024 /* Was 256. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 913:
914: #define GNUPLOTPROGRAM "gnuplot"
915: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
916: #define FILENAMELENGTH 132
917:
918: #define GLOCK_ERROR_NOPATH -1 /* empty path */
919: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
920:
1.144 brouard 921: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
922: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 923:
924: #define NINTERVMAX 8
1.144 brouard 925: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
926: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
927: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 928: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 929: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
930: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 931: #define MAXN 20000
1.144 brouard 932: #define YEARM 12. /**< Number of months per year */
1.218 brouard 933: /* #define AGESUP 130 */
934: #define AGESUP 150
935: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 936: #define AGEBASE 40
1.194 brouard 937: #define AGEOVERFLOW 1.e20
1.164 brouard 938: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 939: #ifdef _WIN32
940: #define DIRSEPARATOR '\\'
941: #define CHARSEPARATOR "\\"
942: #define ODIRSEPARATOR '/'
943: #else
1.126 brouard 944: #define DIRSEPARATOR '/'
945: #define CHARSEPARATOR "/"
946: #define ODIRSEPARATOR '\\'
947: #endif
948:
1.249 ! brouard 949: /* $Id: imach.c,v 1.248 2016/09/07 14:10:18 brouard Exp $ */
1.126 brouard 950: /* $State: Exp $ */
1.196 brouard 951: #include "version.h"
952: char version[]=__IMACH_VERSION__;
1.224 brouard 953: 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";
1.249 ! brouard 954: char fullversion[]="$Revision: 1.248 $ $Date: 2016/09/07 14:10:18 $";
1.126 brouard 955: char strstart[80];
956: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 957: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 958: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 959: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
960: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
961: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 962: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
963: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 964: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
965: int cptcovprodnoage=0; /**< Number of covariate products without age */
966: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 967: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
968: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 969: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 970: int nsd=0; /**< Total number of single dummy variables (output) */
971: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 972: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 973: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 974: int ntveff=0; /**< ntveff number of effective time varying variables */
975: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 976: int cptcov=0; /* Working variable */
1.218 brouard 977: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 978: int npar=NPARMAX;
979: int nlstate=2; /* Number of live states */
980: int ndeath=1; /* Number of dead states */
1.130 brouard 981: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 982: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 983: int popbased=0;
984:
985: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 986: int maxwav=0; /* Maxim number of waves */
987: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
988: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
989: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 990: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 991: int mle=1, weightopt=0;
1.126 brouard 992: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
993: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
994: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
995: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 996: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 997: int selected(int kvar); /* Is covariate kvar selected for printing results */
998:
1.130 brouard 999: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 1000: double **matprod2(); /* test */
1.126 brouard 1001: double **oldm, **newm, **savm; /* Working pointers to matrices */
1002: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 1003: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
1004:
1.136 brouard 1005: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 1006: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 1007: FILE *ficlog, *ficrespow;
1.130 brouard 1008: int globpr=0; /* Global variable for printing or not */
1.126 brouard 1009: double fretone; /* Only one call to likelihood */
1.130 brouard 1010: long ipmx=0; /* Number of contributions */
1.126 brouard 1011: double sw; /* Sum of weights */
1012: char filerespow[FILENAMELENGTH];
1013: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
1014: FILE *ficresilk;
1015: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
1016: FILE *ficresprobmorprev;
1017: FILE *fichtm, *fichtmcov; /* Html File */
1018: FILE *ficreseij;
1019: char filerese[FILENAMELENGTH];
1020: FILE *ficresstdeij;
1021: char fileresstde[FILENAMELENGTH];
1022: FILE *ficrescveij;
1023: char filerescve[FILENAMELENGTH];
1024: FILE *ficresvij;
1025: char fileresv[FILENAMELENGTH];
1026: FILE *ficresvpl;
1027: char fileresvpl[FILENAMELENGTH];
1028: char title[MAXLINE];
1.234 brouard 1029: char model[MAXLINE]; /**< The model line */
1.217 brouard 1030: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 1031: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
1032: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
1033: char command[FILENAMELENGTH];
1034: int outcmd=0;
1035:
1.217 brouard 1036: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 1037: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 1038: char filelog[FILENAMELENGTH]; /* Log file */
1039: char filerest[FILENAMELENGTH];
1040: char fileregp[FILENAMELENGTH];
1041: char popfile[FILENAMELENGTH];
1042:
1043: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1044:
1.157 brouard 1045: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1046: /* struct timezone tzp; */
1047: /* extern int gettimeofday(); */
1048: struct tm tml, *gmtime(), *localtime();
1049:
1050: extern time_t time();
1051:
1052: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1053: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1054: struct tm tm;
1055:
1.126 brouard 1056: char strcurr[80], strfor[80];
1057:
1058: char *endptr;
1059: long lval;
1060: double dval;
1061:
1062: #define NR_END 1
1063: #define FREE_ARG char*
1064: #define FTOL 1.0e-10
1065:
1066: #define NRANSI
1.240 brouard 1067: #define ITMAX 200
1068: #define ITPOWMAX 20 /* This is now multiplied by the number of parameters */
1.126 brouard 1069:
1070: #define TOL 2.0e-4
1071:
1072: #define CGOLD 0.3819660
1073: #define ZEPS 1.0e-10
1074: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1075:
1076: #define GOLD 1.618034
1077: #define GLIMIT 100.0
1078: #define TINY 1.0e-20
1079:
1080: static double maxarg1,maxarg2;
1081: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1082: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1083:
1084: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1085: #define rint(a) floor(a+0.5)
1.166 brouard 1086: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1087: #define mytinydouble 1.0e-16
1.166 brouard 1088: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1089: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1090: /* static double dsqrarg; */
1091: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1092: static double sqrarg;
1093: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1094: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1095: int agegomp= AGEGOMP;
1096:
1097: int imx;
1098: int stepm=1;
1099: /* Stepm, step in month: minimum step interpolation*/
1100:
1101: int estepm;
1102: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1103:
1104: int m,nb;
1105: long *num;
1.197 brouard 1106: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1107: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1108: covariate for which somebody answered excluding
1109: undefined. Usually 2: 0 and 1. */
1110: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1111: covariate for which somebody answered including
1112: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1113: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1114: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1115: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1116: double *ageexmed,*agecens;
1117: double dateintmean=0;
1118:
1119: double *weight;
1120: int **s; /* Status */
1.141 brouard 1121: double *agedc;
1.145 brouard 1122: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1123: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1124: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1125: double **coqvar; /* Fixed quantitative covariate iqv */
1126: double ***cotvar; /* Time varying covariate itv */
1127: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1128: double idx;
1129: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1130: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1131: /*k 1 2 3 4 5 6 7 8 9 */
1132: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1133: /* Tndvar[k] 1 2 3 4 5 */
1134: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1135: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1136: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1137: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1138: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1139: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1140: /* Tprod[i]=k 4 7 */
1141: /* Tage[i]=k 5 8 */
1142: /* */
1143: /* Type */
1144: /* V 1 2 3 4 5 */
1145: /* F F V V V */
1146: /* D Q D D Q */
1147: /* */
1148: int *TvarsD;
1149: int *TvarsDind;
1150: int *TvarsQ;
1151: int *TvarsQind;
1152:
1.235 brouard 1153: #define MAXRESULTLINES 10
1154: int nresult=0;
1155: int TKresult[MAXRESULTLINES];
1.237 brouard 1156: int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1157: int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1.235 brouard 1158: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1159: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.237 brouard 1160: double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.235 brouard 1161: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1162:
1.234 brouard 1163: /* 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 1164: 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 */
1165: 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 */
1166: 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 */
1167: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1168: 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 */
1169: 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 1170: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1171: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1172: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1173: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1174: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1175: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1176: 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 */
1177: 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 */
1178:
1.230 brouard 1179: int *Tvarsel; /**< Selected covariates for output */
1180: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1181: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1182: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1183: 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 1184: int *DummyV; /** Dummy[v] 0=dummy (0 1), 1 quantitative */
1185: int *FixedV; /** FixedV[v] 0 fixed, 1 varying */
1.197 brouard 1186: int *Tage;
1.227 brouard 1187: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1188: 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 1189: 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*/
1190: 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 1191: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1192: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1193: int **Tvard;
1194: int *Tprod;/**< Gives the k position of the k1 product */
1.238 brouard 1195: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3 */
1.227 brouard 1196: int *Tposprod; /**< Gives the k1 product from the k position */
1.238 brouard 1197: /* if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2) */
1198: /* Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5(V3*V2)]=2 (2nd product without age) */
1.227 brouard 1199: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1200: double *lsurv, *lpop, *tpop;
1201:
1.231 brouard 1202: #define FD 1; /* Fixed dummy covariate */
1203: #define FQ 2; /* Fixed quantitative covariate */
1204: #define FP 3; /* Fixed product covariate */
1205: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1206: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1207: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1208: #define VD 10; /* Varying dummy covariate */
1209: #define VQ 11; /* Varying quantitative covariate */
1210: #define VP 12; /* Varying product covariate */
1211: #define VPDD 13; /* Varying product dummy*dummy covariate */
1212: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1213: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1214: #define APFD 16; /* Age product * fixed dummy covariate */
1215: #define APFQ 17; /* Age product * fixed quantitative covariate */
1216: #define APVD 18; /* Age product * varying dummy covariate */
1217: #define APVQ 19; /* Age product * varying quantitative covariate */
1218:
1219: #define FTYPE 1; /* Fixed covariate */
1220: #define VTYPE 2; /* Varying covariate (loop in wave) */
1221: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1222:
1223: struct kmodel{
1224: int maintype; /* main type */
1225: int subtype; /* subtype */
1226: };
1227: struct kmodel modell[NCOVMAX];
1228:
1.143 brouard 1229: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1230: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1231:
1232: /**************** split *************************/
1233: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1234: {
1235: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1236: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1237: */
1238: char *ss; /* pointer */
1.186 brouard 1239: int l1=0, l2=0; /* length counters */
1.126 brouard 1240:
1241: l1 = strlen(path ); /* length of path */
1242: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1243: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1244: if ( ss == NULL ) { /* no directory, so determine current directory */
1245: strcpy( name, path ); /* we got the fullname name because no directory */
1246: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1247: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1248: /* get current working directory */
1249: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1250: #ifdef WIN32
1251: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1252: #else
1253: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1254: #endif
1.126 brouard 1255: return( GLOCK_ERROR_GETCWD );
1256: }
1257: /* got dirc from getcwd*/
1258: printf(" DIRC = %s \n",dirc);
1.205 brouard 1259: } else { /* strip directory from path */
1.126 brouard 1260: ss++; /* after this, the filename */
1261: l2 = strlen( ss ); /* length of filename */
1262: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1263: strcpy( name, ss ); /* save file name */
1264: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1265: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1266: printf(" DIRC2 = %s \n",dirc);
1267: }
1268: /* We add a separator at the end of dirc if not exists */
1269: l1 = strlen( dirc ); /* length of directory */
1270: if( dirc[l1-1] != DIRSEPARATOR ){
1271: dirc[l1] = DIRSEPARATOR;
1272: dirc[l1+1] = 0;
1273: printf(" DIRC3 = %s \n",dirc);
1274: }
1275: ss = strrchr( name, '.' ); /* find last / */
1276: if (ss >0){
1277: ss++;
1278: strcpy(ext,ss); /* save extension */
1279: l1= strlen( name);
1280: l2= strlen(ss)+1;
1281: strncpy( finame, name, l1-l2);
1282: finame[l1-l2]= 0;
1283: }
1284:
1285: return( 0 ); /* we're done */
1286: }
1287:
1288:
1289: /******************************************/
1290:
1291: void replace_back_to_slash(char *s, char*t)
1292: {
1293: int i;
1294: int lg=0;
1295: i=0;
1296: lg=strlen(t);
1297: for(i=0; i<= lg; i++) {
1298: (s[i] = t[i]);
1299: if (t[i]== '\\') s[i]='/';
1300: }
1301: }
1302:
1.132 brouard 1303: char *trimbb(char *out, char *in)
1.137 brouard 1304: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1305: char *s;
1306: s=out;
1307: while (*in != '\0'){
1.137 brouard 1308: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1309: in++;
1310: }
1311: *out++ = *in++;
1312: }
1313: *out='\0';
1314: return s;
1315: }
1316:
1.187 brouard 1317: /* char *substrchaine(char *out, char *in, char *chain) */
1318: /* { */
1319: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1320: /* char *s, *t; */
1321: /* t=in;s=out; */
1322: /* while ((*in != *chain) && (*in != '\0')){ */
1323: /* *out++ = *in++; */
1324: /* } */
1325:
1326: /* /\* *in matches *chain *\/ */
1327: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1328: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1329: /* } */
1330: /* in--; chain--; */
1331: /* while ( (*in != '\0')){ */
1332: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1333: /* *out++ = *in++; */
1334: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1335: /* } */
1336: /* *out='\0'; */
1337: /* out=s; */
1338: /* return out; */
1339: /* } */
1340: char *substrchaine(char *out, char *in, char *chain)
1341: {
1342: /* Substract chain 'chain' from 'in', return and output 'out' */
1343: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1344:
1345: char *strloc;
1346:
1347: strcpy (out, in);
1348: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1349: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1350: if(strloc != NULL){
1351: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1352: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1353: /* strcpy (strloc, strloc +strlen(chain));*/
1354: }
1355: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1356: return out;
1357: }
1358:
1359:
1.145 brouard 1360: char *cutl(char *blocc, char *alocc, char *in, char occ)
1361: {
1.187 brouard 1362: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1363: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1364: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1365: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1366: */
1.160 brouard 1367: char *s, *t;
1.145 brouard 1368: t=in;s=in;
1369: while ((*in != occ) && (*in != '\0')){
1370: *alocc++ = *in++;
1371: }
1372: if( *in == occ){
1373: *(alocc)='\0';
1374: s=++in;
1375: }
1376:
1377: if (s == t) {/* occ not found */
1378: *(alocc-(in-s))='\0';
1379: in=s;
1380: }
1381: while ( *in != '\0'){
1382: *blocc++ = *in++;
1383: }
1384:
1385: *blocc='\0';
1386: return t;
1387: }
1.137 brouard 1388: char *cutv(char *blocc, char *alocc, char *in, char occ)
1389: {
1.187 brouard 1390: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1391: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1392: gives blocc="abcdef2ghi" and alocc="j".
1393: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1394: */
1395: char *s, *t;
1396: t=in;s=in;
1397: while (*in != '\0'){
1398: while( *in == occ){
1399: *blocc++ = *in++;
1400: s=in;
1401: }
1402: *blocc++ = *in++;
1403: }
1404: if (s == t) /* occ not found */
1405: *(blocc-(in-s))='\0';
1406: else
1407: *(blocc-(in-s)-1)='\0';
1408: in=s;
1409: while ( *in != '\0'){
1410: *alocc++ = *in++;
1411: }
1412:
1413: *alocc='\0';
1414: return s;
1415: }
1416:
1.126 brouard 1417: int nbocc(char *s, char occ)
1418: {
1419: int i,j=0;
1420: int lg=20;
1421: i=0;
1422: lg=strlen(s);
1423: for(i=0; i<= lg; i++) {
1.234 brouard 1424: if (s[i] == occ ) j++;
1.126 brouard 1425: }
1426: return j;
1427: }
1428:
1.137 brouard 1429: /* void cutv(char *u,char *v, char*t, char occ) */
1430: /* { */
1431: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1432: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1433: /* gives u="abcdef2ghi" and v="j" *\/ */
1434: /* int i,lg,j,p=0; */
1435: /* i=0; */
1436: /* lg=strlen(t); */
1437: /* for(j=0; j<=lg-1; j++) { */
1438: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1439: /* } */
1.126 brouard 1440:
1.137 brouard 1441: /* for(j=0; j<p; j++) { */
1442: /* (u[j] = t[j]); */
1443: /* } */
1444: /* u[p]='\0'; */
1.126 brouard 1445:
1.137 brouard 1446: /* for(j=0; j<= lg; j++) { */
1447: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1448: /* } */
1449: /* } */
1.126 brouard 1450:
1.160 brouard 1451: #ifdef _WIN32
1452: char * strsep(char **pp, const char *delim)
1453: {
1454: char *p, *q;
1455:
1456: if ((p = *pp) == NULL)
1457: return 0;
1458: if ((q = strpbrk (p, delim)) != NULL)
1459: {
1460: *pp = q + 1;
1461: *q = '\0';
1462: }
1463: else
1464: *pp = 0;
1465: return p;
1466: }
1467: #endif
1468:
1.126 brouard 1469: /********************** nrerror ********************/
1470:
1471: void nrerror(char error_text[])
1472: {
1473: fprintf(stderr,"ERREUR ...\n");
1474: fprintf(stderr,"%s\n",error_text);
1475: exit(EXIT_FAILURE);
1476: }
1477: /*********************** vector *******************/
1478: double *vector(int nl, int nh)
1479: {
1480: double *v;
1481: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1482: if (!v) nrerror("allocation failure in vector");
1483: return v-nl+NR_END;
1484: }
1485:
1486: /************************ free vector ******************/
1487: void free_vector(double*v, int nl, int nh)
1488: {
1489: free((FREE_ARG)(v+nl-NR_END));
1490: }
1491:
1492: /************************ivector *******************************/
1493: int *ivector(long nl,long nh)
1494: {
1495: int *v;
1496: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1497: if (!v) nrerror("allocation failure in ivector");
1498: return v-nl+NR_END;
1499: }
1500:
1501: /******************free ivector **************************/
1502: void free_ivector(int *v, long nl, long nh)
1503: {
1504: free((FREE_ARG)(v+nl-NR_END));
1505: }
1506:
1507: /************************lvector *******************************/
1508: long *lvector(long nl,long nh)
1509: {
1510: long *v;
1511: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1512: if (!v) nrerror("allocation failure in ivector");
1513: return v-nl+NR_END;
1514: }
1515:
1516: /******************free lvector **************************/
1517: void free_lvector(long *v, long nl, long nh)
1518: {
1519: free((FREE_ARG)(v+nl-NR_END));
1520: }
1521:
1522: /******************* imatrix *******************************/
1523: int **imatrix(long nrl, long nrh, long ncl, long nch)
1524: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1525: {
1526: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1527: int **m;
1528:
1529: /* allocate pointers to rows */
1530: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1531: if (!m) nrerror("allocation failure 1 in matrix()");
1532: m += NR_END;
1533: m -= nrl;
1534:
1535:
1536: /* allocate rows and set pointers to them */
1537: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1538: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1539: m[nrl] += NR_END;
1540: m[nrl] -= ncl;
1541:
1542: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1543:
1544: /* return pointer to array of pointers to rows */
1545: return m;
1546: }
1547:
1548: /****************** free_imatrix *************************/
1549: void free_imatrix(m,nrl,nrh,ncl,nch)
1550: int **m;
1551: long nch,ncl,nrh,nrl;
1552: /* free an int matrix allocated by imatrix() */
1553: {
1554: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1555: free((FREE_ARG) (m+nrl-NR_END));
1556: }
1557:
1558: /******************* matrix *******************************/
1559: double **matrix(long nrl, long nrh, long ncl, long nch)
1560: {
1561: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1562: double **m;
1563:
1564: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1565: if (!m) nrerror("allocation failure 1 in matrix()");
1566: m += NR_END;
1567: m -= nrl;
1568:
1569: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1570: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1571: m[nrl] += NR_END;
1572: m[nrl] -= ncl;
1573:
1574: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1575: return m;
1.145 brouard 1576: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1577: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1578: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1579: */
1580: }
1581:
1582: /*************************free matrix ************************/
1583: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1584: {
1585: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1586: free((FREE_ARG)(m+nrl-NR_END));
1587: }
1588:
1589: /******************* ma3x *******************************/
1590: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1591: {
1592: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1593: double ***m;
1594:
1595: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1596: if (!m) nrerror("allocation failure 1 in matrix()");
1597: m += NR_END;
1598: m -= nrl;
1599:
1600: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1601: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1602: m[nrl] += NR_END;
1603: m[nrl] -= ncl;
1604:
1605: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1606:
1607: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1608: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1609: m[nrl][ncl] += NR_END;
1610: m[nrl][ncl] -= nll;
1611: for (j=ncl+1; j<=nch; j++)
1612: m[nrl][j]=m[nrl][j-1]+nlay;
1613:
1614: for (i=nrl+1; i<=nrh; i++) {
1615: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1616: for (j=ncl+1; j<=nch; j++)
1617: m[i][j]=m[i][j-1]+nlay;
1618: }
1619: return m;
1620: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1621: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1622: */
1623: }
1624:
1625: /*************************free ma3x ************************/
1626: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1627: {
1628: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1629: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1630: free((FREE_ARG)(m+nrl-NR_END));
1631: }
1632:
1633: /*************** function subdirf ***********/
1634: char *subdirf(char fileres[])
1635: {
1636: /* Caution optionfilefiname is hidden */
1637: strcpy(tmpout,optionfilefiname);
1638: strcat(tmpout,"/"); /* Add to the right */
1639: strcat(tmpout,fileres);
1640: return tmpout;
1641: }
1642:
1643: /*************** function subdirf2 ***********/
1644: char *subdirf2(char fileres[], char *preop)
1645: {
1646:
1647: /* Caution optionfilefiname is hidden */
1648: strcpy(tmpout,optionfilefiname);
1649: strcat(tmpout,"/");
1650: strcat(tmpout,preop);
1651: strcat(tmpout,fileres);
1652: return tmpout;
1653: }
1654:
1655: /*************** function subdirf3 ***********/
1656: char *subdirf3(char fileres[], char *preop, char *preop2)
1657: {
1658:
1659: /* Caution optionfilefiname is hidden */
1660: strcpy(tmpout,optionfilefiname);
1661: strcat(tmpout,"/");
1662: strcat(tmpout,preop);
1663: strcat(tmpout,preop2);
1664: strcat(tmpout,fileres);
1665: return tmpout;
1666: }
1.213 brouard 1667:
1668: /*************** function subdirfext ***********/
1669: char *subdirfext(char fileres[], char *preop, char *postop)
1670: {
1671:
1672: strcpy(tmpout,preop);
1673: strcat(tmpout,fileres);
1674: strcat(tmpout,postop);
1675: return tmpout;
1676: }
1.126 brouard 1677:
1.213 brouard 1678: /*************** function subdirfext3 ***********/
1679: char *subdirfext3(char fileres[], char *preop, char *postop)
1680: {
1681:
1682: /* Caution optionfilefiname is hidden */
1683: strcpy(tmpout,optionfilefiname);
1684: strcat(tmpout,"/");
1685: strcat(tmpout,preop);
1686: strcat(tmpout,fileres);
1687: strcat(tmpout,postop);
1688: return tmpout;
1689: }
1690:
1.162 brouard 1691: char *asc_diff_time(long time_sec, char ascdiff[])
1692: {
1693: long sec_left, days, hours, minutes;
1694: days = (time_sec) / (60*60*24);
1695: sec_left = (time_sec) % (60*60*24);
1696: hours = (sec_left) / (60*60) ;
1697: sec_left = (sec_left) %(60*60);
1698: minutes = (sec_left) /60;
1699: sec_left = (sec_left) % (60);
1700: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1701: return ascdiff;
1702: }
1703:
1.126 brouard 1704: /***************** f1dim *************************/
1705: extern int ncom;
1706: extern double *pcom,*xicom;
1707: extern double (*nrfunc)(double []);
1708:
1709: double f1dim(double x)
1710: {
1711: int j;
1712: double f;
1713: double *xt;
1714:
1715: xt=vector(1,ncom);
1716: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1717: f=(*nrfunc)(xt);
1718: free_vector(xt,1,ncom);
1719: return f;
1720: }
1721:
1722: /*****************brent *************************/
1723: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1724: {
1725: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1726: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1727: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1728: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1729: * returned function value.
1730: */
1.126 brouard 1731: int iter;
1732: double a,b,d,etemp;
1.159 brouard 1733: double fu=0,fv,fw,fx;
1.164 brouard 1734: double ftemp=0.;
1.126 brouard 1735: double p,q,r,tol1,tol2,u,v,w,x,xm;
1736: double e=0.0;
1737:
1738: a=(ax < cx ? ax : cx);
1739: b=(ax > cx ? ax : cx);
1740: x=w=v=bx;
1741: fw=fv=fx=(*f)(x);
1742: for (iter=1;iter<=ITMAX;iter++) {
1743: xm=0.5*(a+b);
1744: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1745: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1746: printf(".");fflush(stdout);
1747: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1748: #ifdef DEBUGBRENT
1.126 brouard 1749: 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);
1750: 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);
1751: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1752: #endif
1753: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1754: *xmin=x;
1755: return fx;
1756: }
1757: ftemp=fu;
1758: if (fabs(e) > tol1) {
1759: r=(x-w)*(fx-fv);
1760: q=(x-v)*(fx-fw);
1761: p=(x-v)*q-(x-w)*r;
1762: q=2.0*(q-r);
1763: if (q > 0.0) p = -p;
1764: q=fabs(q);
1765: etemp=e;
1766: e=d;
1767: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1768: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1769: else {
1.224 brouard 1770: d=p/q;
1771: u=x+d;
1772: if (u-a < tol2 || b-u < tol2)
1773: d=SIGN(tol1,xm-x);
1.126 brouard 1774: }
1775: } else {
1776: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1777: }
1778: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1779: fu=(*f)(u);
1780: if (fu <= fx) {
1781: if (u >= x) a=x; else b=x;
1782: SHFT(v,w,x,u)
1.183 brouard 1783: SHFT(fv,fw,fx,fu)
1784: } else {
1785: if (u < x) a=u; else b=u;
1786: if (fu <= fw || w == x) {
1.224 brouard 1787: v=w;
1788: w=u;
1789: fv=fw;
1790: fw=fu;
1.183 brouard 1791: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1792: v=u;
1793: fv=fu;
1.183 brouard 1794: }
1795: }
1.126 brouard 1796: }
1797: nrerror("Too many iterations in brent");
1798: *xmin=x;
1799: return fx;
1800: }
1801:
1802: /****************** mnbrak ***********************/
1803:
1804: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1805: double (*func)(double))
1.183 brouard 1806: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1807: the downhill direction (defined by the function as evaluated at the initial points) and returns
1808: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1809: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1810: */
1.126 brouard 1811: double ulim,u,r,q, dum;
1812: double fu;
1.187 brouard 1813:
1814: double scale=10.;
1815: int iterscale=0;
1816:
1817: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1818: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1819:
1820:
1821: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1822: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1823: /* *bx = *ax - (*ax - *bx)/scale; */
1824: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1825: /* } */
1826:
1.126 brouard 1827: if (*fb > *fa) {
1828: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1829: SHFT(dum,*fb,*fa,dum)
1830: }
1.126 brouard 1831: *cx=(*bx)+GOLD*(*bx-*ax);
1832: *fc=(*func)(*cx);
1.183 brouard 1833: #ifdef DEBUG
1.224 brouard 1834: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1835: 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 1836: #endif
1.224 brouard 1837: 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 1838: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1839: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1840: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1841: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1842: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1843: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1844: fu=(*func)(u);
1.163 brouard 1845: #ifdef DEBUG
1846: /* f(x)=A(x-u)**2+f(u) */
1847: double A, fparabu;
1848: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1849: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1850: 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);
1851: 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 1852: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1853: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1854: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1855: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1856: #endif
1.184 brouard 1857: #ifdef MNBRAKORIGINAL
1.183 brouard 1858: #else
1.191 brouard 1859: /* if (fu > *fc) { */
1860: /* #ifdef DEBUG */
1861: /* printf("mnbrak4 fu > fc \n"); */
1862: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1863: /* #endif */
1864: /* /\* 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 *\\/ *\/ */
1865: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1866: /* dum=u; /\* Shifting c and u *\/ */
1867: /* u = *cx; */
1868: /* *cx = dum; */
1869: /* dum = fu; */
1870: /* fu = *fc; */
1871: /* *fc =dum; */
1872: /* } else { /\* end *\/ */
1873: /* #ifdef DEBUG */
1874: /* printf("mnbrak3 fu < fc \n"); */
1875: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1876: /* #endif */
1877: /* dum=u; /\* Shifting c and u *\/ */
1878: /* u = *cx; */
1879: /* *cx = dum; */
1880: /* dum = fu; */
1881: /* fu = *fc; */
1882: /* *fc =dum; */
1883: /* } */
1.224 brouard 1884: #ifdef DEBUGMNBRAK
1885: double A, fparabu;
1886: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1887: fparabu= *fa - A*(*ax-u)*(*ax-u);
1888: 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);
1889: 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 1890: #endif
1.191 brouard 1891: dum=u; /* Shifting c and u */
1892: u = *cx;
1893: *cx = dum;
1894: dum = fu;
1895: fu = *fc;
1896: *fc =dum;
1.183 brouard 1897: #endif
1.162 brouard 1898: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1899: #ifdef DEBUG
1.224 brouard 1900: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1901: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1902: #endif
1.126 brouard 1903: fu=(*func)(u);
1904: if (fu < *fc) {
1.183 brouard 1905: #ifdef DEBUG
1.224 brouard 1906: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1907: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1908: #endif
1909: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1910: SHFT(*fb,*fc,fu,(*func)(u))
1911: #ifdef DEBUG
1912: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1913: #endif
1914: }
1.162 brouard 1915: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1916: #ifdef DEBUG
1.224 brouard 1917: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1918: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1919: #endif
1.126 brouard 1920: u=ulim;
1921: fu=(*func)(u);
1.183 brouard 1922: } else { /* u could be left to b (if r > q parabola has a maximum) */
1923: #ifdef DEBUG
1.224 brouard 1924: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1925: 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 1926: #endif
1.126 brouard 1927: u=(*cx)+GOLD*(*cx-*bx);
1928: fu=(*func)(u);
1.224 brouard 1929: #ifdef DEBUG
1930: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1931: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1932: #endif
1.183 brouard 1933: } /* end tests */
1.126 brouard 1934: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1935: SHFT(*fa,*fb,*fc,fu)
1936: #ifdef DEBUG
1.224 brouard 1937: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1938: 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 1939: #endif
1940: } /* 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 1941: }
1942:
1943: /*************** linmin ************************/
1.162 brouard 1944: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1945: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1946: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1947: the value of func at the returned location p . This is actually all accomplished by calling the
1948: routines mnbrak and brent .*/
1.126 brouard 1949: int ncom;
1950: double *pcom,*xicom;
1951: double (*nrfunc)(double []);
1952:
1.224 brouard 1953: #ifdef LINMINORIGINAL
1.126 brouard 1954: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1955: #else
1956: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1957: #endif
1.126 brouard 1958: {
1959: double brent(double ax, double bx, double cx,
1960: double (*f)(double), double tol, double *xmin);
1961: double f1dim(double x);
1962: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
1963: double *fc, double (*func)(double));
1964: int j;
1965: double xx,xmin,bx,ax;
1966: double fx,fb,fa;
1.187 brouard 1967:
1.203 brouard 1968: #ifdef LINMINORIGINAL
1969: #else
1970: double scale=10., axs, xxs; /* Scale added for infinity */
1971: #endif
1972:
1.126 brouard 1973: ncom=n;
1974: pcom=vector(1,n);
1975: xicom=vector(1,n);
1976: nrfunc=func;
1977: for (j=1;j<=n;j++) {
1978: pcom[j]=p[j];
1.202 brouard 1979: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 1980: }
1.187 brouard 1981:
1.203 brouard 1982: #ifdef LINMINORIGINAL
1983: xx=1.;
1984: #else
1985: axs=0.0;
1986: xxs=1.;
1987: do{
1988: xx= xxs;
1989: #endif
1.187 brouard 1990: ax=0.;
1991: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
1992: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
1993: /* 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)) */
1994: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
1995: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
1996: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
1997: /* 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 1998: #ifdef LINMINORIGINAL
1999: #else
2000: if (fx != fx){
1.224 brouard 2001: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
2002: printf("|");
2003: fprintf(ficlog,"|");
1.203 brouard 2004: #ifdef DEBUGLINMIN
1.224 brouard 2005: 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 2006: #endif
2007: }
1.224 brouard 2008: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 2009: #endif
2010:
1.191 brouard 2011: #ifdef DEBUGLINMIN
2012: 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 2013: 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 2014: #endif
1.224 brouard 2015: #ifdef LINMINORIGINAL
2016: #else
2017: if(fb == fx){ /* Flat function in the direction */
2018: xmin=xx;
2019: *flat=1;
2020: }else{
2021: *flat=0;
2022: #endif
2023: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 2024: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
2025: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
2026: /* fmin = f(p[j] + xmin * xi[j]) */
2027: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
2028: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 2029: #ifdef DEBUG
1.224 brouard 2030: 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);
2031: 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);
2032: #endif
2033: #ifdef LINMINORIGINAL
2034: #else
2035: }
1.126 brouard 2036: #endif
1.191 brouard 2037: #ifdef DEBUGLINMIN
2038: printf("linmin end ");
1.202 brouard 2039: fprintf(ficlog,"linmin end ");
1.191 brouard 2040: #endif
1.126 brouard 2041: for (j=1;j<=n;j++) {
1.203 brouard 2042: #ifdef LINMINORIGINAL
2043: xi[j] *= xmin;
2044: #else
2045: #ifdef DEBUGLINMIN
2046: if(xxs <1.0)
2047: printf(" before xi[%d]=%12.8f", j,xi[j]);
2048: #endif
2049: 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) */
2050: #ifdef DEBUGLINMIN
2051: if(xxs <1.0)
2052: 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 );
2053: #endif
2054: #endif
1.187 brouard 2055: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2056: }
1.191 brouard 2057: #ifdef DEBUGLINMIN
1.203 brouard 2058: printf("\n");
1.191 brouard 2059: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2060: 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 2061: for (j=1;j<=n;j++) {
1.202 brouard 2062: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2063: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2064: if(j % ncovmodel == 0){
1.191 brouard 2065: printf("\n");
1.202 brouard 2066: fprintf(ficlog,"\n");
2067: }
1.191 brouard 2068: }
1.203 brouard 2069: #else
1.191 brouard 2070: #endif
1.126 brouard 2071: free_vector(xicom,1,n);
2072: free_vector(pcom,1,n);
2073: }
2074:
2075:
2076: /*************** powell ************************/
1.162 brouard 2077: /*
2078: Minimization of a function func of n variables. Input consists of an initial starting point
2079: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2080: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2081: such that failure to decrease by more than this amount on one iteration signals doneness. On
2082: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2083: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2084: */
1.224 brouard 2085: #ifdef LINMINORIGINAL
2086: #else
2087: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2088: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2089: #endif
1.126 brouard 2090: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2091: double (*func)(double []))
2092: {
1.224 brouard 2093: #ifdef LINMINORIGINAL
2094: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2095: double (*func)(double []));
1.224 brouard 2096: #else
1.241 brouard 2097: void linmin(double p[], double xi[], int n, double *fret,
2098: double (*func)(double []),int *flat);
1.224 brouard 2099: #endif
1.239 brouard 2100: int i,ibig,j,jk,k;
1.126 brouard 2101: double del,t,*pt,*ptt,*xit;
1.181 brouard 2102: double directest;
1.126 brouard 2103: double fp,fptt;
2104: double *xits;
2105: int niterf, itmp;
1.224 brouard 2106: #ifdef LINMINORIGINAL
2107: #else
2108:
2109: flatdir=ivector(1,n);
2110: for (j=1;j<=n;j++) flatdir[j]=0;
2111: #endif
1.126 brouard 2112:
2113: pt=vector(1,n);
2114: ptt=vector(1,n);
2115: xit=vector(1,n);
2116: xits=vector(1,n);
2117: *fret=(*func)(p);
2118: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2119: rcurr_time = time(NULL);
1.126 brouard 2120: for (*iter=1;;++(*iter)) {
1.187 brouard 2121: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2122: ibig=0;
2123: del=0.0;
1.157 brouard 2124: rlast_time=rcurr_time;
2125: /* (void) gettimeofday(&curr_time,&tzp); */
2126: rcurr_time = time(NULL);
2127: curr_time = *localtime(&rcurr_time);
2128: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2129: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2130: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2131: for (i=1;i<=n;i++) {
1.126 brouard 2132: fprintf(ficrespow," %.12lf", p[i]);
2133: }
1.239 brouard 2134: fprintf(ficrespow,"\n");fflush(ficrespow);
2135: printf("\n#model= 1 + age ");
2136: fprintf(ficlog,"\n#model= 1 + age ");
2137: if(nagesqr==1){
1.241 brouard 2138: printf(" + age*age ");
2139: fprintf(ficlog," + age*age ");
1.239 brouard 2140: }
2141: for(j=1;j <=ncovmodel-2;j++){
2142: if(Typevar[j]==0) {
2143: printf(" + V%d ",Tvar[j]);
2144: fprintf(ficlog," + V%d ",Tvar[j]);
2145: }else if(Typevar[j]==1) {
2146: printf(" + V%d*age ",Tvar[j]);
2147: fprintf(ficlog," + V%d*age ",Tvar[j]);
2148: }else if(Typevar[j]==2) {
2149: printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2150: fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2151: }
2152: }
1.126 brouard 2153: printf("\n");
1.239 brouard 2154: /* printf("12 47.0114589 0.0154322 33.2424412 0.3279905 2.3731903 */
2155: /* 13 -21.5392400 0.1118147 1.2680506 1.2973408 -1.0663662 */
1.126 brouard 2156: fprintf(ficlog,"\n");
1.239 brouard 2157: for(i=1,jk=1; i <=nlstate; i++){
2158: for(k=1; k <=(nlstate+ndeath); k++){
2159: if (k != i) {
2160: printf("%d%d ",i,k);
2161: fprintf(ficlog,"%d%d ",i,k);
2162: for(j=1; j <=ncovmodel; j++){
2163: printf("%12.7f ",p[jk]);
2164: fprintf(ficlog,"%12.7f ",p[jk]);
2165: jk++;
2166: }
2167: printf("\n");
2168: fprintf(ficlog,"\n");
2169: }
2170: }
2171: }
1.241 brouard 2172: if(*iter <=3 && *iter >1){
1.157 brouard 2173: tml = *localtime(&rcurr_time);
2174: strcpy(strcurr,asctime(&tml));
2175: rforecast_time=rcurr_time;
1.126 brouard 2176: itmp = strlen(strcurr);
2177: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.241 brouard 2178: strcurr[itmp-1]='\0';
1.162 brouard 2179: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2180: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2181: for(niterf=10;niterf<=30;niterf+=10){
1.241 brouard 2182: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2183: forecast_time = *localtime(&rforecast_time);
2184: strcpy(strfor,asctime(&forecast_time));
2185: itmp = strlen(strfor);
2186: if(strfor[itmp-1]=='\n')
2187: strfor[itmp-1]='\0';
2188: 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);
2189: 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 2190: }
2191: }
1.187 brouard 2192: for (i=1;i<=n;i++) { /* For each direction i */
2193: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2194: fptt=(*fret);
2195: #ifdef DEBUG
1.203 brouard 2196: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2197: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2198: #endif
1.203 brouard 2199: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2200: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2201: #ifdef LINMINORIGINAL
1.188 brouard 2202: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2203: #else
2204: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2205: flatdir[i]=flat; /* Function is vanishing in that direction i */
2206: #endif
2207: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2208: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2209: /* because that direction will be replaced unless the gain del is small */
2210: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2211: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2212: /* with the new direction. */
2213: del=fabs(fptt-(*fret));
2214: ibig=i;
1.126 brouard 2215: }
2216: #ifdef DEBUG
2217: printf("%d %.12e",i,(*fret));
2218: fprintf(ficlog,"%d %.12e",i,(*fret));
2219: for (j=1;j<=n;j++) {
1.224 brouard 2220: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2221: printf(" x(%d)=%.12e",j,xit[j]);
2222: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2223: }
2224: for(j=1;j<=n;j++) {
1.225 brouard 2225: printf(" p(%d)=%.12e",j,p[j]);
2226: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2227: }
2228: printf("\n");
2229: fprintf(ficlog,"\n");
2230: #endif
1.187 brouard 2231: } /* end loop on each direction i */
2232: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2233: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2234: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2235: for(j=1;j<=n;j++) {
1.225 brouard 2236: if(flatdir[j] >0){
2237: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2238: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2239: }
2240: /* printf("\n"); */
2241: /* fprintf(ficlog,"\n"); */
2242: }
1.243 brouard 2243: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */
2244: if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */
1.188 brouard 2245: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2246: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2247: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2248: /* decreased of more than 3.84 */
2249: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2250: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2251: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2252:
1.188 brouard 2253: /* Starting the program with initial values given by a former maximization will simply change */
2254: /* the scales of the directions and the directions, because the are reset to canonical directions */
2255: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2256: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2257: #ifdef DEBUG
2258: int k[2],l;
2259: k[0]=1;
2260: k[1]=-1;
2261: printf("Max: %.12e",(*func)(p));
2262: fprintf(ficlog,"Max: %.12e",(*func)(p));
2263: for (j=1;j<=n;j++) {
2264: printf(" %.12e",p[j]);
2265: fprintf(ficlog," %.12e",p[j]);
2266: }
2267: printf("\n");
2268: fprintf(ficlog,"\n");
2269: for(l=0;l<=1;l++) {
2270: for (j=1;j<=n;j++) {
2271: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2272: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2273: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2274: }
2275: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2276: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2277: }
2278: #endif
2279:
1.224 brouard 2280: #ifdef LINMINORIGINAL
2281: #else
2282: free_ivector(flatdir,1,n);
2283: #endif
1.126 brouard 2284: free_vector(xit,1,n);
2285: free_vector(xits,1,n);
2286: free_vector(ptt,1,n);
2287: free_vector(pt,1,n);
2288: return;
1.192 brouard 2289: } /* enough precision */
1.240 brouard 2290: if (*iter == ITMAX*n) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2291: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2292: ptt[j]=2.0*p[j]-pt[j];
2293: xit[j]=p[j]-pt[j];
2294: pt[j]=p[j];
2295: }
1.181 brouard 2296: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2297: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2298: if (*iter <=4) {
1.225 brouard 2299: #else
2300: #endif
1.224 brouard 2301: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2302: #else
1.161 brouard 2303: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2304: #endif
1.162 brouard 2305: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2306: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2307: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2308: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2309: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2310: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2311: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2312: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2313: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2314: /* Even if f3 <f1, directest can be negative and t >0 */
2315: /* mu² and del² are equal when f3=f1 */
2316: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2317: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2318: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2319: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2320: #ifdef NRCORIGINAL
2321: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2322: #else
2323: 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 2324: t= t- del*SQR(fp-fptt);
1.183 brouard 2325: #endif
1.202 brouard 2326: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2327: #ifdef DEBUG
1.181 brouard 2328: 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);
2329: 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 2330: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2331: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2332: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2333: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2334: 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);
2335: 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);
2336: #endif
1.183 brouard 2337: #ifdef POWELLORIGINAL
2338: if (t < 0.0) { /* Then we use it for new direction */
2339: #else
1.182 brouard 2340: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2341: 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 2342: 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 2343: 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 2344: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2345: }
1.181 brouard 2346: if (directest < 0.0) { /* Then we use it for new direction */
2347: #endif
1.191 brouard 2348: #ifdef DEBUGLINMIN
1.234 brouard 2349: printf("Before linmin in direction P%d-P0\n",n);
2350: for (j=1;j<=n;j++) {
2351: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2352: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2353: if(j % ncovmodel == 0){
2354: printf("\n");
2355: fprintf(ficlog,"\n");
2356: }
2357: }
1.224 brouard 2358: #endif
2359: #ifdef LINMINORIGINAL
1.234 brouard 2360: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2361: #else
1.234 brouard 2362: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2363: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2364: #endif
1.234 brouard 2365:
1.191 brouard 2366: #ifdef DEBUGLINMIN
1.234 brouard 2367: for (j=1;j<=n;j++) {
2368: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2369: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2370: if(j % ncovmodel == 0){
2371: printf("\n");
2372: fprintf(ficlog,"\n");
2373: }
2374: }
1.224 brouard 2375: #endif
1.234 brouard 2376: for (j=1;j<=n;j++) {
2377: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2378: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2379: }
1.224 brouard 2380: #ifdef LINMINORIGINAL
2381: #else
1.234 brouard 2382: for (j=1, flatd=0;j<=n;j++) {
2383: if(flatdir[j]>0)
2384: flatd++;
2385: }
2386: if(flatd >0){
2387: printf("%d flat directions\n",flatd);
2388: fprintf(ficlog,"%d flat directions\n",flatd);
2389: for (j=1;j<=n;j++) {
2390: if(flatdir[j]>0){
2391: printf("%d ",j);
2392: fprintf(ficlog,"%d ",j);
2393: }
2394: }
2395: printf("\n");
2396: fprintf(ficlog,"\n");
2397: }
1.191 brouard 2398: #endif
1.234 brouard 2399: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2400: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2401:
1.126 brouard 2402: #ifdef DEBUG
1.234 brouard 2403: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2404: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2405: for(j=1;j<=n;j++){
2406: printf(" %lf",xit[j]);
2407: fprintf(ficlog," %lf",xit[j]);
2408: }
2409: printf("\n");
2410: fprintf(ficlog,"\n");
1.126 brouard 2411: #endif
1.192 brouard 2412: } /* end of t or directest negative */
1.224 brouard 2413: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2414: #else
1.234 brouard 2415: } /* end if (fptt < fp) */
1.192 brouard 2416: #endif
1.225 brouard 2417: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2418: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2419: #else
1.224 brouard 2420: #endif
1.234 brouard 2421: } /* loop iteration */
1.126 brouard 2422: }
1.234 brouard 2423:
1.126 brouard 2424: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2425:
1.235 brouard 2426: 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 2427: {
1.235 brouard 2428: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2429: (and selected quantitative values in nres)
2430: by left multiplying the unit
1.234 brouard 2431: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2432: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2433: /* Wx is row vector: population in state 1, population in state 2, population dead */
2434: /* or prevalence in state 1, prevalence in state 2, 0 */
2435: /* newm is the matrix after multiplications, its rows are identical at a factor */
2436: /* Initial matrix pimij */
2437: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2438: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2439: /* 0, 0 , 1} */
2440: /*
2441: * and after some iteration: */
2442: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2443: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2444: /* 0, 0 , 1} */
2445: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2446: /* {0.51571254859325999, 0.4842874514067399, */
2447: /* 0.51326036147820708, 0.48673963852179264} */
2448: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2449:
1.126 brouard 2450: int i, ii,j,k;
1.209 brouard 2451: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2452: /* double **matprod2(); */ /* test */
1.218 brouard 2453: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2454: double **newm;
1.209 brouard 2455: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2456: int ncvloop=0;
1.169 brouard 2457:
1.209 brouard 2458: min=vector(1,nlstate);
2459: max=vector(1,nlstate);
2460: meandiff=vector(1,nlstate);
2461:
1.218 brouard 2462: /* Starting with matrix unity */
1.126 brouard 2463: for (ii=1;ii<=nlstate+ndeath;ii++)
2464: for (j=1;j<=nlstate+ndeath;j++){
2465: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2466: }
1.169 brouard 2467:
2468: cov[1]=1.;
2469:
2470: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2471: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2472: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2473: ncvloop++;
1.126 brouard 2474: newm=savm;
2475: /* Covariates have to be included here again */
1.138 brouard 2476: cov[2]=agefin;
1.187 brouard 2477: if(nagesqr==1)
2478: cov[3]= agefin*agefin;;
1.234 brouard 2479: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2480: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2481: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2482: /* 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 2483: }
2484: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2485: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2486: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2487: /* 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 2488: }
1.237 brouard 2489: for (k=1; k<=cptcovage;k++){ /* For product with age */
1.234 brouard 2490: if(Dummy[Tvar[Tage[k]]]){
2491: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2492: } else{
1.235 brouard 2493: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2494: }
1.235 brouard 2495: /* 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 2496: }
1.237 brouard 2497: for (k=1; k<=cptcovprod;k++){ /* For product without age */
1.235 brouard 2498: /* 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 2499: if(Dummy[Tvard[k][1]==0]){
2500: if(Dummy[Tvard[k][2]==0]){
2501: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2502: }else{
2503: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2504: }
2505: }else{
2506: if(Dummy[Tvard[k][2]==0]){
2507: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2508: }else{
2509: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2510: }
2511: }
1.234 brouard 2512: }
1.138 brouard 2513: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2514: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2515: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2516: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2517: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2518: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2519: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2520:
1.126 brouard 2521: savm=oldm;
2522: oldm=newm;
1.209 brouard 2523:
2524: for(j=1; j<=nlstate; j++){
2525: max[j]=0.;
2526: min[j]=1.;
2527: }
2528: for(i=1;i<=nlstate;i++){
2529: sumnew=0;
2530: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2531: for(j=1; j<=nlstate; j++){
2532: prlim[i][j]= newm[i][j]/(1-sumnew);
2533: max[j]=FMAX(max[j],prlim[i][j]);
2534: min[j]=FMIN(min[j],prlim[i][j]);
2535: }
2536: }
2537:
1.126 brouard 2538: maxmax=0.;
1.209 brouard 2539: for(j=1; j<=nlstate; j++){
2540: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2541: maxmax=FMAX(maxmax,meandiff[j]);
2542: /* 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 2543: } /* j loop */
1.203 brouard 2544: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2545: /* 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 2546: if(maxmax < ftolpl){
1.209 brouard 2547: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2548: free_vector(min,1,nlstate);
2549: free_vector(max,1,nlstate);
2550: free_vector(meandiff,1,nlstate);
1.126 brouard 2551: return prlim;
2552: }
1.169 brouard 2553: } /* age loop */
1.208 brouard 2554: /* After some age loop it doesn't converge */
1.209 brouard 2555: 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 2556: 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 2557: /* 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); */
2558: free_vector(min,1,nlstate);
2559: free_vector(max,1,nlstate);
2560: free_vector(meandiff,1,nlstate);
1.208 brouard 2561:
1.169 brouard 2562: return prlim; /* should not reach here */
1.126 brouard 2563: }
2564:
1.217 brouard 2565:
2566: /**** Back Prevalence limit (stable or period prevalence) ****************/
2567:
1.218 brouard 2568: /* 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) */
2569: /* 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 2570: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij, int nres)
1.217 brouard 2571: {
1.218 brouard 2572: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2573: matrix by transitions matrix until convergence is reached with precision ftolpl */
2574: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2575: /* Wx is row vector: population in state 1, population in state 2, population dead */
2576: /* or prevalence in state 1, prevalence in state 2, 0 */
2577: /* newm is the matrix after multiplications, its rows are identical at a factor */
2578: /* Initial matrix pimij */
2579: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2580: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2581: /* 0, 0 , 1} */
2582: /*
2583: * and after some iteration: */
2584: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2585: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2586: /* 0, 0 , 1} */
2587: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2588: /* {0.51571254859325999, 0.4842874514067399, */
2589: /* 0.51326036147820708, 0.48673963852179264} */
2590: /* If we start from prlim again, prlim tends to a constant matrix */
2591:
2592: int i, ii,j,k;
1.247 brouard 2593: int first=0;
1.217 brouard 2594: double *min, *max, *meandiff, maxmax,sumnew=0.;
2595: /* double **matprod2(); */ /* test */
2596: double **out, cov[NCOVMAX+1], **bmij();
2597: double **newm;
1.218 brouard 2598: double **dnewm, **doldm, **dsavm; /* for use */
2599: double **oldm, **savm; /* for use */
2600:
1.217 brouard 2601: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2602: int ncvloop=0;
2603:
2604: min=vector(1,nlstate);
2605: max=vector(1,nlstate);
2606: meandiff=vector(1,nlstate);
2607:
1.218 brouard 2608: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2609: oldm=oldms; savm=savms;
2610:
2611: /* Starting with matrix unity */
2612: for (ii=1;ii<=nlstate+ndeath;ii++)
2613: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2614: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2615: }
2616:
2617: cov[1]=1.;
2618:
2619: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2620: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2621: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2622: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2623: ncvloop++;
1.218 brouard 2624: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2625: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2626: /* Covariates have to be included here again */
2627: cov[2]=agefin;
2628: if(nagesqr==1)
2629: cov[3]= agefin*agefin;;
1.242 brouard 2630: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2631: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2632: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
2633: /* 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)); */
2634: }
2635: /* for (k=1; k<=cptcovn;k++) { */
2636: /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */
2637: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2638: /* /\* 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])]); *\/ */
2639: /* } */
2640: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2641: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2642: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2643: /* 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]); */
2644: }
2645: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2]; */
2646: /* for (k=1; k<=cptcovprod;k++) /\* Useless *\/ */
2647: /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */
2648: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */
2649: for (k=1; k<=cptcovage;k++){ /* For product with age */
2650: if(Dummy[Tvar[Tage[k]]]){
2651: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2652: } else{
2653: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2654: }
2655: /* 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]); */
2656: }
2657: for (k=1; k<=cptcovprod;k++){ /* For product without age */
2658: /* 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]); */
2659: if(Dummy[Tvard[k][1]==0]){
2660: if(Dummy[Tvard[k][2]==0]){
2661: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2662: }else{
2663: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2664: }
2665: }else{
2666: if(Dummy[Tvard[k][2]==0]){
2667: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2668: }else{
2669: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2670: }
2671: }
1.217 brouard 2672: }
2673:
2674: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2675: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2676: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2677: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2678: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2679: /* ij should be linked to the correct index of cov */
2680: /* age and covariate values ij are in 'cov', but we need to pass
2681: * ij for the observed prevalence at age and status and covariate
2682: * number: prevacurrent[(int)agefin][ii][ij]
2683: */
2684: /* 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 *\/ */
2685: /* 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 *\/ */
2686: 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 2687: savm=oldm;
2688: oldm=newm;
2689: for(j=1; j<=nlstate; j++){
2690: max[j]=0.;
2691: min[j]=1.;
2692: }
2693: for(j=1; j<=nlstate; j++){
2694: for(i=1;i<=nlstate;i++){
1.234 brouard 2695: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2696: bprlim[i][j]= newm[i][j];
2697: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2698: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2699: }
2700: }
1.218 brouard 2701:
1.217 brouard 2702: maxmax=0.;
2703: for(i=1; i<=nlstate; i++){
2704: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2705: maxmax=FMAX(maxmax,meandiff[i]);
2706: /* 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); */
2707: } /* j loop */
2708: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2709: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2710: if(maxmax < ftolpl){
1.220 brouard 2711: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2712: free_vector(min,1,nlstate);
2713: free_vector(max,1,nlstate);
2714: free_vector(meandiff,1,nlstate);
2715: return bprlim;
2716: }
2717: } /* age loop */
2718: /* After some age loop it doesn't converge */
1.247 brouard 2719: if(first){
2720: first=1;
2721: 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\
2722: 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);
2723: }
2724: 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 2725: 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);
2726: /* 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); */
2727: free_vector(min,1,nlstate);
2728: free_vector(max,1,nlstate);
2729: free_vector(meandiff,1,nlstate);
2730:
2731: return bprlim; /* should not reach here */
2732: }
2733:
1.126 brouard 2734: /*************** transition probabilities ***************/
2735:
2736: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2737: {
1.138 brouard 2738: /* According to parameters values stored in x and the covariate's values stored in cov,
2739: computes the probability to be observed in state j being in state i by appying the
2740: model to the ncovmodel covariates (including constant and age).
2741: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2742: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2743: ncth covariate in the global vector x is given by the formula:
2744: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2745: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2746: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2747: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2748: Outputs ps[i][j] the probability to be observed in j being in j according to
2749: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2750: */
2751: double s1, lnpijopii;
1.126 brouard 2752: /*double t34;*/
1.164 brouard 2753: int i,j, nc, ii, jj;
1.126 brouard 2754:
1.223 brouard 2755: for(i=1; i<= nlstate; i++){
2756: for(j=1; j<i;j++){
2757: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2758: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2759: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2760: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2761: }
2762: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2763: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2764: }
2765: for(j=i+1; j<=nlstate+ndeath;j++){
2766: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2767: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2768: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2769: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2770: }
2771: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2772: }
2773: }
1.218 brouard 2774:
1.223 brouard 2775: for(i=1; i<= nlstate; i++){
2776: s1=0;
2777: for(j=1; j<i; j++){
2778: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2779: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2780: }
2781: for(j=i+1; j<=nlstate+ndeath; j++){
2782: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2783: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2784: }
2785: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2786: ps[i][i]=1./(s1+1.);
2787: /* Computing other pijs */
2788: for(j=1; j<i; j++)
2789: ps[i][j]= exp(ps[i][j])*ps[i][i];
2790: for(j=i+1; j<=nlstate+ndeath; j++)
2791: ps[i][j]= exp(ps[i][j])*ps[i][i];
2792: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2793: } /* end i */
1.218 brouard 2794:
1.223 brouard 2795: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2796: for(jj=1; jj<= nlstate+ndeath; jj++){
2797: ps[ii][jj]=0;
2798: ps[ii][ii]=1;
2799: }
2800: }
1.218 brouard 2801:
2802:
1.223 brouard 2803: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2804: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2805: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2806: /* } */
2807: /* printf("\n "); */
2808: /* } */
2809: /* printf("\n ");printf("%lf ",cov[2]);*/
2810: /*
2811: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2812: goto end;*/
1.223 brouard 2813: return ps;
1.126 brouard 2814: }
2815:
1.218 brouard 2816: /*************** backward transition probabilities ***************/
2817:
2818: /* 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 ) */
2819: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2820: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2821: {
1.222 brouard 2822: /* Computes the backward probability at age agefin and covariate ij
2823: * and returns in **ps as well as **bmij.
2824: */
1.218 brouard 2825: int i, ii, j,k;
1.222 brouard 2826:
2827: double **out, **pmij();
2828: double sumnew=0.;
1.218 brouard 2829: double agefin;
1.222 brouard 2830:
2831: double **dnewm, **dsavm, **doldm;
2832: double **bbmij;
2833:
1.218 brouard 2834: doldm=ddoldms; /* global pointers */
1.222 brouard 2835: dnewm=ddnewms;
2836: dsavm=ddsavms;
2837:
2838: agefin=cov[2];
2839: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2840: the observed prevalence (with this covariate ij) */
2841: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2842: /* We do have the matrix Px in savm and we need pij */
2843: for (j=1;j<=nlstate+ndeath;j++){
2844: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2845: for (ii=1;ii<=nlstate;ii++){
2846: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2847: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2848: for (ii=1;ii<=nlstate+ndeath;ii++){
2849: if(sumnew >= 1.e-10){
2850: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2851: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2852: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2853: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2854: /* }else */
2855: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2856: }else{
1.242 brouard 2857: ;
2858: /* 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 2859: }
2860: } /*End ii */
2861: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2862: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2863: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2864: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2865: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2866: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2867: /* left Product of this matrix by diag matrix of prevalences (savm) */
2868: for (j=1;j<=nlstate+ndeath;j++){
2869: for (ii=1;ii<=nlstate+ndeath;ii++){
2870: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2871: }
2872: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2873: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2874: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2875: /* end bmij */
2876: return ps;
1.218 brouard 2877: }
1.217 brouard 2878: /*************** transition probabilities ***************/
2879:
1.218 brouard 2880: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2881: {
2882: /* According to parameters values stored in x and the covariate's values stored in cov,
2883: computes the probability to be observed in state j being in state i by appying the
2884: model to the ncovmodel covariates (including constant and age).
2885: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2886: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2887: ncth covariate in the global vector x is given by the formula:
2888: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2889: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2890: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2891: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2892: Outputs ps[i][j] the probability to be observed in j being in j according to
2893: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2894: */
2895: double s1, lnpijopii;
2896: /*double t34;*/
2897: int i,j, nc, ii, jj;
2898:
1.234 brouard 2899: for(i=1; i<= nlstate; i++){
2900: for(j=1; j<i;j++){
2901: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2902: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2903: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2904: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2905: }
2906: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2907: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2908: }
2909: for(j=i+1; j<=nlstate+ndeath;j++){
2910: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2911: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2912: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2913: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2914: }
2915: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2916: }
2917: }
2918:
2919: for(i=1; i<= nlstate; i++){
2920: s1=0;
2921: for(j=1; j<i; j++){
2922: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2923: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2924: }
2925: for(j=i+1; j<=nlstate+ndeath; j++){
2926: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2927: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2928: }
2929: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2930: ps[i][i]=1./(s1+1.);
2931: /* Computing other pijs */
2932: for(j=1; j<i; j++)
2933: ps[i][j]= exp(ps[i][j])*ps[i][i];
2934: for(j=i+1; j<=nlstate+ndeath; j++)
2935: ps[i][j]= exp(ps[i][j])*ps[i][i];
2936: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2937: } /* end i */
2938:
2939: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2940: for(jj=1; jj<= nlstate+ndeath; jj++){
2941: ps[ii][jj]=0;
2942: ps[ii][ii]=1;
2943: }
2944: }
2945: /* Added for backcast */ /* Transposed matrix too */
2946: for(jj=1; jj<= nlstate+ndeath; jj++){
2947: s1=0.;
2948: for(ii=1; ii<= nlstate+ndeath; ii++){
2949: s1+=ps[ii][jj];
2950: }
2951: for(ii=1; ii<= nlstate; ii++){
2952: ps[ii][jj]=ps[ii][jj]/s1;
2953: }
2954: }
2955: /* Transposition */
2956: for(jj=1; jj<= nlstate+ndeath; jj++){
2957: for(ii=jj; ii<= nlstate+ndeath; ii++){
2958: s1=ps[ii][jj];
2959: ps[ii][jj]=ps[jj][ii];
2960: ps[jj][ii]=s1;
2961: }
2962: }
2963: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2964: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2965: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2966: /* } */
2967: /* printf("\n "); */
2968: /* } */
2969: /* printf("\n ");printf("%lf ",cov[2]);*/
2970: /*
2971: for(i=1; i<= npar; i++) printf("%f ",x[i]);
2972: goto end;*/
2973: return ps;
1.217 brouard 2974: }
2975:
2976:
1.126 brouard 2977: /**************** Product of 2 matrices ******************/
2978:
1.145 brouard 2979: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 2980: {
2981: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
2982: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
2983: /* in, b, out are matrice of pointers which should have been initialized
2984: before: only the contents of out is modified. The function returns
2985: a pointer to pointers identical to out */
1.145 brouard 2986: int i, j, k;
1.126 brouard 2987: for(i=nrl; i<= nrh; i++)
1.145 brouard 2988: for(k=ncolol; k<=ncoloh; k++){
2989: out[i][k]=0.;
2990: for(j=ncl; j<=nch; j++)
2991: out[i][k] +=in[i][j]*b[j][k];
2992: }
1.126 brouard 2993: return out;
2994: }
2995:
2996:
2997: /************* Higher Matrix Product ***************/
2998:
1.235 brouard 2999: 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 3000: {
1.218 brouard 3001: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 3002: 'nhstepm*hstepm*stepm' months (i.e. until
3003: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3004: nhstepm*hstepm matrices.
3005: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3006: (typically every 2 years instead of every month which is too big
3007: for the memory).
3008: Model is determined by parameters x and covariates have to be
3009: included manually here.
3010:
3011: */
3012:
3013: int i, j, d, h, k;
1.131 brouard 3014: double **out, cov[NCOVMAX+1];
1.126 brouard 3015: double **newm;
1.187 brouard 3016: double agexact;
1.214 brouard 3017: double agebegin, ageend;
1.126 brouard 3018:
3019: /* Hstepm could be zero and should return the unit matrix */
3020: for (i=1;i<=nlstate+ndeath;i++)
3021: for (j=1;j<=nlstate+ndeath;j++){
3022: oldm[i][j]=(i==j ? 1.0 : 0.0);
3023: po[i][j][0]=(i==j ? 1.0 : 0.0);
3024: }
3025: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3026: for(h=1; h <=nhstepm; h++){
3027: for(d=1; d <=hstepm; d++){
3028: newm=savm;
3029: /* Covariates have to be included here again */
3030: cov[1]=1.;
1.214 brouard 3031: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 3032: cov[2]=agexact;
3033: if(nagesqr==1)
1.227 brouard 3034: cov[3]= agexact*agexact;
1.235 brouard 3035: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
3036: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
3037: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
3038: /* 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)); */
3039: }
3040: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
3041: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
3042: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
3043: /* 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]); */
3044: }
3045: for (k=1; k<=cptcovage;k++){
3046: if(Dummy[Tvar[Tage[k]]]){
3047: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3048: } else{
3049: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
3050: }
3051: /* 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]); */
3052: }
3053: for (k=1; k<=cptcovprod;k++){ /* */
3054: /* 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]); */
3055: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
3056: }
3057: /* for (k=1; k<=cptcovn;k++) */
3058: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
3059: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
3060: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
3061: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
3062: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 3063:
3064:
1.126 brouard 3065: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3066: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3067: /* right multiplication of oldm by the current matrix */
1.126 brouard 3068: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
3069: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 3070: /* if((int)age == 70){ */
3071: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3072: /* for(i=1; i<=nlstate+ndeath; i++) { */
3073: /* printf("%d pmmij ",i); */
3074: /* for(j=1;j<=nlstate+ndeath;j++) { */
3075: /* printf("%f ",pmmij[i][j]); */
3076: /* } */
3077: /* printf(" oldm "); */
3078: /* for(j=1;j<=nlstate+ndeath;j++) { */
3079: /* printf("%f ",oldm[i][j]); */
3080: /* } */
3081: /* printf("\n"); */
3082: /* } */
3083: /* } */
1.126 brouard 3084: savm=oldm;
3085: oldm=newm;
3086: }
3087: for(i=1; i<=nlstate+ndeath; i++)
3088: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 3089: po[i][j][h]=newm[i][j];
3090: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 3091: }
1.128 brouard 3092: /*printf("h=%d ",h);*/
1.126 brouard 3093: } /* end h */
1.218 brouard 3094: /* printf("\n H=%d \n",h); */
1.126 brouard 3095: return po;
3096: }
3097:
1.217 brouard 3098: /************* Higher Back Matrix Product ***************/
1.218 brouard 3099: /* 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 3100: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 3101: {
1.218 brouard 3102: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 3103: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 3104: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3105: nhstepm*hstepm matrices.
3106: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3107: (typically every 2 years instead of every month which is too big
1.217 brouard 3108: for the memory).
1.218 brouard 3109: Model is determined by parameters x and covariates have to be
3110: included manually here.
1.217 brouard 3111:
1.222 brouard 3112: */
1.217 brouard 3113:
3114: int i, j, d, h, k;
3115: double **out, cov[NCOVMAX+1];
3116: double **newm;
3117: double agexact;
3118: double agebegin, ageend;
1.222 brouard 3119: double **oldm, **savm;
1.217 brouard 3120:
1.222 brouard 3121: oldm=oldms;savm=savms;
1.217 brouard 3122: /* Hstepm could be zero and should return the unit matrix */
3123: for (i=1;i<=nlstate+ndeath;i++)
3124: for (j=1;j<=nlstate+ndeath;j++){
3125: oldm[i][j]=(i==j ? 1.0 : 0.0);
3126: po[i][j][0]=(i==j ? 1.0 : 0.0);
3127: }
3128: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3129: for(h=1; h <=nhstepm; h++){
3130: for(d=1; d <=hstepm; d++){
3131: newm=savm;
3132: /* Covariates have to be included here again */
3133: cov[1]=1.;
3134: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3135: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3136: cov[2]=agexact;
3137: if(nagesqr==1)
1.222 brouard 3138: cov[3]= agexact*agexact;
1.218 brouard 3139: for (k=1; k<=cptcovn;k++)
1.222 brouard 3140: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3141: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3142: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3143: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3144: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3145: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3146: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3147: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3148: /* 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 3149:
3150:
1.217 brouard 3151: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3152: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3153: /* Careful transposed matrix */
1.222 brouard 3154: /* age is in cov[2] */
1.218 brouard 3155: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3156: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3157: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3158: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3159: /* if((int)age == 70){ */
3160: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3161: /* for(i=1; i<=nlstate+ndeath; i++) { */
3162: /* printf("%d pmmij ",i); */
3163: /* for(j=1;j<=nlstate+ndeath;j++) { */
3164: /* printf("%f ",pmmij[i][j]); */
3165: /* } */
3166: /* printf(" oldm "); */
3167: /* for(j=1;j<=nlstate+ndeath;j++) { */
3168: /* printf("%f ",oldm[i][j]); */
3169: /* } */
3170: /* printf("\n"); */
3171: /* } */
3172: /* } */
3173: savm=oldm;
3174: oldm=newm;
3175: }
3176: for(i=1; i<=nlstate+ndeath; i++)
3177: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3178: po[i][j][h]=newm[i][j];
3179: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3180: }
3181: /*printf("h=%d ",h);*/
3182: } /* end h */
1.222 brouard 3183: /* printf("\n H=%d \n",h); */
1.217 brouard 3184: return po;
3185: }
3186:
3187:
1.162 brouard 3188: #ifdef NLOPT
3189: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3190: double fret;
3191: double *xt;
3192: int j;
3193: myfunc_data *d2 = (myfunc_data *) pd;
3194: /* xt = (p1-1); */
3195: xt=vector(1,n);
3196: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3197:
3198: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3199: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3200: printf("Function = %.12lf ",fret);
3201: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3202: printf("\n");
3203: free_vector(xt,1,n);
3204: return fret;
3205: }
3206: #endif
1.126 brouard 3207:
3208: /*************** log-likelihood *************/
3209: double func( double *x)
3210: {
1.226 brouard 3211: int i, ii, j, k, mi, d, kk;
3212: int ioffset=0;
3213: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3214: double **out;
3215: double lli; /* Individual log likelihood */
3216: int s1, s2;
1.228 brouard 3217: 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 3218: double bbh, survp;
3219: long ipmx;
3220: double agexact;
3221: /*extern weight */
3222: /* We are differentiating ll according to initial status */
3223: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3224: /*for(i=1;i<imx;i++)
3225: printf(" %d\n",s[4][i]);
3226: */
1.162 brouard 3227:
1.226 brouard 3228: ++countcallfunc;
1.162 brouard 3229:
1.226 brouard 3230: cov[1]=1.;
1.126 brouard 3231:
1.226 brouard 3232: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3233: ioffset=0;
1.226 brouard 3234: if(mle==1){
3235: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3236: /* Computes the values of the ncovmodel covariates of the model
3237: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3238: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3239: to be observed in j being in i according to the model.
3240: */
1.243 brouard 3241: ioffset=2+nagesqr ;
1.233 brouard 3242: /* Fixed */
1.234 brouard 3243: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3244: 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)*/
3245: }
1.226 brouard 3246: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3247: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3248: has been calculated etc */
3249: /* For an individual i, wav[i] gives the number of effective waves */
3250: /* We compute the contribution to Likelihood of each effective transition
3251: mw[mi][i] is real wave of the mi th effectve wave */
3252: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3253: s2=s[mw[mi+1][i]][i];
3254: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3255: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3256: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3257: */
3258: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3259: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3260: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3261: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
1.234 brouard 3262: }
3263: for (ii=1;ii<=nlstate+ndeath;ii++)
3264: for (j=1;j<=nlstate+ndeath;j++){
3265: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3266: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3267: }
3268: for(d=0; d<dh[mi][i]; d++){
3269: newm=savm;
3270: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3271: cov[2]=agexact;
3272: if(nagesqr==1)
3273: cov[3]= agexact*agexact; /* Should be changed here */
3274: for (kk=1; kk<=cptcovage;kk++) {
1.242 brouard 3275: if(!FixedV[Tvar[Tage[kk]]])
1.234 brouard 3276: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
1.242 brouard 3277: else
3278: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
1.234 brouard 3279: }
3280: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3281: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3282: savm=oldm;
3283: oldm=newm;
3284: } /* end mult */
3285:
3286: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3287: /* But now since version 0.9 we anticipate for bias at large stepm.
3288: * If stepm is larger than one month (smallest stepm) and if the exact delay
3289: * (in months) between two waves is not a multiple of stepm, we rounded to
3290: * the nearest (and in case of equal distance, to the lowest) interval but now
3291: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3292: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3293: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3294: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3295: * -stepm/2 to stepm/2 .
3296: * For stepm=1 the results are the same as for previous versions of Imach.
3297: * For stepm > 1 the results are less biased than in previous versions.
3298: */
1.234 brouard 3299: s1=s[mw[mi][i]][i];
3300: s2=s[mw[mi+1][i]][i];
3301: bbh=(double)bh[mi][i]/(double)stepm;
3302: /* bias bh is positive if real duration
3303: * is higher than the multiple of stepm and negative otherwise.
3304: */
3305: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3306: if( s2 > nlstate){
3307: /* i.e. if s2 is a death state and if the date of death is known
3308: then the contribution to the likelihood is the probability to
3309: die between last step unit time and current step unit time,
3310: which is also equal to probability to die before dh
3311: minus probability to die before dh-stepm .
3312: In version up to 0.92 likelihood was computed
3313: as if date of death was unknown. Death was treated as any other
3314: health state: the date of the interview describes the actual state
3315: and not the date of a change in health state. The former idea was
3316: to consider that at each interview the state was recorded
3317: (healthy, disable or death) and IMaCh was corrected; but when we
3318: introduced the exact date of death then we should have modified
3319: the contribution of an exact death to the likelihood. This new
3320: contribution is smaller and very dependent of the step unit
3321: stepm. It is no more the probability to die between last interview
3322: and month of death but the probability to survive from last
3323: interview up to one month before death multiplied by the
3324: probability to die within a month. Thanks to Chris
3325: Jackson for correcting this bug. Former versions increased
3326: mortality artificially. The bad side is that we add another loop
3327: which slows down the processing. The difference can be up to 10%
3328: lower mortality.
3329: */
3330: /* If, at the beginning of the maximization mostly, the
3331: cumulative probability or probability to be dead is
3332: constant (ie = 1) over time d, the difference is equal to
3333: 0. out[s1][3] = savm[s1][3]: probability, being at state
3334: s1 at precedent wave, to be dead a month before current
3335: wave is equal to probability, being at state s1 at
3336: precedent wave, to be dead at mont of the current
3337: wave. Then the observed probability (that this person died)
3338: is null according to current estimated parameter. In fact,
3339: it should be very low but not zero otherwise the log go to
3340: infinity.
3341: */
1.183 brouard 3342: /* #ifdef INFINITYORIGINAL */
3343: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3344: /* #else */
3345: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3346: /* lli=log(mytinydouble); */
3347: /* else */
3348: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3349: /* #endif */
1.226 brouard 3350: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3351:
1.226 brouard 3352: } else if ( s2==-1 ) { /* alive */
3353: for (j=1,survp=0. ; j<=nlstate; j++)
3354: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3355: /*survp += out[s1][j]; */
3356: lli= log(survp);
3357: }
3358: else if (s2==-4) {
3359: for (j=3,survp=0. ; j<=nlstate; j++)
3360: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3361: lli= log(survp);
3362: }
3363: else if (s2==-5) {
3364: for (j=1,survp=0. ; j<=2; j++)
3365: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3366: lli= log(survp);
3367: }
3368: else{
3369: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3370: /* 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 */
3371: }
3372: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3373: /*if(lli ==000.0)*/
3374: /*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); */
3375: ipmx +=1;
3376: sw += weight[i];
3377: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3378: /* if (lli < log(mytinydouble)){ */
3379: /* 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); */
3380: /* 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]); */
3381: /* } */
3382: } /* end of wave */
3383: } /* end of individual */
3384: } else if(mle==2){
3385: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3386: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3387: for(mi=1; mi<= wav[i]-1; mi++){
3388: for (ii=1;ii<=nlstate+ndeath;ii++)
3389: for (j=1;j<=nlstate+ndeath;j++){
3390: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3391: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3392: }
3393: for(d=0; d<=dh[mi][i]; d++){
3394: newm=savm;
3395: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3396: cov[2]=agexact;
3397: if(nagesqr==1)
3398: cov[3]= agexact*agexact;
3399: for (kk=1; kk<=cptcovage;kk++) {
3400: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3401: }
3402: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3403: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3404: savm=oldm;
3405: oldm=newm;
3406: } /* end mult */
3407:
3408: s1=s[mw[mi][i]][i];
3409: s2=s[mw[mi+1][i]][i];
3410: bbh=(double)bh[mi][i]/(double)stepm;
3411: 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 */
3412: ipmx +=1;
3413: sw += weight[i];
3414: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3415: } /* end of wave */
3416: } /* end of individual */
3417: } else if(mle==3){ /* exponential inter-extrapolation */
3418: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3419: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3420: for(mi=1; mi<= wav[i]-1; mi++){
3421: for (ii=1;ii<=nlstate+ndeath;ii++)
3422: for (j=1;j<=nlstate+ndeath;j++){
3423: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3424: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3425: }
3426: for(d=0; d<dh[mi][i]; d++){
3427: newm=savm;
3428: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3429: cov[2]=agexact;
3430: if(nagesqr==1)
3431: cov[3]= agexact*agexact;
3432: for (kk=1; kk<=cptcovage;kk++) {
3433: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3434: }
3435: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3436: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3437: savm=oldm;
3438: oldm=newm;
3439: } /* end mult */
3440:
3441: s1=s[mw[mi][i]][i];
3442: s2=s[mw[mi+1][i]][i];
3443: bbh=(double)bh[mi][i]/(double)stepm;
3444: 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 */
3445: ipmx +=1;
3446: sw += weight[i];
3447: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3448: } /* end of wave */
3449: } /* end of individual */
3450: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3451: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3452: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3453: for(mi=1; mi<= wav[i]-1; mi++){
3454: for (ii=1;ii<=nlstate+ndeath;ii++)
3455: for (j=1;j<=nlstate+ndeath;j++){
3456: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3457: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3458: }
3459: for(d=0; d<dh[mi][i]; d++){
3460: newm=savm;
3461: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3462: cov[2]=agexact;
3463: if(nagesqr==1)
3464: cov[3]= agexact*agexact;
3465: for (kk=1; kk<=cptcovage;kk++) {
3466: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3467: }
1.126 brouard 3468:
1.226 brouard 3469: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3470: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3471: savm=oldm;
3472: oldm=newm;
3473: } /* end mult */
3474:
3475: s1=s[mw[mi][i]][i];
3476: s2=s[mw[mi+1][i]][i];
3477: if( s2 > nlstate){
3478: lli=log(out[s1][s2] - savm[s1][s2]);
3479: } else if ( s2==-1 ) { /* alive */
3480: for (j=1,survp=0. ; j<=nlstate; j++)
3481: survp += out[s1][j];
3482: lli= log(survp);
3483: }else{
3484: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3485: }
3486: ipmx +=1;
3487: sw += weight[i];
3488: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3489: /* 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 3490: } /* end of wave */
3491: } /* end of individual */
3492: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3493: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3494: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3495: for(mi=1; mi<= wav[i]-1; mi++){
3496: for (ii=1;ii<=nlstate+ndeath;ii++)
3497: for (j=1;j<=nlstate+ndeath;j++){
3498: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3499: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3500: }
3501: for(d=0; d<dh[mi][i]; d++){
3502: newm=savm;
3503: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3504: cov[2]=agexact;
3505: if(nagesqr==1)
3506: cov[3]= agexact*agexact;
3507: for (kk=1; kk<=cptcovage;kk++) {
3508: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3509: }
1.126 brouard 3510:
1.226 brouard 3511: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3512: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3513: savm=oldm;
3514: oldm=newm;
3515: } /* end mult */
3516:
3517: s1=s[mw[mi][i]][i];
3518: s2=s[mw[mi+1][i]][i];
3519: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3520: ipmx +=1;
3521: sw += weight[i];
3522: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3523: /*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]);*/
3524: } /* end of wave */
3525: } /* end of individual */
3526: } /* End of if */
3527: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3528: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3529: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3530: return -l;
1.126 brouard 3531: }
3532:
3533: /*************** log-likelihood *************/
3534: double funcone( double *x)
3535: {
1.228 brouard 3536: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3537: int i, ii, j, k, mi, d, kk;
1.228 brouard 3538: int ioffset=0;
1.131 brouard 3539: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3540: double **out;
3541: double lli; /* Individual log likelihood */
3542: double llt;
3543: int s1, s2;
1.228 brouard 3544: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3545:
1.126 brouard 3546: double bbh, survp;
1.187 brouard 3547: double agexact;
1.214 brouard 3548: double agebegin, ageend;
1.126 brouard 3549: /*extern weight */
3550: /* We are differentiating ll according to initial status */
3551: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3552: /*for(i=1;i<imx;i++)
3553: printf(" %d\n",s[4][i]);
3554: */
3555: cov[1]=1.;
3556:
3557: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3558: ioffset=0;
3559: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.243 brouard 3560: /* ioffset=2+nagesqr+cptcovage; */
3561: ioffset=2+nagesqr;
1.232 brouard 3562: /* Fixed */
1.224 brouard 3563: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3564: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3565: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3566: 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)*/
3567: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3568: /* cov[2+6]=covar[Tvar[6]][i]; */
3569: /* cov[2+6]=covar[2][i]; V2 */
3570: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3571: /* cov[2+7]=covar[Tvar[7]][i]; */
3572: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3573: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3574: /* cov[2+9]=covar[Tvar[9]][i]; */
3575: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3576: }
1.232 brouard 3577: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3578: /* 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?)*\/ */
3579: /* } */
1.231 brouard 3580: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3581: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3582: /* } */
1.225 brouard 3583:
1.233 brouard 3584:
3585: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3586: /* Wave varying (but not age varying) */
3587: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3588: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3589: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
3590: }
1.232 brouard 3591: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.242 brouard 3592: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3593: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
3594: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3595: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
3596: /* 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 3597: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
1.242 brouard 3598: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3599: /* /\* 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]); *\/ */
3600: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
1.232 brouard 3601: /* } */
1.126 brouard 3602: for (ii=1;ii<=nlstate+ndeath;ii++)
1.242 brouard 3603: for (j=1;j<=nlstate+ndeath;j++){
3604: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3605: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3606: }
1.214 brouard 3607:
3608: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3609: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3610: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.247 brouard 3611: /* for(d=0; d<=0; d++){ /\* Delay between two effective waves Only one matrix to speed up*\/ */
1.242 brouard 3612: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3613: and mw[mi+1][i]. dh depends on stepm.*/
3614: newm=savm;
1.247 brouard 3615: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM; /* Here d is needed */
1.242 brouard 3616: cov[2]=agexact;
3617: if(nagesqr==1)
3618: cov[3]= agexact*agexact;
3619: for (kk=1; kk<=cptcovage;kk++) {
3620: if(!FixedV[Tvar[Tage[kk]]])
3621: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3622: else
3623: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
3624: }
3625: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3626: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3627: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3628: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3629: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3630: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3631: savm=oldm;
3632: oldm=newm;
1.126 brouard 3633: } /* end mult */
3634:
3635: s1=s[mw[mi][i]][i];
3636: s2=s[mw[mi+1][i]][i];
1.217 brouard 3637: /* if(s2==-1){ */
3638: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3639: /* /\* exit(1); *\/ */
3640: /* } */
1.126 brouard 3641: bbh=(double)bh[mi][i]/(double)stepm;
3642: /* bias is positive if real duration
3643: * is higher than the multiple of stepm and negative otherwise.
3644: */
3645: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.242 brouard 3646: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3647: } else if ( s2==-1 ) { /* alive */
1.242 brouard 3648: for (j=1,survp=0. ; j<=nlstate; j++)
3649: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3650: lli= log(survp);
1.126 brouard 3651: }else if (mle==1){
1.242 brouard 3652: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3653: } else if(mle==2){
1.242 brouard 3654: 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 3655: } else if(mle==3){ /* exponential inter-extrapolation */
1.242 brouard 3656: 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 3657: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.242 brouard 3658: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3659: } else{ /* mle=0 back to 1 */
1.242 brouard 3660: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3661: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3662: } /* End of if */
3663: ipmx +=1;
3664: sw += weight[i];
3665: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3666: /*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 3667: if(globpr){
1.246 brouard 3668: fprintf(ficresilk,"%09ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3669: %11.6f %11.6f %11.6f ", \
1.242 brouard 3670: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3671: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3672: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3673: llt +=ll[k]*gipmx/gsw;
3674: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3675: }
3676: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3677: }
1.232 brouard 3678: } /* end of wave */
3679: } /* end of individual */
3680: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3681: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3682: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3683: if(globpr==0){ /* First time we count the contributions and weights */
3684: gipmx=ipmx;
3685: gsw=sw;
3686: }
3687: return -l;
1.126 brouard 3688: }
3689:
3690:
3691: /*************** function likelione ***********/
3692: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3693: {
3694: /* This routine should help understanding what is done with
3695: the selection of individuals/waves and
3696: to check the exact contribution to the likelihood.
3697: Plotting could be done.
3698: */
3699: int k;
3700:
3701: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3702: strcpy(fileresilk,"ILK_");
1.202 brouard 3703: strcat(fileresilk,fileresu);
1.126 brouard 3704: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3705: printf("Problem with resultfile: %s\n", fileresilk);
3706: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3707: }
1.214 brouard 3708: 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");
3709: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3710: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3711: for(k=1; k<=nlstate; k++)
3712: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3713: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3714: }
3715:
3716: *fretone=(*funcone)(p);
3717: if(*globpri !=0){
3718: fclose(ficresilk);
1.205 brouard 3719: if (mle ==0)
3720: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3721: else if(mle >=1)
3722: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3723: 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 3724:
1.208 brouard 3725:
3726: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3727: 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 3728: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3729: }
1.207 brouard 3730: 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 3731: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3732: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3733: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3734: fflush(fichtm);
1.205 brouard 3735: }
1.126 brouard 3736: return;
3737: }
3738:
3739:
3740: /*********** Maximum Likelihood Estimation ***************/
3741:
3742: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3743: {
1.165 brouard 3744: int i,j, iter=0;
1.126 brouard 3745: double **xi;
3746: double fret;
3747: double fretone; /* Only one call to likelihood */
3748: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3749:
3750: #ifdef NLOPT
3751: int creturn;
3752: nlopt_opt opt;
3753: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3754: double *lb;
3755: double minf; /* the minimum objective value, upon return */
3756: double * p1; /* Shifted parameters from 0 instead of 1 */
3757: myfunc_data dinst, *d = &dinst;
3758: #endif
3759:
3760:
1.126 brouard 3761: xi=matrix(1,npar,1,npar);
3762: for (i=1;i<=npar;i++)
3763: for (j=1;j<=npar;j++)
3764: xi[i][j]=(i==j ? 1.0 : 0.0);
3765: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3766: strcpy(filerespow,"POW_");
1.126 brouard 3767: strcat(filerespow,fileres);
3768: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3769: printf("Problem with resultfile: %s\n", filerespow);
3770: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3771: }
3772: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3773: for (i=1;i<=nlstate;i++)
3774: for(j=1;j<=nlstate+ndeath;j++)
3775: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3776: fprintf(ficrespow,"\n");
1.162 brouard 3777: #ifdef POWELL
1.126 brouard 3778: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3779: #endif
1.126 brouard 3780:
1.162 brouard 3781: #ifdef NLOPT
3782: #ifdef NEWUOA
3783: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3784: #else
3785: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3786: #endif
3787: lb=vector(0,npar-1);
3788: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3789: nlopt_set_lower_bounds(opt, lb);
3790: nlopt_set_initial_step1(opt, 0.1);
3791:
3792: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3793: d->function = func;
3794: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3795: nlopt_set_min_objective(opt, myfunc, d);
3796: nlopt_set_xtol_rel(opt, ftol);
3797: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3798: printf("nlopt failed! %d\n",creturn);
3799: }
3800: else {
3801: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3802: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3803: iter=1; /* not equal */
3804: }
3805: nlopt_destroy(opt);
3806: #endif
1.126 brouard 3807: free_matrix(xi,1,npar,1,npar);
3808: fclose(ficrespow);
1.203 brouard 3809: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3810: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3811: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3812:
3813: }
3814:
3815: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3816: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3817: {
3818: double **a,**y,*x,pd;
1.203 brouard 3819: /* double **hess; */
1.164 brouard 3820: int i, j;
1.126 brouard 3821: int *indx;
3822:
3823: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3824: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3825: void lubksb(double **a, int npar, int *indx, double b[]) ;
3826: void ludcmp(double **a, int npar, int *indx, double *d) ;
3827: double gompertz(double p[]);
1.203 brouard 3828: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3829:
3830: printf("\nCalculation of the hessian matrix. Wait...\n");
3831: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3832: for (i=1;i<=npar;i++){
1.203 brouard 3833: printf("%d-",i);fflush(stdout);
3834: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3835:
3836: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3837:
3838: /* printf(" %f ",p[i]);
3839: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3840: }
3841:
3842: for (i=1;i<=npar;i++) {
3843: for (j=1;j<=npar;j++) {
3844: if (j>i) {
1.203 brouard 3845: printf(".%d-%d",i,j);fflush(stdout);
3846: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3847: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3848:
3849: hess[j][i]=hess[i][j];
3850: /*printf(" %lf ",hess[i][j]);*/
3851: }
3852: }
3853: }
3854: printf("\n");
3855: fprintf(ficlog,"\n");
3856:
3857: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3858: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3859:
3860: a=matrix(1,npar,1,npar);
3861: y=matrix(1,npar,1,npar);
3862: x=vector(1,npar);
3863: indx=ivector(1,npar);
3864: for (i=1;i<=npar;i++)
3865: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3866: ludcmp(a,npar,indx,&pd);
3867:
3868: for (j=1;j<=npar;j++) {
3869: for (i=1;i<=npar;i++) x[i]=0;
3870: x[j]=1;
3871: lubksb(a,npar,indx,x);
3872: for (i=1;i<=npar;i++){
3873: matcov[i][j]=x[i];
3874: }
3875: }
3876:
3877: printf("\n#Hessian matrix#\n");
3878: fprintf(ficlog,"\n#Hessian matrix#\n");
3879: for (i=1;i<=npar;i++) {
3880: for (j=1;j<=npar;j++) {
1.203 brouard 3881: printf("%.6e ",hess[i][j]);
3882: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3883: }
3884: printf("\n");
3885: fprintf(ficlog,"\n");
3886: }
3887:
1.203 brouard 3888: /* printf("\n#Covariance matrix#\n"); */
3889: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3890: /* for (i=1;i<=npar;i++) { */
3891: /* for (j=1;j<=npar;j++) { */
3892: /* printf("%.6e ",matcov[i][j]); */
3893: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3894: /* } */
3895: /* printf("\n"); */
3896: /* fprintf(ficlog,"\n"); */
3897: /* } */
3898:
1.126 brouard 3899: /* Recompute Inverse */
1.203 brouard 3900: /* for (i=1;i<=npar;i++) */
3901: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3902: /* ludcmp(a,npar,indx,&pd); */
3903:
3904: /* printf("\n#Hessian matrix recomputed#\n"); */
3905:
3906: /* for (j=1;j<=npar;j++) { */
3907: /* for (i=1;i<=npar;i++) x[i]=0; */
3908: /* x[j]=1; */
3909: /* lubksb(a,npar,indx,x); */
3910: /* for (i=1;i<=npar;i++){ */
3911: /* y[i][j]=x[i]; */
3912: /* printf("%.3e ",y[i][j]); */
3913: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3914: /* } */
3915: /* printf("\n"); */
3916: /* fprintf(ficlog,"\n"); */
3917: /* } */
3918:
3919: /* Verifying the inverse matrix */
3920: #ifdef DEBUGHESS
3921: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3922:
1.203 brouard 3923: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3924: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3925:
3926: for (j=1;j<=npar;j++) {
3927: for (i=1;i<=npar;i++){
1.203 brouard 3928: printf("%.2f ",y[i][j]);
3929: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3930: }
3931: printf("\n");
3932: fprintf(ficlog,"\n");
3933: }
1.203 brouard 3934: #endif
1.126 brouard 3935:
3936: free_matrix(a,1,npar,1,npar);
3937: free_matrix(y,1,npar,1,npar);
3938: free_vector(x,1,npar);
3939: free_ivector(indx,1,npar);
1.203 brouard 3940: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3941:
3942:
3943: }
3944:
3945: /*************** hessian matrix ****************/
3946: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3947: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3948: int i;
3949: int l=1, lmax=20;
1.203 brouard 3950: double k1,k2, res, fx;
1.132 brouard 3951: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3952: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3953: int k=0,kmax=10;
3954: double l1;
3955:
3956: fx=func(x);
3957: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3958: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3959: l1=pow(10,l);
3960: delts=delt;
3961: for(k=1 ; k <kmax; k=k+1){
3962: delt = delta*(l1*k);
3963: p2[theta]=x[theta] +delt;
1.145 brouard 3964: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 3965: p2[theta]=x[theta]-delt;
3966: k2=func(p2)-fx;
3967: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 3968: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 3969:
1.203 brouard 3970: #ifdef DEBUGHESSII
1.126 brouard 3971: 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);
3972: 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);
3973: #endif
3974: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
3975: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
3976: k=kmax;
3977: }
3978: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 3979: k=kmax; l=lmax*10;
1.126 brouard 3980: }
3981: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
3982: delts=delt;
3983: }
1.203 brouard 3984: } /* End loop k */
1.126 brouard 3985: }
3986: delti[theta]=delts;
3987: return res;
3988:
3989: }
3990:
1.203 brouard 3991: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 3992: {
3993: int i;
1.164 brouard 3994: int l=1, lmax=20;
1.126 brouard 3995: double k1,k2,k3,k4,res,fx;
1.132 brouard 3996: double p2[MAXPARM+1];
1.203 brouard 3997: int k, kmax=1;
3998: double v1, v2, cv12, lc1, lc2;
1.208 brouard 3999:
4000: int firstime=0;
1.203 brouard 4001:
1.126 brouard 4002: fx=func(x);
1.203 brouard 4003: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 4004: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 4005: p2[thetai]=x[thetai]+delti[thetai]*k;
4006: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4007: k1=func(p2)-fx;
4008:
1.203 brouard 4009: p2[thetai]=x[thetai]+delti[thetai]*k;
4010: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4011: k2=func(p2)-fx;
4012:
1.203 brouard 4013: p2[thetai]=x[thetai]-delti[thetai]*k;
4014: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4015: k3=func(p2)-fx;
4016:
1.203 brouard 4017: p2[thetai]=x[thetai]-delti[thetai]*k;
4018: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4019: k4=func(p2)-fx;
1.203 brouard 4020: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
4021: if(k1*k2*k3*k4 <0.){
1.208 brouard 4022: firstime=1;
1.203 brouard 4023: kmax=kmax+10;
1.208 brouard 4024: }
4025: if(kmax >=10 || firstime ==1){
1.246 brouard 4026: 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);
4027: 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 4028: 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);
4029: 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);
4030: }
4031: #ifdef DEBUGHESSIJ
4032: v1=hess[thetai][thetai];
4033: v2=hess[thetaj][thetaj];
4034: cv12=res;
4035: /* Computing eigen value of Hessian matrix */
4036: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4037: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4038: if ((lc2 <0) || (lc1 <0) ){
4039: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4040: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4041: 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);
4042: 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);
4043: }
1.126 brouard 4044: #endif
4045: }
4046: return res;
4047: }
4048:
1.203 brouard 4049: /* Not done yet: Was supposed to fix if not exactly at the maximum */
4050: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
4051: /* { */
4052: /* int i; */
4053: /* int l=1, lmax=20; */
4054: /* double k1,k2,k3,k4,res,fx; */
4055: /* double p2[MAXPARM+1]; */
4056: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
4057: /* int k=0,kmax=10; */
4058: /* double l1; */
4059:
4060: /* fx=func(x); */
4061: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
4062: /* l1=pow(10,l); */
4063: /* delts=delt; */
4064: /* for(k=1 ; k <kmax; k=k+1){ */
4065: /* delt = delti*(l1*k); */
4066: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
4067: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4068: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4069: /* k1=func(p2)-fx; */
4070:
4071: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4072: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4073: /* k2=func(p2)-fx; */
4074:
4075: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4076: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4077: /* k3=func(p2)-fx; */
4078:
4079: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4080: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4081: /* k4=func(p2)-fx; */
4082: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
4083: /* #ifdef DEBUGHESSIJ */
4084: /* 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); */
4085: /* 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); */
4086: /* #endif */
4087: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
4088: /* k=kmax; */
4089: /* } */
4090: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
4091: /* k=kmax; l=lmax*10; */
4092: /* } */
4093: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
4094: /* delts=delt; */
4095: /* } */
4096: /* } /\* End loop k *\/ */
4097: /* } */
4098: /* delti[theta]=delts; */
4099: /* return res; */
4100: /* } */
4101:
4102:
1.126 brouard 4103: /************** Inverse of matrix **************/
4104: void ludcmp(double **a, int n, int *indx, double *d)
4105: {
4106: int i,imax,j,k;
4107: double big,dum,sum,temp;
4108: double *vv;
4109:
4110: vv=vector(1,n);
4111: *d=1.0;
4112: for (i=1;i<=n;i++) {
4113: big=0.0;
4114: for (j=1;j<=n;j++)
4115: if ((temp=fabs(a[i][j])) > big) big=temp;
4116: if (big == 0.0) nrerror("Singular matrix in routine ludcmp");
4117: vv[i]=1.0/big;
4118: }
4119: for (j=1;j<=n;j++) {
4120: for (i=1;i<j;i++) {
4121: sum=a[i][j];
4122: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
4123: a[i][j]=sum;
4124: }
4125: big=0.0;
4126: for (i=j;i<=n;i++) {
4127: sum=a[i][j];
4128: for (k=1;k<j;k++)
4129: sum -= a[i][k]*a[k][j];
4130: a[i][j]=sum;
4131: if ( (dum=vv[i]*fabs(sum)) >= big) {
4132: big=dum;
4133: imax=i;
4134: }
4135: }
4136: if (j != imax) {
4137: for (k=1;k<=n;k++) {
4138: dum=a[imax][k];
4139: a[imax][k]=a[j][k];
4140: a[j][k]=dum;
4141: }
4142: *d = -(*d);
4143: vv[imax]=vv[j];
4144: }
4145: indx[j]=imax;
4146: if (a[j][j] == 0.0) a[j][j]=TINY;
4147: if (j != n) {
4148: dum=1.0/(a[j][j]);
4149: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4150: }
4151: }
4152: free_vector(vv,1,n); /* Doesn't work */
4153: ;
4154: }
4155:
4156: void lubksb(double **a, int n, int *indx, double b[])
4157: {
4158: int i,ii=0,ip,j;
4159: double sum;
4160:
4161: for (i=1;i<=n;i++) {
4162: ip=indx[i];
4163: sum=b[ip];
4164: b[ip]=b[i];
4165: if (ii)
4166: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4167: else if (sum) ii=i;
4168: b[i]=sum;
4169: }
4170: for (i=n;i>=1;i--) {
4171: sum=b[i];
4172: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4173: b[i]=sum/a[i][i];
4174: }
4175: }
4176:
4177: void pstamp(FILE *fichier)
4178: {
1.196 brouard 4179: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4180: }
4181:
4182: /************ Frequencies ********************/
1.226 brouard 4183: void freqsummary(char fileres[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
4184: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4185: int firstpass, int lastpass, int stepm, int weightopt, char model[])
4186: { /* Some frequencies */
4187:
1.227 brouard 4188: int i, m, jk, j1, bool, z1,j, k, iv;
1.226 brouard 4189: int iind=0, iage=0;
4190: int mi; /* Effective wave */
4191: int first;
4192: double ***freq; /* Frequencies */
4193: double *meanq;
4194: double **meanqt;
4195: double *pp, **prop, *posprop, *pospropt;
4196: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4197: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4198: double agebegin, ageend;
4199:
4200: pp=vector(1,nlstate);
4201: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4202: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4203: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4204: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4205: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4206: meanqt=matrix(1,lastpass,1,nqtveff);
4207: strcpy(fileresp,"P_");
4208: strcat(fileresp,fileresu);
4209: /*strcat(fileresphtm,fileresu);*/
4210: if((ficresp=fopen(fileresp,"w"))==NULL) {
4211: printf("Problem with prevalence resultfile: %s\n", fileresp);
4212: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4213: exit(0);
4214: }
1.240 brouard 4215:
1.226 brouard 4216: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4217: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4218: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4219: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4220: fflush(ficlog);
4221: exit(70);
4222: }
4223: else{
4224: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4225: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4226: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4227: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4228: }
1.237 brouard 4229: 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 4230:
1.226 brouard 4231: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4232: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4233: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4234: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4235: fflush(ficlog);
4236: exit(70);
1.240 brouard 4237: } else{
1.226 brouard 4238: 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 4239: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4240: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4241: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4242: }
1.240 brouard 4243: 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);
4244:
1.226 brouard 4245: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4246: j1=0;
1.126 brouard 4247:
1.227 brouard 4248: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4249: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4250: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.240 brouard 4251:
4252:
1.226 brouard 4253: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4254: reference=low_education V1=0,V2=0
4255: med_educ V1=1 V2=0,
4256: high_educ V1=0 V2=1
4257: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4258: */
1.249 ! brouard 4259: dateintsum=0;
! 4260: k2cpt=0;
! 4261:
! 4262: for (j = 0; j <= cptcoveff; j+=cptcoveff){
! 4263: first=1;
1.227 brouard 4264: 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 */
1.226 brouard 4265: posproptt=0.;
4266: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4267: scanf("%d", i);*/
4268: for (i=-5; i<=nlstate+ndeath; i++)
4269: for (jk=-5; jk<=nlstate+ndeath; jk++)
1.240 brouard 4270: for(m=iagemin; m <= iagemax+3; m++)
4271: freq[i][jk][m]=0;
4272:
1.226 brouard 4273: for (i=1; i<=nlstate; i++) {
4274: for(m=iagemin; m <= iagemax+3; m++)
1.240 brouard 4275: prop[i][m]=0;
1.226 brouard 4276: posprop[i]=0;
4277: pospropt[i]=0;
4278: }
1.227 brouard 4279: /* for (z1=1; z1<= nqfveff; z1++) { */
4280: /* meanq[z1]+=0.; */
4281: /* for(m=1;m<=lastpass;m++){ */
4282: /* meanqt[m][z1]=0.; */
4283: /* } */
4284: /* } */
1.240 brouard 4285:
1.249 ! brouard 4286: /* dateintsum=0; */
! 4287: /* k2cpt=0; */
! 4288:
1.227 brouard 4289: /* For that combination of covariate j1, we count and print the frequencies in one pass */
1.226 brouard 4290: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4291: bool=1;
1.249 ! brouard 4292: if(j !=0){
1.227 brouard 4293: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
1.234 brouard 4294: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
1.227 brouard 4295: /* for (z1=1; z1<= nqfveff; z1++) { */
4296: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4297: /* } */
1.234 brouard 4298: for (z1=1; z1<=cptcoveff; z1++) {
4299: /* if(Tvaraff[z1] ==-20){ */
4300: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4301: /* }else if(Tvaraff[z1] ==-10){ */
4302: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4303: /* }else */
4304: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){
4305: /* Tests if this individual iind responded to j1 (V4=1 V3=0) */
4306: bool=0;
4307: /* 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",
4308: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4309: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4310: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4311: } /* Onlyf fixed */
4312: } /* end z1 */
4313: } /* cptcovn > 0 */
1.227 brouard 4314: } /* end any */
1.249 ! brouard 4315: }/* end j==0 */
1.227 brouard 4316: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
1.234 brouard 4317: /* for(m=firstpass; m<=lastpass; m++){ */
4318: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4319: m=mw[mi][iind];
1.249 ! brouard 4320: if(j!=0){
1.234 brouard 4321: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4322: for (z1=1; z1<=cptcoveff; z1++) {
4323: if( Fixed[Tmodelind[z1]]==1){
4324: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4325: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
1.249 ! brouard 4326: bool=0; /* not selected */
1.234 brouard 4327: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4328: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4329: bool=0;
4330: }
4331: }
4332: }
4333: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
1.249 ! brouard 4334: } /* end j==0 */
1.234 brouard 4335: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4336: if(bool==1){
4337: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4338: and mw[mi+1][iind]. dh depends on stepm. */
4339: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4340: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4341: if(m >=firstpass && m <=lastpass){
4342: k2=anint[m][iind]+(mint[m][iind]/12.);
4343: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4344: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4345: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4346: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4347: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4348: if (m<lastpass) {
4349: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4350: /* 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]); */
4351: if(s[m][iind]==-1)
4352: 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.));
4353: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4354: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4355: 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 */
4356: }
4357: } /* end if between passes */
1.249 ! brouard 4358: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99) && (j==0)) {
! 4359: dateintsum=dateintsum+k2; /* on all covariates ?*/
1.234 brouard 4360: k2cpt++;
4361: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
4362: }
4363: } /* end bool 2 */
4364: } /* end m */
1.226 brouard 4365: } /* end bool */
4366: } /* end iind = 1 to imx */
4367: /* prop[s][age] is feeded for any initial and valid live state as well as
4368: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
1.240 brouard 4369:
4370:
1.226 brouard 4371: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4372: pstamp(ficresp);
1.249 ! brouard 4373: if (cptcoveff>0 && j!=0){
1.226 brouard 4374: fprintf(ficresp, "\n#********** Variable ");
4375: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4376: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
1.240 brouard 4377: fprintf(ficlog, "\n#********** Variable ");
1.227 brouard 4378: for (z1=1; z1<=cptcoveff; z1++){
1.240 brouard 4379: if(DummyV[z1]){
4380: fprintf(ficresp, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4381: fprintf(ficresphtm, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4382: fprintf(ficresphtmfr, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4383: fprintf(ficlog, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4384: }else{
4385: fprintf(ficresp, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4386: fprintf(ficresphtm, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4387: fprintf(ficresphtmfr, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4388: fprintf(ficlog, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4389: }
1.226 brouard 4390: }
4391: fprintf(ficresp, "**********\n#");
4392: fprintf(ficresphtm, "**********</h3>\n");
4393: fprintf(ficresphtmfr, "**********</h3>\n");
4394: fprintf(ficlog, "**********\n");
4395: }
4396: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4397: for(i=1; i<=nlstate;i++) {
1.240 brouard 4398: fprintf(ficresp, " Age Prev(%d) N(%d) N ",i,i);
1.226 brouard 4399: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4400: }
4401: fprintf(ficresp, "\n");
4402: fprintf(ficresphtm, "\n");
1.240 brouard 4403:
1.226 brouard 4404: /* Header of frequency table by age */
4405: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4406: fprintf(ficresphtmfr,"<th>Age</th> ");
4407: for(jk=-1; jk <=nlstate+ndeath; jk++){
4408: for(m=-1; m <=nlstate+ndeath; m++){
1.234 brouard 4409: if(jk!=0 && m!=0)
4410: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.226 brouard 4411: }
4412: }
4413: fprintf(ficresphtmfr, "\n");
1.240 brouard 4414:
1.226 brouard 4415: /* For each age */
4416: for(iage=iagemin; iage <= iagemax+3; iage++){
4417: fprintf(ficresphtm,"<tr>");
4418: if(iage==iagemax+1){
1.240 brouard 4419: fprintf(ficlog,"1");
4420: fprintf(ficresphtmfr,"<tr><th>0</th> ");
1.226 brouard 4421: }else if(iage==iagemax+2){
1.240 brouard 4422: fprintf(ficlog,"0");
4423: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
1.226 brouard 4424: }else if(iage==iagemax+3){
1.240 brouard 4425: fprintf(ficlog,"Total");
4426: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
1.226 brouard 4427: }else{
1.240 brouard 4428: if(first==1){
4429: first=0;
4430: printf("See log file for details...\n");
4431: }
4432: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4433: fprintf(ficlog,"Age %d", iage);
1.226 brouard 4434: }
4435: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4436: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4437: pp[jk] += freq[jk][m][iage];
1.226 brouard 4438: }
4439: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4440: for(m=-1, pos=0; m <=0 ; m++)
4441: pos += freq[jk][m][iage];
4442: if(pp[jk]>=1.e-10){
4443: if(first==1){
4444: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4445: }
4446: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4447: }else{
4448: if(first==1)
4449: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4450: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4451: }
1.226 brouard 4452: }
1.240 brouard 4453:
1.226 brouard 4454: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4455: /* posprop[jk]=0; */
4456: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4457: pp[jk] += freq[jk][m][iage];
1.226 brouard 4458: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
1.240 brouard 4459:
1.226 brouard 4460: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
1.240 brouard 4461: pos += pp[jk]; /* pos is the total number of transitions until this age */
4462: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4463: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4464: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
4465: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
1.226 brouard 4466: }
4467: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4468: if(pos>=1.e-5){
4469: if(first==1)
4470: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4471: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4472: }else{
4473: if(first==1)
4474: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4475: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4476: }
4477: if( iage <= iagemax){
4478: if(pos>=1.e-5){
4479: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4480: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4481: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4482: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4483: }
4484: else{
4485: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4486: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4487: }
4488: }
4489: pospropt[jk] +=posprop[jk];
1.226 brouard 4490: } /* end loop jk */
4491: /* pospropt=0.; */
4492: for(jk=-1; jk <=nlstate+ndeath; jk++){
1.240 brouard 4493: for(m=-1; m <=nlstate+ndeath; m++){
4494: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4495: if(first==1){
4496: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4497: }
4498: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4499: }
4500: if(jk!=0 && m!=0)
4501: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
4502: }
1.226 brouard 4503: } /* end loop jk */
4504: posproptt=0.;
4505: for(jk=1; jk <=nlstate; jk++){
1.240 brouard 4506: posproptt += pospropt[jk];
1.226 brouard 4507: }
4508: fprintf(ficresphtmfr,"</tr>\n ");
4509: if(iage <= iagemax){
1.240 brouard 4510: fprintf(ficresp,"\n");
4511: fprintf(ficresphtm,"</tr>\n");
1.226 brouard 4512: }
4513: if(first==1)
1.240 brouard 4514: printf("Others in log...\n");
1.226 brouard 4515: fprintf(ficlog,"\n");
4516: } /* end loop age iage */
4517: fprintf(ficresphtm,"<tr><th>Tot</th>");
4518: for(jk=1; jk <=nlstate ; jk++){
4519: if(posproptt < 1.e-5){
1.240 brouard 4520: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
1.226 brouard 4521: }else{
1.240 brouard 4522: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.226 brouard 4523: }
4524: }
4525: fprintf(ficresphtm,"</tr>\n");
4526: fprintf(ficresphtm,"</table>\n");
4527: fprintf(ficresphtmfr,"</table>\n");
4528: if(posproptt < 1.e-5){
4529: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4530: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4531: fprintf(ficres,"\n This combination (%d) is not valid and no result will be produced\n\n",j1);
4532: invalidvarcomb[j1]=1;
4533: }else{
4534: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4535: invalidvarcomb[j1]=0;
4536: }
4537: fprintf(ficresphtmfr,"</table>\n");
4538: } /* end selected combination of covariate j1 */
1.249 ! brouard 4539: if(j==0){ /* We can estimate starting values from the occurences in each case */
! 4540: for(jk=-1; jk <=nlstate+ndeath; jk++){
! 4541: for(m=-1; m <=nlstate+ndeath; m++){
! 4542: /* param[i]|j][k]= freq[jk][m][iagemax+3] */
! 4543: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
! 4544: if(first==1){
! 4545: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
! 4546: }
! 4547: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
! 4548: }
! 4549: }
! 4550: } /* end loop jk */
! 4551: }
! 4552: } /* end j */
1.226 brouard 4553: dateintmean=dateintsum/k2cpt;
1.240 brouard 4554:
1.226 brouard 4555: fclose(ficresp);
4556: fclose(ficresphtm);
4557: fclose(ficresphtmfr);
4558: free_vector(meanq,1,nqfveff);
4559: free_matrix(meanqt,1,lastpass,1,nqtveff);
4560: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4561: free_vector(pospropt,1,nlstate);
4562: free_vector(posprop,1,nlstate);
4563: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4564: free_vector(pp,1,nlstate);
4565: /* End of freqsummary */
4566: }
1.126 brouard 4567:
4568: /************ Prevalence ********************/
1.227 brouard 4569: 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)
4570: {
4571: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4572: in each health status at the date of interview (if between dateprev1 and dateprev2).
4573: We still use firstpass and lastpass as another selection.
4574: */
1.126 brouard 4575:
1.227 brouard 4576: int i, m, jk, j1, bool, z1,j, iv;
4577: int mi; /* Effective wave */
4578: int iage;
4579: double agebegin, ageend;
4580:
4581: double **prop;
4582: double posprop;
4583: double y2; /* in fractional years */
4584: int iagemin, iagemax;
4585: int first; /** to stop verbosity which is redirected to log file */
4586:
4587: iagemin= (int) agemin;
4588: iagemax= (int) agemax;
4589: /*pp=vector(1,nlstate);*/
4590: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4591: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4592: j1=0;
1.222 brouard 4593:
1.227 brouard 4594: /*j=cptcoveff;*/
4595: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4596:
1.227 brouard 4597: first=1;
4598: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4599: for (i=1; i<=nlstate; i++)
4600: for(iage=iagemin-AGEMARGE; iage <= iagemax+3+AGEMARGE; iage++)
4601: prop[i][iage]=0.0;
4602: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4603: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4604: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4605:
4606: for (i=1; i<=imx; i++) { /* Each individual */
4607: bool=1;
4608: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4609: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4610: m=mw[mi][i];
4611: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4612: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4613: for (z1=1; z1<=cptcoveff; z1++){
4614: if( Fixed[Tmodelind[z1]]==1){
4615: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4616: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4617: bool=0;
4618: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4619: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4620: bool=0;
4621: }
4622: }
4623: if(bool==1){ /* Otherwise we skip that wave/person */
4624: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4625: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4626: if(m >=firstpass && m <=lastpass){
4627: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4628: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4629: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4630: if(agev[m][i]==1) agev[m][i]=iagemax+2;
4631: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+3+AGEMARGE){
4632: 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);
4633: exit(1);
4634: }
4635: if (s[m][i]>0 && s[m][i]<=nlstate) {
4636: /*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]]);*/
4637: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4638: prop[s[m][i]][iagemax+3] += weight[i];
4639: } /* end valid statuses */
4640: } /* end selection of dates */
4641: } /* end selection of waves */
4642: } /* end bool */
4643: } /* end wave */
4644: } /* end individual */
4645: for(i=iagemin; i <= iagemax+3; i++){
4646: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4647: posprop += prop[jk][i];
4648: }
4649:
4650: for(jk=1; jk <=nlstate ; jk++){
4651: if( i <= iagemax){
4652: if(posprop>=1.e-5){
4653: probs[i][jk][j1]= prop[jk][i]/posprop;
4654: } else{
4655: if(first==1){
4656: first=0;
4657: 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]);
4658: }
4659: }
4660: }
4661: }/* end jk */
4662: }/* end i */
1.222 brouard 4663: /*} *//* end i1 */
1.227 brouard 4664: } /* end j1 */
1.222 brouard 4665:
1.227 brouard 4666: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4667: /*free_vector(pp,1,nlstate);*/
4668: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4669: } /* End of prevalence */
1.126 brouard 4670:
4671: /************* Waves Concatenation ***************/
4672:
4673: 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)
4674: {
4675: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4676: Death is a valid wave (if date is known).
4677: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4678: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4679: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4680: */
1.126 brouard 4681:
1.224 brouard 4682: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4683: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4684: double sum=0., jmean=0.;*/
1.224 brouard 4685: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4686: int j, k=0,jk, ju, jl;
4687: double sum=0.;
4688: first=0;
1.214 brouard 4689: firstwo=0;
1.217 brouard 4690: firsthree=0;
1.218 brouard 4691: firstfour=0;
1.164 brouard 4692: jmin=100000;
1.126 brouard 4693: jmax=-1;
4694: jmean=0.;
1.224 brouard 4695:
4696: /* Treating live states */
1.214 brouard 4697: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4698: mi=0; /* First valid wave */
1.227 brouard 4699: mli=0; /* Last valid wave */
1.126 brouard 4700: m=firstpass;
1.214 brouard 4701: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4702: 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 */
4703: mli=m-1;/* mw[++mi][i]=m-1; */
4704: }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 */
4705: mw[++mi][i]=m;
4706: mli=m;
1.224 brouard 4707: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4708: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4709: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4710: }
1.227 brouard 4711: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4712: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4713: break;
1.224 brouard 4714: #else
1.227 brouard 4715: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4716: if(firsthree == 0){
4717: 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);
4718: firsthree=1;
4719: }
4720: 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);
4721: mw[++mi][i]=m;
4722: mli=m;
4723: }
4724: if(s[m][i]==-2){ /* Vital status is really unknown */
4725: nbwarn++;
4726: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4727: 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);
4728: 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);
4729: }
4730: break;
4731: }
4732: break;
1.224 brouard 4733: #endif
1.227 brouard 4734: }/* End m >= lastpass */
1.126 brouard 4735: }/* end while */
1.224 brouard 4736:
1.227 brouard 4737: /* 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 4738: /* After last pass */
1.224 brouard 4739: /* Treating death states */
1.214 brouard 4740: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4741: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4742: /* } */
1.126 brouard 4743: mi++; /* Death is another wave */
4744: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4745: /* Only death is a correct wave */
1.126 brouard 4746: mw[mi][i]=m;
1.224 brouard 4747: }
4748: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.227 brouard 4749: else if ((int) andc[i] != 9999) { /* Status is negative. A death occured after lastpass, we can't take it into account because of potential bias */
1.216 brouard 4750: /* m++; */
4751: /* mi++; */
4752: /* s[m][i]=nlstate+1; /\* We are setting the status to the last of non live state *\/ */
4753: /* mw[mi][i]=m; */
1.218 brouard 4754: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4755: 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 */
4756: nbwarn++;
4757: if(firstfiv==0){
4758: 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 );
4759: firstfiv=1;
4760: }else{
4761: 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 );
4762: }
4763: }else{ /* Death occured afer last wave potential bias */
4764: nberr++;
4765: if(firstwo==0){
4766: 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.\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 );
4767: firstwo=1;
4768: }
4769: 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.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
4770: }
1.218 brouard 4771: }else{ /* end date of interview is known */
1.227 brouard 4772: /* death is known but not confirmed by death status at any wave */
4773: if(firstfour==0){
4774: 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 );
4775: firstfour=1;
4776: }
4777: 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 4778: }
1.224 brouard 4779: } /* end if date of death is known */
4780: #endif
4781: wav[i]=mi; /* mi should be the last effective wave (or mli) */
4782: /* wav[i]=mw[mi][i]; */
1.126 brouard 4783: if(mi==0){
4784: nbwarn++;
4785: if(first==0){
1.227 brouard 4786: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
4787: first=1;
1.126 brouard 4788: }
4789: if(first==1){
1.227 brouard 4790: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 4791: }
4792: } /* end mi==0 */
4793: } /* End individuals */
1.214 brouard 4794: /* wav and mw are no more changed */
1.223 brouard 4795:
1.214 brouard 4796:
1.126 brouard 4797: for(i=1; i<=imx; i++){
4798: for(mi=1; mi<wav[i];mi++){
4799: if (stepm <=0)
1.227 brouard 4800: dh[mi][i]=1;
1.126 brouard 4801: else{
1.227 brouard 4802: if (s[mw[mi+1][i]][i] > nlstate) { /* A death */
4803: if (agedc[i] < 2*AGESUP) {
4804: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
4805: if(j==0) j=1; /* Survives at least one month after exam */
4806: else if(j<0){
4807: nberr++;
4808: 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]);
4809: j=1; /* Temporary Dangerous patch */
4810: 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);
4811: 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]);
4812: 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);
4813: }
4814: k=k+1;
4815: if (j >= jmax){
4816: jmax=j;
4817: ijmax=i;
4818: }
4819: if (j <= jmin){
4820: jmin=j;
4821: ijmin=i;
4822: }
4823: sum=sum+j;
4824: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
4825: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
4826: }
4827: }
4828: else{
4829: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 4830: /* 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 4831:
1.227 brouard 4832: k=k+1;
4833: if (j >= jmax) {
4834: jmax=j;
4835: ijmax=i;
4836: }
4837: else if (j <= jmin){
4838: jmin=j;
4839: ijmin=i;
4840: }
4841: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
4842: /*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]);*/
4843: if(j<0){
4844: nberr++;
4845: 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]);
4846: 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]);
4847: }
4848: sum=sum+j;
4849: }
4850: jk= j/stepm;
4851: jl= j -jk*stepm;
4852: ju= j -(jk+1)*stepm;
4853: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
4854: if(jl==0){
4855: dh[mi][i]=jk;
4856: bh[mi][i]=0;
4857: }else{ /* We want a negative bias in order to only have interpolation ie
4858: * to avoid the price of an extra matrix product in likelihood */
4859: dh[mi][i]=jk+1;
4860: bh[mi][i]=ju;
4861: }
4862: }else{
4863: if(jl <= -ju){
4864: dh[mi][i]=jk;
4865: bh[mi][i]=jl; /* bias is positive if real duration
4866: * is higher than the multiple of stepm and negative otherwise.
4867: */
4868: }
4869: else{
4870: dh[mi][i]=jk+1;
4871: bh[mi][i]=ju;
4872: }
4873: if(dh[mi][i]==0){
4874: dh[mi][i]=1; /* At least one step */
4875: bh[mi][i]=ju; /* At least one step */
4876: /* 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);*/
4877: }
4878: } /* end if mle */
1.126 brouard 4879: }
4880: } /* end wave */
4881: }
4882: jmean=sum/k;
4883: 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 4884: 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 4885: }
1.126 brouard 4886:
4887: /*********** Tricode ****************************/
1.220 brouard 4888: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.242 brouard 4889: {
4890: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
4891: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
4892: * Boring subroutine which should only output nbcode[Tvar[j]][k]
4893: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
4894: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
4895: */
1.130 brouard 4896:
1.242 brouard 4897: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
4898: int modmaxcovj=0; /* Modality max of covariates j */
4899: int cptcode=0; /* Modality max of covariates j */
4900: int modmincovj=0; /* Modality min of covariates j */
1.145 brouard 4901:
4902:
1.242 brouard 4903: /* cptcoveff=0; */
4904: /* *cptcov=0; */
1.126 brouard 4905:
1.242 brouard 4906: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 4907:
1.242 brouard 4908: /* Loop on covariates without age and products and no quantitative variable */
4909: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
4910: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
4911: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
4912: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
4913: switch(Fixed[k]) {
4914: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
4915: 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*/
4916: ij=(int)(covar[Tvar[k]][i]);
4917: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
4918: * If product of Vn*Vm, still boolean *:
4919: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
4920: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
4921: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
4922: modality of the nth covariate of individual i. */
4923: if (ij > modmaxcovj)
4924: modmaxcovj=ij;
4925: else if (ij < modmincovj)
4926: modmincovj=ij;
4927: if ((ij < -1) && (ij > NCOVMAX)){
4928: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
4929: exit(1);
4930: }else
4931: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
4932: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
4933: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
4934: /* getting the maximum value of the modality of the covariate
4935: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
4936: female ies 1, then modmaxcovj=1.
4937: */
4938: } /* end for loop on individuals i */
4939: printf(" Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4940: fprintf(ficlog," Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4941: cptcode=modmaxcovj;
4942: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
4943: /*for (i=0; i<=cptcode; i++) {*/
4944: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
4945: printf("Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4946: fprintf(ficlog, "Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4947: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
4948: if( j != -1){
4949: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
4950: covariate for which somebody answered excluding
4951: undefined. Usually 2: 0 and 1. */
4952: }
4953: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
4954: covariate for which somebody answered including
4955: undefined. Usually 3: -1, 0 and 1. */
4956: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
4957: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
4958: } /* Ndum[-1] number of undefined modalities */
1.231 brouard 4959:
1.242 brouard 4960: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
4961: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
4962: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
4963: /* modmincovj=3; modmaxcovj = 7; */
4964: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
4965: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
4966: /* defining two dummy variables: variables V1_1 and V1_2.*/
4967: /* nbcode[Tvar[j]][ij]=k; */
4968: /* nbcode[Tvar[j]][1]=0; */
4969: /* nbcode[Tvar[j]][2]=1; */
4970: /* nbcode[Tvar[j]][3]=2; */
4971: /* To be continued (not working yet). */
4972: ij=0; /* ij is similar to i but can jump over null modalities */
4973: 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*/
4974: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
4975: break;
4976: }
4977: ij++;
4978: 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*/
4979: cptcode = ij; /* New max modality for covar j */
4980: } /* end of loop on modality i=-1 to 1 or more */
4981: break;
4982: case 1: /* Testing on varying covariate, could be simple and
4983: * should look at waves or product of fixed *
4984: * varying. No time to test -1, assuming 0 and 1 only */
4985: ij=0;
4986: for(i=0; i<=1;i++){
4987: nbcode[Tvar[k]][++ij]=i;
4988: }
4989: break;
4990: default:
4991: break;
4992: } /* end switch */
4993: } /* end dummy test */
4994:
4995: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
4996: /* /\*recode from 0 *\/ */
4997: /* k is a modality. If we have model=V1+V1*sex */
4998: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
4999: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
5000: /* } */
5001: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
5002: /* if (ij > ncodemax[j]) { */
5003: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5004: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5005: /* break; */
5006: /* } */
5007: /* } /\* end of loop on modality k *\/ */
5008: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
5009:
5010: for (k=-1; k< maxncov; k++) Ndum[k]=0;
5011: /* Look at fixed dummy (single or product) covariates to check empty modalities */
5012: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
5013: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
5014: 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 */
5015: 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 */
5016: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
5017: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
5018:
5019: ij=0;
5020: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
5021: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
5022: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
5023: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
5024: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
5025: /* If product not in single variable we don't print results */
5026: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
5027: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
5028: 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*/
5029: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
5030: 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 */
5031: if(Fixed[k]!=0)
5032: anyvaryingduminmodel=1;
5033: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
5034: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
5035: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
5036: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
5037: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
5038: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
5039: }
5040: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
5041: /* ij--; */
5042: /* cptcoveff=ij; /\*Number of total covariates*\/ */
5043: *cptcov=ij; /*Number of total real effective covariates: effective
5044: * because they can be excluded from the model and real
5045: * if in the model but excluded because missing values, but how to get k from ij?*/
5046: for(j=ij+1; j<= cptcovt; j++){
5047: Tvaraff[j]=0;
5048: Tmodelind[j]=0;
5049: }
5050: for(j=ntveff+1; j<= cptcovt; j++){
5051: TmodelInvind[j]=0;
5052: }
5053: /* To be sorted */
5054: ;
5055: }
1.126 brouard 5056:
1.145 brouard 5057:
1.126 brouard 5058: /*********** Health Expectancies ****************/
5059:
1.235 brouard 5060: 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 5061:
5062: {
5063: /* Health expectancies, no variances */
1.164 brouard 5064: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 5065: int nhstepma, nstepma; /* Decreasing with age */
5066: double age, agelim, hf;
5067: double ***p3mat;
5068: double eip;
5069:
1.238 brouard 5070: /* pstamp(ficreseij); */
1.126 brouard 5071: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
5072: fprintf(ficreseij,"# Age");
5073: for(i=1; i<=nlstate;i++){
5074: for(j=1; j<=nlstate;j++){
5075: fprintf(ficreseij," e%1d%1d ",i,j);
5076: }
5077: fprintf(ficreseij," e%1d. ",i);
5078: }
5079: fprintf(ficreseij,"\n");
5080:
5081:
5082: if(estepm < stepm){
5083: printf ("Problem %d lower than %d\n",estepm, stepm);
5084: }
5085: else hstepm=estepm;
5086: /* We compute the life expectancy from trapezoids spaced every estepm months
5087: * This is mainly to measure the difference between two models: for example
5088: * if stepm=24 months pijx are given only every 2 years and by summing them
5089: * we are calculating an estimate of the Life Expectancy assuming a linear
5090: * progression in between and thus overestimating or underestimating according
5091: * to the curvature of the survival function. If, for the same date, we
5092: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5093: * to compare the new estimate of Life expectancy with the same linear
5094: * hypothesis. A more precise result, taking into account a more precise
5095: * curvature will be obtained if estepm is as small as stepm. */
5096:
5097: /* For example we decided to compute the life expectancy with the smallest unit */
5098: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5099: nhstepm is the number of hstepm from age to agelim
5100: nstepm is the number of stepm from age to agelin.
5101: Look at hpijx to understand the reason of that which relies in memory size
5102: and note for a fixed period like estepm months */
5103: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5104: survival function given by stepm (the optimization length). Unfortunately it
5105: means that if the survival funtion is printed only each two years of age and if
5106: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5107: results. So we changed our mind and took the option of the best precision.
5108: */
5109: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5110:
5111: agelim=AGESUP;
5112: /* If stepm=6 months */
5113: /* Computed by stepm unit matrices, product of hstepm matrices, stored
5114: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
5115:
5116: /* nhstepm age range expressed in number of stepm */
5117: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5118: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5119: /* if (stepm >= YEARM) hstepm=1;*/
5120: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5121: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5122:
5123: for (age=bage; age<=fage; age ++){
5124: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5125: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5126: /* if (stepm >= YEARM) hstepm=1;*/
5127: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
5128:
5129: /* If stepm=6 months */
5130: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5131: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5132:
1.235 brouard 5133: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 5134:
5135: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5136:
5137: printf("%d|",(int)age);fflush(stdout);
5138: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5139:
5140: /* Computing expectancies */
5141: for(i=1; i<=nlstate;i++)
5142: for(j=1; j<=nlstate;j++)
5143: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5144: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
5145:
5146: /* 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]);*/
5147:
5148: }
5149:
5150: fprintf(ficreseij,"%3.0f",age );
5151: for(i=1; i<=nlstate;i++){
5152: eip=0;
5153: for(j=1; j<=nlstate;j++){
5154: eip +=eij[i][j][(int)age];
5155: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
5156: }
5157: fprintf(ficreseij,"%9.4f", eip );
5158: }
5159: fprintf(ficreseij,"\n");
5160:
5161: }
5162: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5163: printf("\n");
5164: fprintf(ficlog,"\n");
5165:
5166: }
5167:
1.235 brouard 5168: 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 5169:
5170: {
5171: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5172: to initial status i, ei. .
1.126 brouard 5173: */
5174: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5175: int nhstepma, nstepma; /* Decreasing with age */
5176: double age, agelim, hf;
5177: double ***p3matp, ***p3matm, ***varhe;
5178: double **dnewm,**doldm;
5179: double *xp, *xm;
5180: double **gp, **gm;
5181: double ***gradg, ***trgradg;
5182: int theta;
5183:
5184: double eip, vip;
5185:
5186: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5187: xp=vector(1,npar);
5188: xm=vector(1,npar);
5189: dnewm=matrix(1,nlstate*nlstate,1,npar);
5190: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5191:
5192: pstamp(ficresstdeij);
5193: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5194: fprintf(ficresstdeij,"# Age");
5195: for(i=1; i<=nlstate;i++){
5196: for(j=1; j<=nlstate;j++)
5197: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5198: fprintf(ficresstdeij," e%1d. ",i);
5199: }
5200: fprintf(ficresstdeij,"\n");
5201:
5202: pstamp(ficrescveij);
5203: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5204: fprintf(ficrescveij,"# Age");
5205: for(i=1; i<=nlstate;i++)
5206: for(j=1; j<=nlstate;j++){
5207: cptj= (j-1)*nlstate+i;
5208: for(i2=1; i2<=nlstate;i2++)
5209: for(j2=1; j2<=nlstate;j2++){
5210: cptj2= (j2-1)*nlstate+i2;
5211: if(cptj2 <= cptj)
5212: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5213: }
5214: }
5215: fprintf(ficrescveij,"\n");
5216:
5217: if(estepm < stepm){
5218: printf ("Problem %d lower than %d\n",estepm, stepm);
5219: }
5220: else hstepm=estepm;
5221: /* We compute the life expectancy from trapezoids spaced every estepm months
5222: * This is mainly to measure the difference between two models: for example
5223: * if stepm=24 months pijx are given only every 2 years and by summing them
5224: * we are calculating an estimate of the Life Expectancy assuming a linear
5225: * progression in between and thus overestimating or underestimating according
5226: * to the curvature of the survival function. If, for the same date, we
5227: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5228: * to compare the new estimate of Life expectancy with the same linear
5229: * hypothesis. A more precise result, taking into account a more precise
5230: * curvature will be obtained if estepm is as small as stepm. */
5231:
5232: /* For example we decided to compute the life expectancy with the smallest unit */
5233: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5234: nhstepm is the number of hstepm from age to agelim
5235: nstepm is the number of stepm from age to agelin.
5236: Look at hpijx to understand the reason of that which relies in memory size
5237: and note for a fixed period like estepm months */
5238: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5239: survival function given by stepm (the optimization length). Unfortunately it
5240: means that if the survival funtion is printed only each two years of age and if
5241: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5242: results. So we changed our mind and took the option of the best precision.
5243: */
5244: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5245:
5246: /* If stepm=6 months */
5247: /* nhstepm age range expressed in number of stepm */
5248: agelim=AGESUP;
5249: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5250: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5251: /* if (stepm >= YEARM) hstepm=1;*/
5252: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5253:
5254: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5255: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5256: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5257: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5258: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5259: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5260:
5261: for (age=bage; age<=fage; age ++){
5262: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5263: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5264: /* if (stepm >= YEARM) hstepm=1;*/
5265: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5266:
1.126 brouard 5267: /* If stepm=6 months */
5268: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5269: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5270:
5271: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5272:
1.126 brouard 5273: /* Computing Variances of health expectancies */
5274: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5275: decrease memory allocation */
5276: for(theta=1; theta <=npar; theta++){
5277: for(i=1; i<=npar; i++){
1.222 brouard 5278: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5279: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5280: }
1.235 brouard 5281: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5282: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5283:
1.126 brouard 5284: for(j=1; j<= nlstate; j++){
1.222 brouard 5285: for(i=1; i<=nlstate; i++){
5286: for(h=0; h<=nhstepm-1; h++){
5287: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5288: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5289: }
5290: }
1.126 brouard 5291: }
1.218 brouard 5292:
1.126 brouard 5293: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5294: for(h=0; h<=nhstepm-1; h++){
5295: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5296: }
1.126 brouard 5297: }/* End theta */
5298:
5299:
5300: for(h=0; h<=nhstepm-1; h++)
5301: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5302: for(theta=1; theta <=npar; theta++)
5303: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5304:
1.218 brouard 5305:
1.222 brouard 5306: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5307: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5308: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5309:
1.222 brouard 5310: printf("%d|",(int)age);fflush(stdout);
5311: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5312: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5313: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5314: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5315: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5316: for(ij=1;ij<=nlstate*nlstate;ij++)
5317: for(ji=1;ji<=nlstate*nlstate;ji++)
5318: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5319: }
5320: }
1.218 brouard 5321:
1.126 brouard 5322: /* Computing expectancies */
1.235 brouard 5323: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5324: for(i=1; i<=nlstate;i++)
5325: for(j=1; j<=nlstate;j++)
1.222 brouard 5326: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5327: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5328:
1.222 brouard 5329: /* 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 5330:
1.222 brouard 5331: }
1.218 brouard 5332:
1.126 brouard 5333: fprintf(ficresstdeij,"%3.0f",age );
5334: for(i=1; i<=nlstate;i++){
5335: eip=0.;
5336: vip=0.;
5337: for(j=1; j<=nlstate;j++){
1.222 brouard 5338: eip += eij[i][j][(int)age];
5339: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5340: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5341: 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 5342: }
5343: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5344: }
5345: fprintf(ficresstdeij,"\n");
1.218 brouard 5346:
1.126 brouard 5347: fprintf(ficrescveij,"%3.0f",age );
5348: for(i=1; i<=nlstate;i++)
5349: for(j=1; j<=nlstate;j++){
1.222 brouard 5350: cptj= (j-1)*nlstate+i;
5351: for(i2=1; i2<=nlstate;i2++)
5352: for(j2=1; j2<=nlstate;j2++){
5353: cptj2= (j2-1)*nlstate+i2;
5354: if(cptj2 <= cptj)
5355: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5356: }
1.126 brouard 5357: }
5358: fprintf(ficrescveij,"\n");
1.218 brouard 5359:
1.126 brouard 5360: }
5361: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5362: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5363: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5364: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5365: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5366: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5367: printf("\n");
5368: fprintf(ficlog,"\n");
1.218 brouard 5369:
1.126 brouard 5370: free_vector(xm,1,npar);
5371: free_vector(xp,1,npar);
5372: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5373: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5374: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5375: }
1.218 brouard 5376:
1.126 brouard 5377: /************ Variance ******************/
1.235 brouard 5378: 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 5379: {
5380: /* Variance of health expectancies */
5381: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5382: /* double **newm;*/
5383: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5384:
5385: /* int movingaverage(); */
5386: double **dnewm,**doldm;
5387: double **dnewmp,**doldmp;
5388: int i, j, nhstepm, hstepm, h, nstepm ;
5389: int k;
5390: double *xp;
5391: double **gp, **gm; /* for var eij */
5392: double ***gradg, ***trgradg; /*for var eij */
5393: double **gradgp, **trgradgp; /* for var p point j */
5394: double *gpp, *gmp; /* for var p point j */
5395: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5396: double ***p3mat;
5397: double age,agelim, hf;
5398: /* double ***mobaverage; */
5399: int theta;
5400: char digit[4];
5401: char digitp[25];
5402:
5403: char fileresprobmorprev[FILENAMELENGTH];
5404:
5405: if(popbased==1){
5406: if(mobilav!=0)
5407: strcpy(digitp,"-POPULBASED-MOBILAV_");
5408: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5409: }
5410: else
5411: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5412:
1.218 brouard 5413: /* if (mobilav!=0) { */
5414: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5415: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5416: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5417: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5418: /* } */
5419: /* } */
5420:
5421: strcpy(fileresprobmorprev,"PRMORPREV-");
5422: sprintf(digit,"%-d",ij);
5423: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5424: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5425: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5426: strcat(fileresprobmorprev,fileresu);
5427: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5428: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5429: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5430: }
5431: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5432: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5433: pstamp(ficresprobmorprev);
5434: 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 5435: fprintf(ficresprobmorprev,"# Selected quantitative variables and dummies");
5436: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
5437: fprintf(ficresprobmorprev," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
5438: }
5439: for(j=1;j<=cptcoveff;j++)
5440: fprintf(ficresprobmorprev,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(ij,j)]);
5441: fprintf(ficresprobmorprev,"\n");
5442:
1.218 brouard 5443: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5444: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5445: fprintf(ficresprobmorprev," p.%-d SE",j);
5446: for(i=1; i<=nlstate;i++)
5447: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5448: }
5449: fprintf(ficresprobmorprev,"\n");
5450:
5451: fprintf(ficgp,"\n# Routine varevsij");
5452: fprintf(ficgp,"\nunset title \n");
5453: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5454: 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");
5455: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5456: /* } */
5457: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5458: pstamp(ficresvij);
5459: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5460: if(popbased==1)
5461: 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);
5462: else
5463: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5464: fprintf(ficresvij,"# Age");
5465: for(i=1; i<=nlstate;i++)
5466: for(j=1; j<=nlstate;j++)
5467: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5468: fprintf(ficresvij,"\n");
5469:
5470: xp=vector(1,npar);
5471: dnewm=matrix(1,nlstate,1,npar);
5472: doldm=matrix(1,nlstate,1,nlstate);
5473: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5474: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5475:
5476: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5477: gpp=vector(nlstate+1,nlstate+ndeath);
5478: gmp=vector(nlstate+1,nlstate+ndeath);
5479: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5480:
1.218 brouard 5481: if(estepm < stepm){
5482: printf ("Problem %d lower than %d\n",estepm, stepm);
5483: }
5484: else hstepm=estepm;
5485: /* For example we decided to compute the life expectancy with the smallest unit */
5486: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5487: nhstepm is the number of hstepm from age to agelim
5488: nstepm is the number of stepm from age to agelim.
5489: Look at function hpijx to understand why because of memory size limitations,
5490: we decided (b) to get a life expectancy respecting the most precise curvature of the
5491: survival function given by stepm (the optimization length). Unfortunately it
5492: means that if the survival funtion is printed every two years of age and if
5493: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5494: results. So we changed our mind and took the option of the best precision.
5495: */
5496: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5497: agelim = AGESUP;
5498: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5499: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5500: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5501: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5502: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5503: gp=matrix(0,nhstepm,1,nlstate);
5504: gm=matrix(0,nhstepm,1,nlstate);
5505:
5506:
5507: for(theta=1; theta <=npar; theta++){
5508: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5509: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5510: }
5511:
1.242 brouard 5512: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5513:
5514: if (popbased==1) {
5515: if(mobilav ==0){
5516: for(i=1; i<=nlstate;i++)
5517: prlim[i][i]=probs[(int)age][i][ij];
5518: }else{ /* mobilav */
5519: for(i=1; i<=nlstate;i++)
5520: prlim[i][i]=mobaverage[(int)age][i][ij];
5521: }
5522: }
5523:
1.235 brouard 5524: 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 5525: for(j=1; j<= nlstate; j++){
5526: for(h=0; h<=nhstepm; h++){
5527: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5528: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5529: }
5530: }
5531: /* Next for computing probability of death (h=1 means
5532: computed over hstepm matrices product = hstepm*stepm months)
5533: as a weighted average of prlim.
5534: */
5535: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5536: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5537: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5538: }
5539: /* end probability of death */
5540:
5541: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5542: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5543:
1.242 brouard 5544: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nres);
1.218 brouard 5545:
5546: if (popbased==1) {
5547: if(mobilav ==0){
5548: for(i=1; i<=nlstate;i++)
5549: prlim[i][i]=probs[(int)age][i][ij];
5550: }else{ /* mobilav */
5551: for(i=1; i<=nlstate;i++)
5552: prlim[i][i]=mobaverage[(int)age][i][ij];
5553: }
5554: }
5555:
1.235 brouard 5556: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5557:
5558: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5559: for(h=0; h<=nhstepm; h++){
5560: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5561: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5562: }
5563: }
5564: /* This for computing probability of death (h=1 means
5565: computed over hstepm matrices product = hstepm*stepm months)
5566: as a weighted average of prlim.
5567: */
5568: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5569: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5570: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5571: }
5572: /* end probability of death */
5573:
5574: for(j=1; j<= nlstate; j++) /* vareij */
5575: for(h=0; h<=nhstepm; h++){
5576: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5577: }
5578:
5579: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5580: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5581: }
5582:
5583: } /* End theta */
5584:
5585: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5586:
5587: for(h=0; h<=nhstepm; h++) /* veij */
5588: for(j=1; j<=nlstate;j++)
5589: for(theta=1; theta <=npar; theta++)
5590: trgradg[h][j][theta]=gradg[h][theta][j];
5591:
5592: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5593: for(theta=1; theta <=npar; theta++)
5594: trgradgp[j][theta]=gradgp[theta][j];
5595:
5596:
5597: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5598: for(i=1;i<=nlstate;i++)
5599: for(j=1;j<=nlstate;j++)
5600: vareij[i][j][(int)age] =0.;
5601:
5602: for(h=0;h<=nhstepm;h++){
5603: for(k=0;k<=nhstepm;k++){
5604: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5605: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5606: for(i=1;i<=nlstate;i++)
5607: for(j=1;j<=nlstate;j++)
5608: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5609: }
5610: }
5611:
5612: /* pptj */
5613: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5614: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5615: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5616: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5617: varppt[j][i]=doldmp[j][i];
5618: /* end ppptj */
5619: /* x centered again */
5620:
1.242 brouard 5621: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5622:
5623: if (popbased==1) {
5624: if(mobilav ==0){
5625: for(i=1; i<=nlstate;i++)
5626: prlim[i][i]=probs[(int)age][i][ij];
5627: }else{ /* mobilav */
5628: for(i=1; i<=nlstate;i++)
5629: prlim[i][i]=mobaverage[(int)age][i][ij];
5630: }
5631: }
5632:
5633: /* This for computing probability of death (h=1 means
5634: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5635: as a weighted average of prlim.
5636: */
1.235 brouard 5637: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5638: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5639: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5640: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5641: }
5642: /* end probability of death */
5643:
5644: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5645: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5646: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5647: for(i=1; i<=nlstate;i++){
5648: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5649: }
5650: }
5651: fprintf(ficresprobmorprev,"\n");
5652:
5653: fprintf(ficresvij,"%.0f ",age );
5654: for(i=1; i<=nlstate;i++)
5655: for(j=1; j<=nlstate;j++){
5656: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5657: }
5658: fprintf(ficresvij,"\n");
5659: free_matrix(gp,0,nhstepm,1,nlstate);
5660: free_matrix(gm,0,nhstepm,1,nlstate);
5661: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5662: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5663: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5664: } /* End age */
5665: free_vector(gpp,nlstate+1,nlstate+ndeath);
5666: free_vector(gmp,nlstate+1,nlstate+ndeath);
5667: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5668: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5669: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5670: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5671: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5672: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5673: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5674: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5675: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5676: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5677: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5678: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5679: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5680: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5681: 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);
5682: /* 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 5683: */
1.218 brouard 5684: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5685: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5686:
1.218 brouard 5687: free_vector(xp,1,npar);
5688: free_matrix(doldm,1,nlstate,1,nlstate);
5689: free_matrix(dnewm,1,nlstate,1,npar);
5690: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5691: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5692: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5693: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5694: fclose(ficresprobmorprev);
5695: fflush(ficgp);
5696: fflush(fichtm);
5697: } /* end varevsij */
1.126 brouard 5698:
5699: /************ Variance of prevlim ******************/
1.235 brouard 5700: 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 5701: {
1.205 brouard 5702: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5703: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5704:
1.126 brouard 5705: double **dnewm,**doldm;
5706: int i, j, nhstepm, hstepm;
5707: double *xp;
5708: double *gp, *gm;
5709: double **gradg, **trgradg;
1.208 brouard 5710: double **mgm, **mgp;
1.126 brouard 5711: double age,agelim;
5712: int theta;
5713:
5714: pstamp(ficresvpl);
5715: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
1.241 brouard 5716: fprintf(ficresvpl,"# Age ");
5717: if(nresult >=1)
5718: fprintf(ficresvpl," Result# ");
1.126 brouard 5719: for(i=1; i<=nlstate;i++)
5720: fprintf(ficresvpl," %1d-%1d",i,i);
5721: fprintf(ficresvpl,"\n");
5722:
5723: xp=vector(1,npar);
5724: dnewm=matrix(1,nlstate,1,npar);
5725: doldm=matrix(1,nlstate,1,nlstate);
5726:
5727: hstepm=1*YEARM; /* Every year of age */
5728: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5729: agelim = AGESUP;
5730: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5731: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5732: if (stepm >= YEARM) hstepm=1;
5733: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5734: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5735: mgp=matrix(1,npar,1,nlstate);
5736: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5737: gp=vector(1,nlstate);
5738: gm=vector(1,nlstate);
5739:
5740: for(theta=1; theta <=npar; theta++){
5741: for(i=1; i<=npar; i++){ /* Computes gradient */
5742: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5743: }
1.209 brouard 5744: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5745: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5746: else
1.235 brouard 5747: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5748: for(i=1;i<=nlstate;i++){
1.126 brouard 5749: gp[i] = prlim[i][i];
1.208 brouard 5750: mgp[theta][i] = prlim[i][i];
5751: }
1.126 brouard 5752: for(i=1; i<=npar; i++) /* Computes gradient */
5753: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5754: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5755: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5756: else
1.235 brouard 5757: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5758: for(i=1;i<=nlstate;i++){
1.126 brouard 5759: gm[i] = prlim[i][i];
1.208 brouard 5760: mgm[theta][i] = prlim[i][i];
5761: }
1.126 brouard 5762: for(i=1;i<=nlstate;i++)
5763: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5764: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5765: } /* End theta */
5766:
5767: trgradg =matrix(1,nlstate,1,npar);
5768:
5769: for(j=1; j<=nlstate;j++)
5770: for(theta=1; theta <=npar; theta++)
5771: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 5772: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5773: /* printf("\nmgm mgp %d ",(int)age); */
5774: /* for(j=1; j<=nlstate;j++){ */
5775: /* printf(" %d ",j); */
5776: /* for(theta=1; theta <=npar; theta++) */
5777: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
5778: /* printf("\n "); */
5779: /* } */
5780: /* } */
5781: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5782: /* printf("\n gradg %d ",(int)age); */
5783: /* for(j=1; j<=nlstate;j++){ */
5784: /* printf("%d ",j); */
5785: /* for(theta=1; theta <=npar; theta++) */
5786: /* printf("%d %lf ",theta,gradg[theta][j]); */
5787: /* printf("\n "); */
5788: /* } */
5789: /* } */
1.126 brouard 5790:
5791: for(i=1;i<=nlstate;i++)
5792: varpl[i][(int)age] =0.;
1.209 brouard 5793: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 5794: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5795: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
5796: }else{
1.126 brouard 5797: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5798: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 5799: }
1.126 brouard 5800: for(i=1;i<=nlstate;i++)
5801: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
5802:
5803: fprintf(ficresvpl,"%.0f ",age );
1.241 brouard 5804: if(nresult >=1)
5805: fprintf(ficresvpl,"%d ",nres );
1.126 brouard 5806: for(i=1; i<=nlstate;i++)
5807: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
5808: fprintf(ficresvpl,"\n");
5809: free_vector(gp,1,nlstate);
5810: free_vector(gm,1,nlstate);
1.208 brouard 5811: free_matrix(mgm,1,npar,1,nlstate);
5812: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 5813: free_matrix(gradg,1,npar,1,nlstate);
5814: free_matrix(trgradg,1,nlstate,1,npar);
5815: } /* End age */
5816:
5817: free_vector(xp,1,npar);
5818: free_matrix(doldm,1,nlstate,1,npar);
5819: free_matrix(dnewm,1,nlstate,1,nlstate);
5820:
5821: }
5822:
5823: /************ Variance of one-step probabilities ******************/
5824: 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 5825: {
5826: int i, j=0, k1, l1, tj;
5827: int k2, l2, j1, z1;
5828: int k=0, l;
5829: int first=1, first1, first2;
5830: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
5831: double **dnewm,**doldm;
5832: double *xp;
5833: double *gp, *gm;
5834: double **gradg, **trgradg;
5835: double **mu;
5836: double age, cov[NCOVMAX+1];
5837: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
5838: int theta;
5839: char fileresprob[FILENAMELENGTH];
5840: char fileresprobcov[FILENAMELENGTH];
5841: char fileresprobcor[FILENAMELENGTH];
5842: double ***varpij;
5843:
5844: strcpy(fileresprob,"PROB_");
5845: strcat(fileresprob,fileres);
5846: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
5847: printf("Problem with resultfile: %s\n", fileresprob);
5848: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
5849: }
5850: strcpy(fileresprobcov,"PROBCOV_");
5851: strcat(fileresprobcov,fileresu);
5852: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
5853: printf("Problem with resultfile: %s\n", fileresprobcov);
5854: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
5855: }
5856: strcpy(fileresprobcor,"PROBCOR_");
5857: strcat(fileresprobcor,fileresu);
5858: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
5859: printf("Problem with resultfile: %s\n", fileresprobcor);
5860: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
5861: }
5862: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5863: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5864: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5865: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5866: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5867: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5868: pstamp(ficresprob);
5869: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
5870: fprintf(ficresprob,"# Age");
5871: pstamp(ficresprobcov);
5872: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
5873: fprintf(ficresprobcov,"# Age");
5874: pstamp(ficresprobcor);
5875: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
5876: fprintf(ficresprobcor,"# Age");
1.126 brouard 5877:
5878:
1.222 brouard 5879: for(i=1; i<=nlstate;i++)
5880: for(j=1; j<=(nlstate+ndeath);j++){
5881: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
5882: fprintf(ficresprobcov," p%1d-%1d ",i,j);
5883: fprintf(ficresprobcor," p%1d-%1d ",i,j);
5884: }
5885: /* fprintf(ficresprob,"\n");
5886: fprintf(ficresprobcov,"\n");
5887: fprintf(ficresprobcor,"\n");
5888: */
5889: xp=vector(1,npar);
5890: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5891: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5892: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
5893: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
5894: first=1;
5895: fprintf(ficgp,"\n# Routine varprob");
5896: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
5897: fprintf(fichtm,"\n");
5898:
5899: 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);
5900: 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);
5901: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 5902: and drawn. It helps understanding how is the covariance between two incidences.\
5903: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 5904: 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 5905: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
5906: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
5907: standard deviations wide on each axis. <br>\
5908: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
5909: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
5910: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
5911:
1.222 brouard 5912: cov[1]=1;
5913: /* tj=cptcoveff; */
1.225 brouard 5914: tj = (int) pow(2,cptcoveff);
1.222 brouard 5915: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
5916: j1=0;
1.224 brouard 5917: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 5918: if (cptcovn>0) {
5919: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 5920: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5921: fprintf(ficresprob, "**********\n#\n");
5922: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 5923: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5924: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 5925:
1.222 brouard 5926: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 5927: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5928: fprintf(ficgp, "**********\n#\n");
1.220 brouard 5929:
5930:
1.222 brouard 5931: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 5932: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5933: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 5934:
1.222 brouard 5935: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 5936: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5937: fprintf(ficresprobcor, "**********\n#");
5938: if(invalidvarcomb[j1]){
5939: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
5940: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
5941: continue;
5942: }
5943: }
5944: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
5945: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5946: gp=vector(1,(nlstate)*(nlstate+ndeath));
5947: gm=vector(1,(nlstate)*(nlstate+ndeath));
5948: for (age=bage; age<=fage; age ++){
5949: cov[2]=age;
5950: if(nagesqr==1)
5951: cov[3]= age*age;
5952: for (k=1; k<=cptcovn;k++) {
5953: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
5954: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
5955: * 1 1 1 1 1
5956: * 2 2 1 1 1
5957: * 3 1 2 1 1
5958: */
5959: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
5960: }
5961: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
5962: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
5963: for (k=1; k<=cptcovprod;k++)
5964: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 5965:
5966:
1.222 brouard 5967: for(theta=1; theta <=npar; theta++){
5968: for(i=1; i<=npar; i++)
5969: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 5970:
1.222 brouard 5971: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 5972:
1.222 brouard 5973: k=0;
5974: for(i=1; i<= (nlstate); i++){
5975: for(j=1; j<=(nlstate+ndeath);j++){
5976: k=k+1;
5977: gp[k]=pmmij[i][j];
5978: }
5979: }
1.220 brouard 5980:
1.222 brouard 5981: for(i=1; i<=npar; i++)
5982: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 5983:
1.222 brouard 5984: pmij(pmmij,cov,ncovmodel,xp,nlstate);
5985: k=0;
5986: for(i=1; i<=(nlstate); i++){
5987: for(j=1; j<=(nlstate+ndeath);j++){
5988: k=k+1;
5989: gm[k]=pmmij[i][j];
5990: }
5991: }
1.220 brouard 5992:
1.222 brouard 5993: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
5994: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
5995: }
1.126 brouard 5996:
1.222 brouard 5997: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
5998: for(theta=1; theta <=npar; theta++)
5999: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 6000:
1.222 brouard 6001: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
6002: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 6003:
1.222 brouard 6004: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 6005:
1.222 brouard 6006: k=0;
6007: for(i=1; i<=(nlstate); i++){
6008: for(j=1; j<=(nlstate+ndeath);j++){
6009: k=k+1;
6010: mu[k][(int) age]=pmmij[i][j];
6011: }
6012: }
6013: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
6014: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
6015: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 6016:
1.222 brouard 6017: /*printf("\n%d ",(int)age);
6018: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6019: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6020: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6021: }*/
1.220 brouard 6022:
1.222 brouard 6023: fprintf(ficresprob,"\n%d ",(int)age);
6024: fprintf(ficresprobcov,"\n%d ",(int)age);
6025: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 6026:
1.222 brouard 6027: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
6028: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
6029: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6030: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
6031: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
6032: }
6033: i=0;
6034: for (k=1; k<=(nlstate);k++){
6035: for (l=1; l<=(nlstate+ndeath);l++){
6036: i++;
6037: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
6038: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
6039: for (j=1; j<=i;j++){
6040: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
6041: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
6042: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
6043: }
6044: }
6045: }/* end of loop for state */
6046: } /* end of loop for age */
6047: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
6048: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
6049: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6050: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6051:
6052: /* Confidence intervalle of pij */
6053: /*
6054: fprintf(ficgp,"\nunset parametric;unset label");
6055: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
6056: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
6057: 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);
6058: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
6059: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
6060: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
6061: */
6062:
6063: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
6064: first1=1;first2=2;
6065: for (k2=1; k2<=(nlstate);k2++){
6066: for (l2=1; l2<=(nlstate+ndeath);l2++){
6067: if(l2==k2) continue;
6068: j=(k2-1)*(nlstate+ndeath)+l2;
6069: for (k1=1; k1<=(nlstate);k1++){
6070: for (l1=1; l1<=(nlstate+ndeath);l1++){
6071: if(l1==k1) continue;
6072: i=(k1-1)*(nlstate+ndeath)+l1;
6073: if(i<=j) continue;
6074: for (age=bage; age<=fage; age ++){
6075: if ((int)age %5==0){
6076: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
6077: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
6078: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
6079: mu1=mu[i][(int) age]/stepm*YEARM ;
6080: mu2=mu[j][(int) age]/stepm*YEARM;
6081: c12=cv12/sqrt(v1*v2);
6082: /* Computing eigen value of matrix of covariance */
6083: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6084: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6085: if ((lc2 <0) || (lc1 <0) ){
6086: if(first2==1){
6087: first1=0;
6088: 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);
6089: }
6090: 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);
6091: /* lc1=fabs(lc1); */ /* If we want to have them positive */
6092: /* lc2=fabs(lc2); */
6093: }
1.220 brouard 6094:
1.222 brouard 6095: /* Eigen vectors */
6096: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
6097: /*v21=sqrt(1.-v11*v11); *//* error */
6098: v21=(lc1-v1)/cv12*v11;
6099: v12=-v21;
6100: v22=v11;
6101: tnalp=v21/v11;
6102: if(first1==1){
6103: first1=0;
6104: 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);
6105: }
6106: 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);
6107: /*printf(fignu*/
6108: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
6109: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
6110: if(first==1){
6111: first=0;
6112: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
6113: fprintf(ficgp,"\nset parametric;unset label");
6114: 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);
6115: fprintf(ficgp,"\nset ter svg size 640, 480");
6116: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 6117: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 6118: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 6119: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
6120: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6121: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6122: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
6123: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6124: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6125: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6126: 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", \
6127: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6128: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6129: }else{
6130: first=0;
6131: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
6132: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6133: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6134: 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", \
6135: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6136: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6137: }/* if first */
6138: } /* age mod 5 */
6139: } /* end loop age */
6140: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6141: first=1;
6142: } /*l12 */
6143: } /* k12 */
6144: } /*l1 */
6145: }/* k1 */
6146: } /* loop on combination of covariates j1 */
6147: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
6148: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
6149: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6150: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
6151: free_vector(xp,1,npar);
6152: fclose(ficresprob);
6153: fclose(ficresprobcov);
6154: fclose(ficresprobcor);
6155: fflush(ficgp);
6156: fflush(fichtmcov);
6157: }
1.126 brouard 6158:
6159:
6160: /******************* Printing html file ***********/
1.201 brouard 6161: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 6162: int lastpass, int stepm, int weightopt, char model[],\
6163: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.217 brouard 6164: int popforecast, int prevfcast, int backcast, int estepm , \
1.213 brouard 6165: double jprev1, double mprev1,double anprev1, double dateprev1, \
6166: double jprev2, double mprev2,double anprev2, double dateprev2){
1.237 brouard 6167: int jj1, k1, i1, cpt, k4, nres;
1.126 brouard 6168:
6169: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
6170: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
6171: </ul>");
1.237 brouard 6172: fprintf(fichtm,"<ul><li> model=1+age+%s\n \
6173: </ul>", model);
1.214 brouard 6174: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
6175: 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",
6176: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
6177: 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 6178: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
6179: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 6180: fprintf(fichtm,"\
6181: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 6182: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6183: fprintf(fichtm,"\
1.217 brouard 6184: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6185: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6186: fprintf(fichtm,"\
1.126 brouard 6187: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6188: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6189: fprintf(fichtm,"\
1.217 brouard 6190: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6191: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6192: fprintf(fichtm,"\
1.211 brouard 6193: - (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 6194: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6195: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6196: if(prevfcast==1){
6197: fprintf(fichtm,"\
6198: - Prevalence projections by age and states: \
1.201 brouard 6199: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6200: }
1.126 brouard 6201:
1.222 brouard 6202: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 6203:
1.225 brouard 6204: m=pow(2,cptcoveff);
1.222 brouard 6205: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6206:
1.222 brouard 6207: jj1=0;
1.237 brouard 6208:
6209: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.241 brouard 6210: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
1.237 brouard 6211: if(TKresult[nres]!= k1)
6212: continue;
1.220 brouard 6213:
1.222 brouard 6214: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6215: jj1++;
6216: if (cptcovn > 0) {
6217: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6218: for (cpt=1; cpt<=cptcoveff;cpt++){
1.237 brouard 6219: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6220: printf(" V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);fflush(stdout);
6221: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6222: /* printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout); */
1.222 brouard 6223: }
1.237 brouard 6224: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6225: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6226: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);fflush(stdout);
6227: }
6228:
1.230 brouard 6229: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6230: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6231: if(invalidvarcomb[k1]){
6232: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6233: printf("\nCombination (%d) ignored because no cases \n",k1);
6234: continue;
6235: }
6236: }
6237: /* aij, bij */
1.241 brouard 6238: 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> \
6239: <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 6240: /* Pij */
1.241 brouard 6241: 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> \
6242: <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 6243: /* Quasi-incidences */
6244: 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 6245: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6246: 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 6247: 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> \
6248: <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 6249: /* Survival functions (period) in state j */
6250: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6251: 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> \
6252: <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 6253: }
6254: /* State specific survival functions (period) */
6255: for(cpt=1; cpt<=nlstate;cpt++){
6256: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6257: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.241 brouard 6258: <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 6259: }
6260: /* Period (stable) prevalence in each health state */
6261: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6262: fprintf(fichtm,"<br>\n- Convergence 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> \
6263: <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 6264: }
6265: if(backcast==1){
6266: /* Period (stable) back prevalence in each health state */
6267: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6268: fprintf(fichtm,"<br>\n- Convergence to period (stable) back 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> \
6269: <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 6270: }
1.217 brouard 6271: }
1.222 brouard 6272: if(prevfcast==1){
6273: /* Projection of prevalence up to period (stable) prevalence in each health state */
6274: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6275: 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> \
6276: <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 6277: }
6278: }
1.220 brouard 6279:
1.222 brouard 6280: for(cpt=1; cpt<=nlstate;cpt++) {
1.241 brouard 6281: 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> \
6282: <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 6283: }
6284: /* } /\* end i1 *\/ */
6285: }/* End k1 */
6286: fprintf(fichtm,"</ul>");
1.126 brouard 6287:
1.222 brouard 6288: fprintf(fichtm,"\
1.126 brouard 6289: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6290: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6291: - 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 6292: But because parameters are usually highly correlated (a higher incidence of disability \
6293: and a higher incidence of recovery can give very close observed transition) it might \
6294: be very useful to look not only at linear confidence intervals estimated from the \
6295: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6296: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6297: covariance matrix of the one-step probabilities. \
6298: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6299:
1.222 brouard 6300: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6301: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6302: fprintf(fichtm,"\
1.126 brouard 6303: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6304: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6305:
1.222 brouard 6306: fprintf(fichtm,"\
1.126 brouard 6307: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6308: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6309: fprintf(fichtm,"\
1.126 brouard 6310: - 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): \
6311: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6312: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6313: fprintf(fichtm,"\
1.126 brouard 6314: - (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): \
6315: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6316: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6317: fprintf(fichtm,"\
1.128 brouard 6318: - 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 6319: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6320: fprintf(fichtm,"\
1.128 brouard 6321: - 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 6322: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6323: fprintf(fichtm,"\
1.126 brouard 6324: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6325: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6326:
6327: /* if(popforecast==1) fprintf(fichtm,"\n */
6328: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6329: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6330: /* <br>",fileres,fileres,fileres,fileres); */
6331: /* else */
6332: /* 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 6333: fflush(fichtm);
6334: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6335:
1.225 brouard 6336: m=pow(2,cptcoveff);
1.222 brouard 6337: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6338:
1.222 brouard 6339: jj1=0;
1.237 brouard 6340:
1.241 brouard 6341: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.222 brouard 6342: for(k1=1; k1<=m;k1++){
1.237 brouard 6343: if(TKresult[nres]!= k1)
6344: continue;
1.222 brouard 6345: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6346: jj1++;
1.126 brouard 6347: if (cptcovn > 0) {
6348: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6349: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.237 brouard 6350: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);
6351: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6352: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6353: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6354: }
6355:
1.126 brouard 6356: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6357:
1.222 brouard 6358: if(invalidvarcomb[k1]){
6359: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6360: continue;
6361: }
1.126 brouard 6362: }
6363: for(cpt=1; cpt<=nlstate;cpt++) {
1.218 brouard 6364: fprintf(fichtm,"\n<br>- Observed (cross-sectional) and period (incidence based) \
1.241 brouard 6365: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d-%d.svg\"> %s_%d-%d-%d.svg</a>\n <br>\
6366: <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 6367: }
6368: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6369: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6370: true period expectancies (those weighted with period prevalences are also\
6371: drawn in addition to the population based expectancies computed using\
1.241 brouard 6372: observed and cahotic prevalences: <a href=\"%s_%d-%d.svg\">%s_%d-%d.svg</a>\n<br>\
6373: <img src=\"%s_%d-%d.svg\">",subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres);
1.222 brouard 6374: /* } /\* end i1 *\/ */
6375: }/* End k1 */
1.241 brouard 6376: }/* End nres */
1.222 brouard 6377: fprintf(fichtm,"</ul>");
6378: fflush(fichtm);
1.126 brouard 6379: }
6380:
6381: /******************* Gnuplot file **************/
1.223 brouard 6382: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6383:
6384: char dirfileres[132],optfileres[132];
1.223 brouard 6385: char gplotcondition[132];
1.237 brouard 6386: 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 6387: int lv=0, vlv=0, kl=0;
1.130 brouard 6388: int ng=0;
1.201 brouard 6389: int vpopbased;
1.223 brouard 6390: int ioffset; /* variable offset for columns */
1.235 brouard 6391: int nres=0; /* Index of resultline */
1.219 brouard 6392:
1.126 brouard 6393: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6394: /* printf("Problem with file %s",optionfilegnuplot); */
6395: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6396: /* } */
6397:
6398: /*#ifdef windows */
6399: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6400: /*#endif */
1.225 brouard 6401: m=pow(2,cptcoveff);
1.126 brouard 6402:
1.202 brouard 6403: /* Contribution to likelihood */
6404: /* Plot the probability implied in the likelihood */
1.223 brouard 6405: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6406: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6407: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6408: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6409: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6410: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6411: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6412: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6413: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6414: 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));
6415: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6416: 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));
6417: for (i=1; i<= nlstate ; i ++) {
6418: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6419: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6420: 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);
6421: for (j=2; j<= nlstate+ndeath ; j ++) {
6422: 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);
6423: }
6424: fprintf(ficgp,";\nset out; unset ylabel;\n");
6425: }
6426: /* 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 */
6427: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6428: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6429: fprintf(ficgp,"\nset out;unset log\n");
6430: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6431:
1.126 brouard 6432: strcpy(dirfileres,optionfilefiname);
6433: strcpy(optfileres,"vpl");
1.223 brouard 6434: /* 1eme*/
1.238 brouard 6435: for (cpt=1; cpt<= nlstate ; cpt ++){ /* For each live state */
6436: for (k1=1; k1<= m ; k1 ++){ /* For each valid combination of covariate */
1.236 brouard 6437: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.238 brouard 6438: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
6439: if(TKresult[nres]!= k1)
6440: continue;
6441: /* We are interested in selected combination by the resultline */
1.246 brouard 6442: /* printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt); */
1.238 brouard 6443: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
6444: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6445: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
6446: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6447: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6448: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6449: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6450: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.246 brouard 6451: /* printf(" V%d=%d ",Tvaraff[k],vlv); */
1.238 brouard 6452: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6453: }
6454: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.246 brouard 6455: /* printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); */
1.238 brouard 6456: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6457: }
1.246 brouard 6458: /* printf("\n#\n"); */
1.238 brouard 6459: fprintf(ficgp,"\n#\n");
6460: if(invalidvarcomb[k1]){
6461: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6462: continue;
6463: }
1.235 brouard 6464:
1.241 brouard 6465: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
6466: fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres);
6467: 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 6468:
1.238 brouard 6469: for (i=1; i<= nlstate ; i ++) {
6470: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6471: else fprintf(ficgp," %%*lf (%%*lf)");
6472: }
1.242 brouard 6473: 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 6474: for (i=1; i<= nlstate ; i ++) {
6475: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6476: else fprintf(ficgp," %%*lf (%%*lf)");
6477: }
1.242 brouard 6478: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2==%d ? $3-1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres);
1.238 brouard 6479: for (i=1; i<= nlstate ; i ++) {
6480: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6481: else fprintf(ficgp," %%*lf (%%*lf)");
6482: }
6483: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6484: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6485: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
1.242 brouard 6486: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1, nres in 2 to be fixed */
1.238 brouard 6487: if(cptcoveff ==0){
1.245 brouard 6488: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
1.238 brouard 6489: }else{
6490: kl=0;
6491: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6492: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6493: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6494: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6495: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6496: vlv= nbcode[Tvaraff[k]][lv];
1.223 brouard 6497: kl++;
1.238 brouard 6498: /* 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 *\/ */
6499: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6500: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6501: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
6502: if(k==cptcoveff){
1.245 brouard 6503: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' w l lt 3",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
1.242 brouard 6504: 2+cptcoveff*2+(cpt-1), cpt ); /* 4 or 6 ?*/
1.238 brouard 6505: }else{
6506: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6507: kl++;
6508: }
6509: } /* end covariate */
6510: } /* end if no covariate */
6511: } /* end if backcast */
6512: fprintf(ficgp,"\nset out \n");
6513: } /* nres */
1.201 brouard 6514: } /* k1 */
6515: } /* cpt */
1.235 brouard 6516:
6517:
1.126 brouard 6518: /*2 eme*/
1.238 brouard 6519: for (k1=1; k1<= m ; k1 ++){
6520: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6521: if(TKresult[nres]!= k1)
6522: continue;
6523: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
6524: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.225 brouard 6525: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6526: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6527: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6528: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6529: vlv= nbcode[Tvaraff[k]][lv];
6530: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6531: }
1.237 brouard 6532: /* for(k=1; k <= ncovds; k++){ */
1.236 brouard 6533: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 6534: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.236 brouard 6535: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6536: }
1.211 brouard 6537: fprintf(ficgp,"\n#\n");
1.223 brouard 6538: if(invalidvarcomb[k1]){
6539: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6540: continue;
6541: }
1.219 brouard 6542:
1.241 brouard 6543: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1,nres);
1.238 brouard 6544: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6545: if(vpopbased==0)
6546: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6547: else
6548: fprintf(ficgp,"\nreplot ");
6549: for (i=1; i<= nlstate+1 ; i ++) {
6550: k=2*i;
6551: 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);
6552: for (j=1; j<= nlstate+1 ; j ++) {
6553: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6554: else fprintf(ficgp," %%*lf (%%*lf)");
6555: }
6556: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6557: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
6558: 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);
6559: for (j=1; j<= nlstate+1 ; j ++) {
6560: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6561: else fprintf(ficgp," %%*lf (%%*lf)");
6562: }
6563: fprintf(ficgp,"\" t\"\" w l lt 0,");
6564: 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);
6565: for (j=1; j<= nlstate+1 ; j ++) {
6566: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6567: else fprintf(ficgp," %%*lf (%%*lf)");
6568: }
6569: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6570: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6571: } /* state */
6572: } /* vpopbased */
1.244 brouard 6573: fprintf(ficgp,"\nset out;set out \"%s_%d-%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1,nres); /* Buggy gnuplot */
1.238 brouard 6574: } /* end nres */
6575: } /* k1 end 2 eme*/
6576:
6577:
6578: /*3eme*/
6579: for (k1=1; k1<= m ; k1 ++){
6580: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.240 brouard 6581: if(TKresult[nres]!= k1)
1.238 brouard 6582: continue;
6583:
6584: for (cpt=1; cpt<= nlstate ; cpt ++) {
6585: fprintf(ficgp,"\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
6586: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6587: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6588: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6589: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6590: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6591: vlv= nbcode[Tvaraff[k]][lv];
6592: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6593: }
6594: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6595: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6596: }
6597: fprintf(ficgp,"\n#\n");
6598: if(invalidvarcomb[k1]){
6599: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6600: continue;
6601: }
6602:
6603: /* k=2+nlstate*(2*cpt-2); */
6604: k=2+(nlstate+1)*(cpt-1);
1.241 brouard 6605: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.238 brouard 6606: fprintf(ficgp,"set ter svg size 640, 480\n\
1.201 brouard 6607: 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 6608: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6609: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6610: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6611: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6612: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6613: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6614:
1.238 brouard 6615: */
6616: for (i=1; i< nlstate ; i ++) {
6617: 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);
6618: /* 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 6619:
1.238 brouard 6620: }
6621: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+nlstate,cpt);
6622: }
6623: } /* end nres */
6624: } /* end kl 3eme */
1.126 brouard 6625:
1.223 brouard 6626: /* 4eme */
1.201 brouard 6627: /* Survival functions (period) from state i in state j by initial state i */
1.238 brouard 6628: for (k1=1; k1<=m; k1++){ /* For each covariate and each value */
6629: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6630: if(TKresult[nres]!= k1)
1.223 brouard 6631: continue;
1.238 brouard 6632: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state cpt*/
6633: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
6634: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6635: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6636: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6637: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6638: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6639: vlv= nbcode[Tvaraff[k]][lv];
6640: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6641: }
6642: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6643: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6644: }
6645: fprintf(ficgp,"\n#\n");
6646: if(invalidvarcomb[k1]){
6647: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6648: continue;
1.223 brouard 6649: }
1.238 brouard 6650:
1.241 brouard 6651: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.238 brouard 6652: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6653: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6654: k=3;
6655: for (i=1; i<= nlstate ; i ++){
6656: if(i==1){
6657: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6658: }else{
6659: fprintf(ficgp,", '' ");
6660: }
6661: l=(nlstate+ndeath)*(i-1)+1;
6662: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6663: for (j=2; j<= nlstate+ndeath ; j ++)
6664: fprintf(ficgp,"+$%d",k+l+j-1);
6665: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
6666: } /* nlstate */
6667: fprintf(ficgp,"\nset out\n");
6668: } /* end cpt state*/
6669: } /* end nres */
6670: } /* end covariate k1 */
6671:
1.220 brouard 6672: /* 5eme */
1.201 brouard 6673: /* Survival functions (period) from state i in state j by final state j */
1.238 brouard 6674: for (k1=1; k1<= m ; k1++){ /* For each covariate combination if any */
6675: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6676: if(TKresult[nres]!= k1)
1.227 brouard 6677: continue;
1.238 brouard 6678: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
6679: 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);
6680: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6681: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6682: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6683: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6684: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6685: vlv= nbcode[Tvaraff[k]][lv];
6686: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6687: }
6688: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6689: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6690: }
6691: fprintf(ficgp,"\n#\n");
6692: if(invalidvarcomb[k1]){
6693: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6694: continue;
6695: }
1.227 brouard 6696:
1.241 brouard 6697: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.238 brouard 6698: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6699: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6700: k=3;
6701: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6702: if(j==1)
6703: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6704: else
6705: fprintf(ficgp,", '' ");
6706: l=(nlstate+ndeath)*(cpt-1) +j;
6707: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6708: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6709: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6710: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
6711: } /* nlstate */
6712: fprintf(ficgp,", '' ");
6713: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6714: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6715: l=(nlstate+ndeath)*(cpt-1) +j;
6716: if(j < nlstate)
6717: fprintf(ficgp,"$%d +",k+l);
6718: else
6719: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
6720: }
6721: fprintf(ficgp,"\nset out\n");
6722: } /* end cpt state*/
6723: } /* end covariate */
6724: } /* end nres */
1.227 brouard 6725:
1.220 brouard 6726: /* 6eme */
1.202 brouard 6727: /* CV preval stable (period) for each covariate */
1.237 brouard 6728: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6729: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6730: if(TKresult[nres]!= k1)
6731: continue;
1.153 brouard 6732: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6733:
1.211 brouard 6734: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6735: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6736: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
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 (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6744: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6745: }
1.211 brouard 6746: fprintf(ficgp,"\n#\n");
1.223 brouard 6747: if(invalidvarcomb[k1]){
1.227 brouard 6748: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6749: continue;
1.223 brouard 6750: }
1.227 brouard 6751:
1.241 brouard 6752: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.126 brouard 6753: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6754: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6755: k=3; /* Offset */
1.153 brouard 6756: for (i=1; i<= nlstate ; i ++){
1.227 brouard 6757: if(i==1)
6758: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6759: else
6760: fprintf(ficgp,", '' ");
6761: l=(nlstate+ndeath)*(i-1)+1;
6762: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6763: for (j=2; j<= nlstate ; j ++)
6764: fprintf(ficgp,"+$%d",k+l+j-1);
6765: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6766: } /* nlstate */
1.201 brouard 6767: fprintf(ficgp,"\nset out\n");
1.153 brouard 6768: } /* end cpt state*/
6769: } /* end covariate */
1.227 brouard 6770:
6771:
1.220 brouard 6772: /* 7eme */
1.218 brouard 6773: if(backcast == 1){
1.217 brouard 6774: /* CV back preval stable (period) for each covariate */
1.237 brouard 6775: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6776: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6777: if(TKresult[nres]!= k1)
6778: continue;
1.218 brouard 6779: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6780: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
6781: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6782: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6783: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6784: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 6785: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 6786: vlv= nbcode[Tvaraff[k]][lv];
6787: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6788: }
1.237 brouard 6789: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6790: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6791: }
1.227 brouard 6792: fprintf(ficgp,"\n#\n");
6793: if(invalidvarcomb[k1]){
6794: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6795: continue;
6796: }
6797:
1.241 brouard 6798: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.227 brouard 6799: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6800: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6801: k=3; /* Offset */
6802: for (i=1; i<= nlstate ; i ++){
6803: if(i==1)
6804: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
6805: else
6806: fprintf(ficgp,", '' ");
6807: /* l=(nlstate+ndeath)*(i-1)+1; */
6808: l=(nlstate+ndeath)*(cpt-1)+1;
6809: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
6810: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
6811: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+(cpt-1)+i-1); /* a vérifier */
6812: /* for (j=2; j<= nlstate ; j ++) */
6813: /* fprintf(ficgp,"+$%d",k+l+j-1); */
6814: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
6815: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
6816: } /* nlstate */
6817: fprintf(ficgp,"\nset out\n");
1.218 brouard 6818: } /* end cpt state*/
6819: } /* end covariate */
6820: } /* End if backcast */
6821:
1.223 brouard 6822: /* 8eme */
1.218 brouard 6823: if(prevfcast==1){
6824: /* Projection from cross-sectional to stable (period) for each covariate */
6825:
1.237 brouard 6826: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6827: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6828: if(TKresult[nres]!= k1)
6829: continue;
1.211 brouard 6830: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6831: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
6832: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
6833: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6834: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6835: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6836: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6837: vlv= nbcode[Tvaraff[k]][lv];
6838: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6839: }
1.237 brouard 6840: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6841: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6842: }
1.227 brouard 6843: fprintf(ficgp,"\n#\n");
6844: if(invalidvarcomb[k1]){
6845: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6846: continue;
6847: }
6848:
6849: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
1.241 brouard 6850: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.227 brouard 6851: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
1.238 brouard 6852: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6853: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
6854: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6855: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6856: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6857: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6858: if(i==1){
6859: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
6860: }else{
6861: fprintf(ficgp,",\\\n '' ");
6862: }
6863: if(cptcoveff ==0){ /* No covariate */
6864: ioffset=2; /* Age is in 2 */
6865: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6866: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6867: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6868: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6869: fprintf(ficgp," u %d:(", ioffset);
6870: if(i==nlstate+1)
6871: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
6872: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6873: else
6874: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
6875: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6876: }else{ /* more than 2 covariates */
6877: if(cptcoveff ==1){
6878: ioffset=4; /* Age is in 4 */
6879: }else{
6880: ioffset=6; /* Age is in 6 */
6881: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6882: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6883: }
6884: fprintf(ficgp," u %d:(",ioffset);
6885: kl=0;
6886: strcpy(gplotcondition,"(");
6887: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
6888: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
6889: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6890: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6891: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6892: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
6893: kl++;
6894: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
6895: kl++;
6896: if(k <cptcoveff && cptcoveff>1)
6897: sprintf(gplotcondition+strlen(gplotcondition)," && ");
6898: }
6899: strcpy(gplotcondition+strlen(gplotcondition),")");
6900: /* 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 *\/ */
6901: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6902: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6903: /* '' 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*/
6904: if(i==nlstate+1){
6905: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
6906: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6907: }else{
6908: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
6909: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6910: }
6911: } /* end if covariate */
6912: } /* nlstate */
6913: fprintf(ficgp,"\nset out\n");
1.223 brouard 6914: } /* end cpt state*/
6915: } /* end covariate */
6916: } /* End if prevfcast */
1.227 brouard 6917:
6918:
1.238 brouard 6919: /* 9eme writing MLE parameters */
6920: fprintf(ficgp,"\n##############\n#9eme MLE estimated parameters\n#############\n");
1.126 brouard 6921: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 6922: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 6923: for(k=1; k <=(nlstate+ndeath); k++){
6924: if (k != i) {
1.227 brouard 6925: fprintf(ficgp,"# current state %d\n",k);
6926: for(j=1; j <=ncovmodel; j++){
6927: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
6928: jk++;
6929: }
6930: fprintf(ficgp,"\n");
1.126 brouard 6931: }
6932: }
1.223 brouard 6933: }
1.187 brouard 6934: fprintf(ficgp,"##############\n#\n");
1.227 brouard 6935:
1.145 brouard 6936: /*goto avoid;*/
1.238 brouard 6937: /* 10eme Graphics of probabilities or incidences using written MLE parameters */
6938: fprintf(ficgp,"\n##############\n#10eme Graphics of probabilities or incidences\n#############\n");
1.187 brouard 6939: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
6940: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
6941: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
6942: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
6943: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6944: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6945: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6946: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6947: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
6948: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6949: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
6950: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
6951: fprintf(ficgp,"#\n");
1.223 brouard 6952: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
1.238 brouard 6953: fprintf(ficgp,"#Number of graphics: first is logit, 2nd is probabilities, third is incidences per year\n");
1.237 brouard 6954: fprintf(ficgp,"#model=%s \n",model);
1.238 brouard 6955: fprintf(ficgp,"# Type of graphic ng=%d\n",ng);
1.237 brouard 6956: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);/* to be checked */
6957: for(jk=1; jk <=m; jk++) /* For each combination of covariate */
6958: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6959: if(TKresult[nres]!= jk)
6960: continue;
6961: fprintf(ficgp,"# Combination of dummy jk=%d and ",jk);
6962: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6963: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6964: }
6965: fprintf(ficgp,"\n#\n");
1.241 brouard 6966: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng,nres);
1.223 brouard 6967: fprintf(ficgp,"\nset ter svg size 640, 480 ");
6968: if (ng==1){
6969: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
6970: fprintf(ficgp,"\nunset log y");
6971: }else if (ng==2){
6972: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
6973: fprintf(ficgp,"\nset log y");
6974: }else if (ng==3){
6975: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
6976: fprintf(ficgp,"\nset log y");
6977: }else
6978: fprintf(ficgp,"\nunset title ");
6979: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
6980: i=1;
6981: for(k2=1; k2<=nlstate; k2++) {
6982: k3=i;
6983: for(k=1; k<=(nlstate+ndeath); k++) {
6984: if (k != k2){
6985: switch( ng) {
6986: case 1:
6987: if(nagesqr==0)
6988: fprintf(ficgp," p%d+p%d*x",i,i+1);
6989: else /* nagesqr =1 */
6990: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6991: break;
6992: case 2: /* ng=2 */
6993: if(nagesqr==0)
6994: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
6995: else /* nagesqr =1 */
6996: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6997: break;
6998: case 3:
6999: if(nagesqr==0)
7000: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
7001: else /* nagesqr =1 */
7002: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
7003: break;
7004: }
7005: ij=1;/* To be checked else nbcode[0][0] wrong */
1.237 brouard 7006: ijp=1; /* product no age */
7007: /* for(j=3; j <=ncovmodel-nagesqr; j++) { */
7008: for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */
1.223 brouard 7009: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
1.237 brouard 7010: if(j==Tage[ij]) { /* Product by age */
7011: if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */
1.238 brouard 7012: if(DummyV[j]==0){
1.237 brouard 7013: fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);;
7014: }else{ /* quantitative */
7015: fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */
7016: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7017: }
7018: ij++;
7019: }
7020: }else if(j==Tprod[ijp]) { /* */
7021: /* printf("Tprod[%d]=%d, j=%d\n", ij, Tprod[ijp], j); */
7022: if(ijp <=cptcovprod) { /* Product */
1.238 brouard 7023: if(DummyV[Tvard[ijp][1]]==0){/* Vn is dummy */
7024: if(DummyV[Tvard[ijp][2]]==0){/* Vn and Vm are dummy */
1.237 brouard 7025: /* 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)]); */
7026: fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tinvresult[nres][Tvard[ijp][2]]);
7027: }else{ /* Vn is dummy and Vm is quanti */
7028: /* fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],Tqinvresult[nres][Tvard[ijp][2]]); */
7029: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7030: }
7031: }else{ /* Vn*Vm Vn is quanti */
1.238 brouard 7032: if(DummyV[Tvard[ijp][2]]==0){
1.237 brouard 7033: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][2]],Tqinvresult[nres][Tvard[ijp][1]]);
7034: }else{ /* Both quanti */
7035: fprintf(ficgp,"+p%d*%f*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7036: }
7037: }
1.238 brouard 7038: ijp++;
1.237 brouard 7039: }
7040: } else{ /* simple covariate */
7041: /* fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,nbcode[Tvar[j]][codtabm(jk,j)]); /\* Valgrind bug nbcode *\/ */
7042: if(Dummy[j]==0){
7043: fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]); /* */
7044: }else{ /* quantitative */
7045: fprintf(ficgp,"+p%d*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* */
1.223 brouard 7046: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7047: }
1.237 brouard 7048: } /* end simple */
7049: } /* end j */
1.223 brouard 7050: }else{
7051: i=i-ncovmodel;
7052: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
7053: fprintf(ficgp," (1.");
7054: }
1.227 brouard 7055:
1.223 brouard 7056: if(ng != 1){
7057: fprintf(ficgp,")/(1");
1.227 brouard 7058:
1.223 brouard 7059: for(k1=1; k1 <=nlstate; k1++){
7060: if(nagesqr==0)
7061: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
7062: else /* nagesqr =1 */
7063: 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 7064:
1.223 brouard 7065: ij=1;
7066: for(j=3; j <=ncovmodel-nagesqr; j++){
1.237 brouard 7067: if((j-2)==Tage[ij]) { /* Bug valgrind */
7068: if(ij <=cptcovage) { /* Bug valgrind */
1.223 brouard 7069: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
7070: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7071: ij++;
7072: }
7073: }
7074: else
1.225 brouard 7075: 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 7076: }
7077: fprintf(ficgp,")");
7078: }
7079: fprintf(ficgp,")");
7080: if(ng ==2)
7081: fprintf(ficgp," t \"p%d%d\" ", k2,k);
7082: else /* ng= 3 */
7083: fprintf(ficgp," t \"i%d%d\" ", k2,k);
7084: }else{ /* end ng <> 1 */
7085: if( k !=k2) /* logit p11 is hard to draw */
7086: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
7087: }
7088: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
7089: fprintf(ficgp,",");
7090: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
7091: fprintf(ficgp,",");
7092: i=i+ncovmodel;
7093: } /* end k */
7094: } /* end k2 */
7095: fprintf(ficgp,"\n set out\n");
7096: } /* end jk */
7097: } /* end ng */
7098: /* avoid: */
7099: fflush(ficgp);
1.126 brouard 7100: } /* end gnuplot */
7101:
7102:
7103: /*************** Moving average **************/
1.219 brouard 7104: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 7105: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 7106:
1.222 brouard 7107: int i, cpt, cptcod;
7108: int modcovmax =1;
7109: int mobilavrange, mob;
7110: int iage=0;
7111:
7112: double sum=0.;
7113: double age;
7114: double *sumnewp, *sumnewm;
7115: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
7116:
7117:
1.225 brouard 7118: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 7119: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
7120:
7121: sumnewp = vector(1,ncovcombmax);
7122: sumnewm = vector(1,ncovcombmax);
7123: agemingood = vector(1,ncovcombmax);
7124: agemaxgood = vector(1,ncovcombmax);
7125:
7126: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7127: sumnewm[cptcod]=0.;
7128: sumnewp[cptcod]=0.;
7129: agemingood[cptcod]=0;
7130: agemaxgood[cptcod]=0;
7131: }
7132: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
7133:
7134: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
7135: if(mobilav==1) mobilavrange=5; /* default */
7136: else mobilavrange=mobilav;
7137: for (age=bage; age<=fage; age++)
7138: for (i=1; i<=nlstate;i++)
7139: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
7140: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
7141: /* We keep the original values on the extreme ages bage, fage and for
7142: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
7143: we use a 5 terms etc. until the borders are no more concerned.
7144: */
7145: for (mob=3;mob <=mobilavrange;mob=mob+2){
7146: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
7147: for (i=1; i<=nlstate;i++){
7148: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7149: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
7150: for (cpt=1;cpt<=(mob-1)/2;cpt++){
7151: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
7152: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
7153: }
7154: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
7155: }
7156: }
7157: }/* end age */
7158: }/* end mob */
7159: }else
7160: return -1;
7161: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7162: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
7163: if(invalidvarcomb[cptcod]){
7164: printf("\nCombination (%d) ignored because no cases \n",cptcod);
7165: continue;
7166: }
1.219 brouard 7167:
1.222 brouard 7168: agemingood[cptcod]=fage-(mob-1)/2;
7169: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
7170: sumnewm[cptcod]=0.;
7171: for (i=1; i<=nlstate;i++){
7172: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7173: }
7174: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7175: agemingood[cptcod]=age;
7176: }else{ /* bad */
7177: for (i=1; i<=nlstate;i++){
7178: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7179: } /* i */
7180: } /* end bad */
7181: }/* age */
7182: sum=0.;
7183: for (i=1; i<=nlstate;i++){
7184: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7185: }
7186: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7187: 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);
7188: /* for (i=1; i<=nlstate;i++){ */
7189: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7190: /* } /\* i *\/ */
7191: } /* end bad */
7192: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
7193: /* From youngest, finding the oldest wrong */
7194: agemaxgood[cptcod]=bage+(mob-1)/2;
7195: for (age=bage+(mob-1)/2; age<=fage; age++){
7196: sumnewm[cptcod]=0.;
7197: for (i=1; i<=nlstate;i++){
7198: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7199: }
7200: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7201: agemaxgood[cptcod]=age;
7202: }else{ /* bad */
7203: for (i=1; i<=nlstate;i++){
7204: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7205: } /* i */
7206: } /* end bad */
7207: }/* age */
7208: sum=0.;
7209: for (i=1; i<=nlstate;i++){
7210: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7211: }
7212: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7213: 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);
7214: /* for (i=1; i<=nlstate;i++){ */
7215: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7216: /* } /\* i *\/ */
7217: } /* end bad */
7218:
7219: for (age=bage; age<=fage; age++){
1.235 brouard 7220: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 7221: sumnewp[cptcod]=0.;
7222: sumnewm[cptcod]=0.;
7223: for (i=1; i<=nlstate;i++){
7224: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
7225: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7226: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
7227: }
7228: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
7229: }
7230: /* printf("\n"); */
7231: /* } */
7232: /* brutal averaging */
7233: for (i=1; i<=nlstate;i++){
7234: for (age=1; age<=bage; age++){
7235: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7236: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7237: }
7238: for (age=fage; age<=AGESUP; age++){
7239: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7240: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7241: }
7242: } /* end i status */
7243: for (i=nlstate+1; i<=nlstate+ndeath;i++){
7244: for (age=1; age<=AGESUP; age++){
7245: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
7246: mobaverage[(int)age][i][cptcod]=0.;
7247: }
7248: }
7249: }/* end cptcod */
7250: free_vector(sumnewm,1, ncovcombmax);
7251: free_vector(sumnewp,1, ncovcombmax);
7252: free_vector(agemaxgood,1, ncovcombmax);
7253: free_vector(agemingood,1, ncovcombmax);
7254: return 0;
7255: }/* End movingaverage */
1.218 brouard 7256:
1.126 brouard 7257:
7258: /************** Forecasting ******************/
1.235 brouard 7259: 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 7260: /* proj1, year, month, day of starting projection
7261: agemin, agemax range of age
7262: dateprev1 dateprev2 range of dates during which prevalence is computed
7263: anproj2 year of en of projection (same day and month as proj1).
7264: */
1.235 brouard 7265: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7266: double agec; /* generic age */
7267: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7268: double *popeffectif,*popcount;
7269: double ***p3mat;
1.218 brouard 7270: /* double ***mobaverage; */
1.126 brouard 7271: char fileresf[FILENAMELENGTH];
7272:
7273: agelim=AGESUP;
1.211 brouard 7274: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7275: in each health status at the date of interview (if between dateprev1 and dateprev2).
7276: We still use firstpass and lastpass as another selection.
7277: */
1.214 brouard 7278: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7279: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7280:
1.201 brouard 7281: strcpy(fileresf,"F_");
7282: strcat(fileresf,fileresu);
1.126 brouard 7283: if((ficresf=fopen(fileresf,"w"))==NULL) {
7284: printf("Problem with forecast resultfile: %s\n", fileresf);
7285: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7286: }
1.235 brouard 7287: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7288: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7289:
1.225 brouard 7290: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7291:
7292:
7293: stepsize=(int) (stepm+YEARM-1)/YEARM;
7294: if (stepm<=12) stepsize=1;
7295: if(estepm < stepm){
7296: printf ("Problem %d lower than %d\n",estepm, stepm);
7297: }
7298: else hstepm=estepm;
7299:
7300: hstepm=hstepm/stepm;
7301: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7302: fractional in yp1 */
7303: anprojmean=yp;
7304: yp2=modf((yp1*12),&yp);
7305: mprojmean=yp;
7306: yp1=modf((yp2*30.5),&yp);
7307: jprojmean=yp;
7308: if(jprojmean==0) jprojmean=1;
7309: if(mprojmean==0) jprojmean=1;
7310:
1.227 brouard 7311: i1=pow(2,cptcoveff);
1.126 brouard 7312: if (cptcovn < 1){i1=1;}
7313:
7314: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7315:
7316: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7317:
1.126 brouard 7318: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7319: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7320: for(k=1; k<=i1;k++){
7321: if(TKresult[nres]!= k)
7322: continue;
1.227 brouard 7323: if(invalidvarcomb[k]){
7324: printf("\nCombination (%d) projection ignored because no cases \n",k);
7325: continue;
7326: }
7327: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7328: for(j=1;j<=cptcoveff;j++) {
7329: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7330: }
1.235 brouard 7331: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 7332: fprintf(ficresf," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.235 brouard 7333: }
1.227 brouard 7334: fprintf(ficresf," yearproj age");
7335: for(j=1; j<=nlstate+ndeath;j++){
7336: for(i=1; i<=nlstate;i++)
7337: fprintf(ficresf," p%d%d",i,j);
7338: fprintf(ficresf," wp.%d",j);
7339: }
7340: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7341: fprintf(ficresf,"\n");
7342: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7343: for (agec=fage; agec>=(ageminpar-1); agec--){
7344: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7345: nhstepm = nhstepm/hstepm;
7346: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7347: oldm=oldms;savm=savms;
1.235 brouard 7348: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7349:
7350: for (h=0; h<=nhstepm; h++){
7351: if (h*hstepm/YEARM*stepm ==yearp) {
7352: fprintf(ficresf,"\n");
7353: for(j=1;j<=cptcoveff;j++)
7354: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7355: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7356: }
7357: for(j=1; j<=nlstate+ndeath;j++) {
7358: ppij=0.;
7359: for(i=1; i<=nlstate;i++) {
7360: if (mobilav==1)
7361: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7362: else {
7363: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7364: }
7365: if (h*hstepm/YEARM*stepm== yearp) {
7366: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7367: }
7368: } /* end i */
7369: if (h*hstepm/YEARM*stepm==yearp) {
7370: fprintf(ficresf," %.3f", ppij);
7371: }
7372: }/* end j */
7373: } /* end h */
7374: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7375: } /* end agec */
7376: } /* end yearp */
7377: } /* end k */
1.219 brouard 7378:
1.126 brouard 7379: fclose(ficresf);
1.215 brouard 7380: printf("End of Computing forecasting \n");
7381: fprintf(ficlog,"End of Computing forecasting\n");
7382:
1.126 brouard 7383: }
7384:
1.218 brouard 7385: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7386: /* 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 7387: /* /\* back1, year, month, day of starting backection */
7388: /* agemin, agemax range of age */
7389: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7390: /* anback2 year of en of backection (same day and month as back1). */
7391: /* *\/ */
7392: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7393: /* double agec; /\* generic age *\/ */
7394: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7395: /* double *popeffectif,*popcount; */
7396: /* double ***p3mat; */
7397: /* /\* double ***mobaverage; *\/ */
7398: /* char fileresfb[FILENAMELENGTH]; */
7399:
7400: /* agelim=AGESUP; */
7401: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7402: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7403: /* We still use firstpass and lastpass as another selection. */
7404: /* *\/ */
7405: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7406: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7407: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7408:
7409: /* strcpy(fileresfb,"FB_"); */
7410: /* strcat(fileresfb,fileresu); */
7411: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7412: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7413: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7414: /* } */
7415: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7416: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7417:
1.225 brouard 7418: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7419:
7420: /* /\* if (mobilav!=0) { *\/ */
7421: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7422: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7423: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7424: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7425: /* /\* } *\/ */
7426: /* /\* } *\/ */
7427:
7428: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7429: /* if (stepm<=12) stepsize=1; */
7430: /* if(estepm < stepm){ */
7431: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7432: /* } */
7433: /* else hstepm=estepm; */
7434:
7435: /* hstepm=hstepm/stepm; */
7436: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7437: /* fractional in yp1 *\/ */
7438: /* anprojmean=yp; */
7439: /* yp2=modf((yp1*12),&yp); */
7440: /* mprojmean=yp; */
7441: /* yp1=modf((yp2*30.5),&yp); */
7442: /* jprojmean=yp; */
7443: /* if(jprojmean==0) jprojmean=1; */
7444: /* if(mprojmean==0) jprojmean=1; */
7445:
1.225 brouard 7446: /* i1=cptcoveff; */
1.218 brouard 7447: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7448:
1.218 brouard 7449: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7450:
1.218 brouard 7451: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7452:
7453: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7454: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7455: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7456: /* k=k+1; */
7457: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7458: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7459: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7460: /* } */
7461: /* fprintf(ficresfb," yearbproj age"); */
7462: /* for(j=1; j<=nlstate+ndeath;j++){ */
7463: /* for(i=1; i<=nlstate;i++) */
7464: /* fprintf(ficresfb," p%d%d",i,j); */
7465: /* fprintf(ficresfb," p.%d",j); */
7466: /* } */
7467: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7468: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7469: /* fprintf(ficresfb,"\n"); */
7470: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7471: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7472: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7473: /* nhstepm = nhstepm/hstepm; */
7474: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7475: /* oldm=oldms;savm=savms; */
7476: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7477: /* for (h=0; h<=nhstepm; h++){ */
7478: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7479: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7480: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7481: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7482: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7483: /* } */
7484: /* for(j=1; j<=nlstate+ndeath;j++) { */
7485: /* ppij=0.; */
7486: /* for(i=1; i<=nlstate;i++) { */
7487: /* if (mobilav==1) */
7488: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7489: /* else { */
7490: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7491: /* } */
7492: /* if (h*hstepm/YEARM*stepm== yearp) { */
7493: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7494: /* } */
7495: /* } /\* end i *\/ */
7496: /* if (h*hstepm/YEARM*stepm==yearp) { */
7497: /* fprintf(ficresfb," %.3f", ppij); */
7498: /* } */
7499: /* }/\* end j *\/ */
7500: /* } /\* end h *\/ */
7501: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7502: /* } /\* end agec *\/ */
7503: /* } /\* end yearp *\/ */
7504: /* } /\* end cptcod *\/ */
7505: /* } /\* end cptcov *\/ */
7506:
7507: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7508:
7509: /* fclose(ficresfb); */
7510: /* printf("End of Computing Back forecasting \n"); */
7511: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7512:
1.218 brouard 7513: /* } */
1.217 brouard 7514:
1.126 brouard 7515: /************** Forecasting *****not tested NB*************/
1.227 brouard 7516: /* 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 7517:
1.227 brouard 7518: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7519: /* int *popage; */
7520: /* double calagedatem, agelim, kk1, kk2; */
7521: /* double *popeffectif,*popcount; */
7522: /* double ***p3mat,***tabpop,***tabpopprev; */
7523: /* /\* double ***mobaverage; *\/ */
7524: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7525:
1.227 brouard 7526: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7527: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7528: /* agelim=AGESUP; */
7529: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7530:
1.227 brouard 7531: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7532:
7533:
1.227 brouard 7534: /* strcpy(filerespop,"POP_"); */
7535: /* strcat(filerespop,fileresu); */
7536: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7537: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7538: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7539: /* } */
7540: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7541: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7542:
1.227 brouard 7543: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7544:
1.227 brouard 7545: /* /\* if (mobilav!=0) { *\/ */
7546: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7547: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7548: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7549: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7550: /* /\* } *\/ */
7551: /* /\* } *\/ */
1.126 brouard 7552:
1.227 brouard 7553: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7554: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7555:
1.227 brouard 7556: /* agelim=AGESUP; */
1.126 brouard 7557:
1.227 brouard 7558: /* hstepm=1; */
7559: /* hstepm=hstepm/stepm; */
1.218 brouard 7560:
1.227 brouard 7561: /* if (popforecast==1) { */
7562: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7563: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7564: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7565: /* } */
7566: /* popage=ivector(0,AGESUP); */
7567: /* popeffectif=vector(0,AGESUP); */
7568: /* popcount=vector(0,AGESUP); */
1.126 brouard 7569:
1.227 brouard 7570: /* i=1; */
7571: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7572:
1.227 brouard 7573: /* imx=i; */
7574: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7575: /* } */
1.218 brouard 7576:
1.227 brouard 7577: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7578: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7579: /* k=k+1; */
7580: /* fprintf(ficrespop,"\n#******"); */
7581: /* for(j=1;j<=cptcoveff;j++) { */
7582: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7583: /* } */
7584: /* fprintf(ficrespop,"******\n"); */
7585: /* fprintf(ficrespop,"# Age"); */
7586: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7587: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7588:
1.227 brouard 7589: /* for (cpt=0; cpt<=0;cpt++) { */
7590: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7591:
1.227 brouard 7592: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7593: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7594: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7595:
1.227 brouard 7596: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7597: /* oldm=oldms;savm=savms; */
7598: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7599:
1.227 brouard 7600: /* for (h=0; h<=nhstepm; h++){ */
7601: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7602: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7603: /* } */
7604: /* for(j=1; j<=nlstate+ndeath;j++) { */
7605: /* kk1=0.;kk2=0; */
7606: /* for(i=1; i<=nlstate;i++) { */
7607: /* if (mobilav==1) */
7608: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7609: /* else { */
7610: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7611: /* } */
7612: /* } */
7613: /* if (h==(int)(calagedatem+12*cpt)){ */
7614: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7615: /* /\*fprintf(ficrespop," %.3f", kk1); */
7616: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7617: /* } */
7618: /* } */
7619: /* for(i=1; i<=nlstate;i++){ */
7620: /* kk1=0.; */
7621: /* for(j=1; j<=nlstate;j++){ */
7622: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7623: /* } */
7624: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7625: /* } */
1.218 brouard 7626:
1.227 brouard 7627: /* if (h==(int)(calagedatem+12*cpt)) */
7628: /* for(j=1; j<=nlstate;j++) */
7629: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7630: /* } */
7631: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7632: /* } */
7633: /* } */
1.218 brouard 7634:
1.227 brouard 7635: /* /\******\/ */
1.218 brouard 7636:
1.227 brouard 7637: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7638: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7639: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7640: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7641: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7642:
1.227 brouard 7643: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7644: /* oldm=oldms;savm=savms; */
7645: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7646: /* for (h=0; h<=nhstepm; h++){ */
7647: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7648: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7649: /* } */
7650: /* for(j=1; j<=nlstate+ndeath;j++) { */
7651: /* kk1=0.;kk2=0; */
7652: /* for(i=1; i<=nlstate;i++) { */
7653: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7654: /* } */
7655: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7656: /* } */
7657: /* } */
7658: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7659: /* } */
7660: /* } */
7661: /* } */
7662: /* } */
1.218 brouard 7663:
1.227 brouard 7664: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7665:
1.227 brouard 7666: /* if (popforecast==1) { */
7667: /* free_ivector(popage,0,AGESUP); */
7668: /* free_vector(popeffectif,0,AGESUP); */
7669: /* free_vector(popcount,0,AGESUP); */
7670: /* } */
7671: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7672: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7673: /* fclose(ficrespop); */
7674: /* } /\* End of popforecast *\/ */
1.218 brouard 7675:
1.126 brouard 7676: int fileappend(FILE *fichier, char *optionfich)
7677: {
7678: if((fichier=fopen(optionfich,"a"))==NULL) {
7679: printf("Problem with file: %s\n", optionfich);
7680: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7681: return (0);
7682: }
7683: fflush(fichier);
7684: return (1);
7685: }
7686:
7687:
7688: /**************** function prwizard **********************/
7689: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7690: {
7691:
7692: /* Wizard to print covariance matrix template */
7693:
1.164 brouard 7694: char ca[32], cb[32];
7695: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7696: int numlinepar;
7697:
7698: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7699: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7700: for(i=1; i <=nlstate; i++){
7701: jj=0;
7702: for(j=1; j <=nlstate+ndeath; j++){
7703: if(j==i) continue;
7704: jj++;
7705: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7706: printf("%1d%1d",i,j);
7707: fprintf(ficparo,"%1d%1d",i,j);
7708: for(k=1; k<=ncovmodel;k++){
7709: /* printf(" %lf",param[i][j][k]); */
7710: /* fprintf(ficparo," %lf",param[i][j][k]); */
7711: printf(" 0.");
7712: fprintf(ficparo," 0.");
7713: }
7714: printf("\n");
7715: fprintf(ficparo,"\n");
7716: }
7717: }
7718: printf("# Scales (for hessian or gradient estimation)\n");
7719: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7720: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
7721: for(i=1; i <=nlstate; i++){
7722: jj=0;
7723: for(j=1; j <=nlstate+ndeath; j++){
7724: if(j==i) continue;
7725: jj++;
7726: fprintf(ficparo,"%1d%1d",i,j);
7727: printf("%1d%1d",i,j);
7728: fflush(stdout);
7729: for(k=1; k<=ncovmodel;k++){
7730: /* printf(" %le",delti3[i][j][k]); */
7731: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7732: printf(" 0.");
7733: fprintf(ficparo," 0.");
7734: }
7735: numlinepar++;
7736: printf("\n");
7737: fprintf(ficparo,"\n");
7738: }
7739: }
7740: printf("# Covariance matrix\n");
7741: /* # 121 Var(a12)\n\ */
7742: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7743: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7744: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7745: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7746: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7747: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7748: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7749: fflush(stdout);
7750: fprintf(ficparo,"# Covariance matrix\n");
7751: /* # 121 Var(a12)\n\ */
7752: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7753: /* # ...\n\ */
7754: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7755:
7756: for(itimes=1;itimes<=2;itimes++){
7757: jj=0;
7758: for(i=1; i <=nlstate; i++){
7759: for(j=1; j <=nlstate+ndeath; j++){
7760: if(j==i) continue;
7761: for(k=1; k<=ncovmodel;k++){
7762: jj++;
7763: ca[0]= k+'a'-1;ca[1]='\0';
7764: if(itimes==1){
7765: printf("#%1d%1d%d",i,j,k);
7766: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7767: }else{
7768: printf("%1d%1d%d",i,j,k);
7769: fprintf(ficparo,"%1d%1d%d",i,j,k);
7770: /* printf(" %.5le",matcov[i][j]); */
7771: }
7772: ll=0;
7773: for(li=1;li <=nlstate; li++){
7774: for(lj=1;lj <=nlstate+ndeath; lj++){
7775: if(lj==li) continue;
7776: for(lk=1;lk<=ncovmodel;lk++){
7777: ll++;
7778: if(ll<=jj){
7779: cb[0]= lk +'a'-1;cb[1]='\0';
7780: if(ll<jj){
7781: if(itimes==1){
7782: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7783: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7784: }else{
7785: printf(" 0.");
7786: fprintf(ficparo," 0.");
7787: }
7788: }else{
7789: if(itimes==1){
7790: printf(" Var(%s%1d%1d)",ca,i,j);
7791: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
7792: }else{
7793: printf(" 0.");
7794: fprintf(ficparo," 0.");
7795: }
7796: }
7797: }
7798: } /* end lk */
7799: } /* end lj */
7800: } /* end li */
7801: printf("\n");
7802: fprintf(ficparo,"\n");
7803: numlinepar++;
7804: } /* end k*/
7805: } /*end j */
7806: } /* end i */
7807: } /* end itimes */
7808:
7809: } /* end of prwizard */
7810: /******************* Gompertz Likelihood ******************************/
7811: double gompertz(double x[])
7812: {
7813: double A,B,L=0.0,sump=0.,num=0.;
7814: int i,n=0; /* n is the size of the sample */
7815:
1.220 brouard 7816: for (i=1;i<=imx ; i++) {
1.126 brouard 7817: sump=sump+weight[i];
7818: /* sump=sump+1;*/
7819: num=num+1;
7820: }
7821:
7822:
7823: /* for (i=0; i<=imx; i++)
7824: 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]);*/
7825:
7826: for (i=1;i<=imx ; i++)
7827: {
7828: if (cens[i] == 1 && wav[i]>1)
7829: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
7830:
7831: if (cens[i] == 0 && wav[i]>1)
7832: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
7833: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
7834:
7835: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7836: if (wav[i] > 1 ) { /* ??? */
7837: L=L+A*weight[i];
7838: /* 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]);*/
7839: }
7840: }
7841:
7842: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7843:
7844: return -2*L*num/sump;
7845: }
7846:
1.136 brouard 7847: #ifdef GSL
7848: /******************* Gompertz_f Likelihood ******************************/
7849: double gompertz_f(const gsl_vector *v, void *params)
7850: {
7851: double A,B,LL=0.0,sump=0.,num=0.;
7852: double *x= (double *) v->data;
7853: int i,n=0; /* n is the size of the sample */
7854:
7855: for (i=0;i<=imx-1 ; i++) {
7856: sump=sump+weight[i];
7857: /* sump=sump+1;*/
7858: num=num+1;
7859: }
7860:
7861:
7862: /* for (i=0; i<=imx; i++)
7863: 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]);*/
7864: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
7865: for (i=1;i<=imx ; i++)
7866: {
7867: if (cens[i] == 1 && wav[i]>1)
7868: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
7869:
7870: if (cens[i] == 0 && wav[i]>1)
7871: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
7872: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
7873:
7874: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7875: if (wav[i] > 1 ) { /* ??? */
7876: LL=LL+A*weight[i];
7877: /* 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]);*/
7878: }
7879: }
7880:
7881: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7882: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
7883:
7884: return -2*LL*num/sump;
7885: }
7886: #endif
7887:
1.126 brouard 7888: /******************* Printing html file ***********/
1.201 brouard 7889: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 7890: int lastpass, int stepm, int weightopt, char model[],\
7891: int imx, double p[],double **matcov,double agemortsup){
7892: int i,k;
7893:
7894: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
7895: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
7896: for (i=1;i<=2;i++)
7897: 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 7898: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 7899: fprintf(fichtm,"</ul>");
7900:
7901: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
7902:
7903: 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>");
7904:
7905: for (k=agegomp;k<(agemortsup-2);k++)
7906: 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]);
7907:
7908:
7909: fflush(fichtm);
7910: }
7911:
7912: /******************* Gnuplot file **************/
1.201 brouard 7913: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 7914:
7915: char dirfileres[132],optfileres[132];
1.164 brouard 7916:
1.126 brouard 7917: int ng;
7918:
7919:
7920: /*#ifdef windows */
7921: fprintf(ficgp,"cd \"%s\" \n",pathc);
7922: /*#endif */
7923:
7924:
7925: strcpy(dirfileres,optionfilefiname);
7926: strcpy(optfileres,"vpl");
1.199 brouard 7927: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 7928: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 7929: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 7930: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 7931: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
7932:
7933: }
7934:
1.136 brouard 7935: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
7936: {
1.126 brouard 7937:
1.136 brouard 7938: /*-------- data file ----------*/
7939: FILE *fic;
7940: char dummy[]=" ";
1.240 brouard 7941: int i=0, j=0, n=0, iv=0, v;
1.223 brouard 7942: int lstra;
1.136 brouard 7943: int linei, month, year,iout;
7944: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 7945: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 7946: char *stratrunc;
1.223 brouard 7947:
1.240 brouard 7948: DummyV=ivector(1,NCOVMAX); /* 1 to 3 */
7949: FixedV=ivector(1,NCOVMAX); /* 1 to 3 */
1.126 brouard 7950:
1.240 brouard 7951: for(v=1; v <=ncovcol;v++){
7952: DummyV[v]=0;
7953: FixedV[v]=0;
7954: }
7955: for(v=ncovcol+1; v <=ncovcol+nqv;v++){
7956: DummyV[v]=1;
7957: FixedV[v]=0;
7958: }
7959: for(v=ncovcol+nqv+1; v <=ncovcol+nqv+ntv;v++){
7960: DummyV[v]=0;
7961: FixedV[v]=1;
7962: }
7963: for(v=ncovcol+nqv+ntv+1; v <=ncovcol+nqv+ntv+nqtv;v++){
7964: DummyV[v]=1;
7965: FixedV[v]=1;
7966: }
7967: for(v=1; v <=ncovcol+nqv+ntv+nqtv;v++){
7968: printf("Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
7969: 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]);
7970: }
1.126 brouard 7971:
1.136 brouard 7972: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 7973: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
7974: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 7975: }
1.126 brouard 7976:
1.136 brouard 7977: i=1;
7978: linei=0;
7979: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
7980: linei=linei+1;
7981: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
7982: if(line[j] == '\t')
7983: line[j] = ' ';
7984: }
7985: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
7986: ;
7987: };
7988: line[j+1]=0; /* Trims blanks at end of line */
7989: if(line[0]=='#'){
7990: fprintf(ficlog,"Comment line\n%s\n",line);
7991: printf("Comment line\n%s\n",line);
7992: continue;
7993: }
7994: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 7995: strcpy(line, linetmp);
1.223 brouard 7996:
7997: /* Loops on waves */
7998: for (j=maxwav;j>=1;j--){
7999: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.238 brouard 8000: cutv(stra, strb, line, ' ');
8001: if(strb[0]=='.') { /* Missing value */
8002: lval=-1;
8003: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
8004: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
8005: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
8006: 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);
8007: 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);
8008: return 1;
8009: }
8010: }else{
8011: errno=0;
8012: /* what_kind_of_number(strb); */
8013: dval=strtod(strb,&endptr);
8014: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
8015: /* if(strb != endptr && *endptr == '\0') */
8016: /* dval=dlval; */
8017: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8018: if( strb[0]=='\0' || (*endptr != '\0')){
8019: 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);
8020: 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);
8021: return 1;
8022: }
8023: cotqvar[j][iv][i]=dval;
8024: cotvar[j][ntv+iv][i]=dval;
8025: }
8026: strcpy(line,stra);
1.223 brouard 8027: }/* end loop ntqv */
1.225 brouard 8028:
1.223 brouard 8029: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.238 brouard 8030: cutv(stra, strb, line, ' ');
8031: if(strb[0]=='.') { /* Missing value */
8032: lval=-1;
8033: }else{
8034: errno=0;
8035: lval=strtol(strb,&endptr,10);
8036: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8037: if( strb[0]=='\0' || (*endptr != '\0')){
8038: 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);
8039: 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);
8040: return 1;
8041: }
8042: }
8043: if(lval <-1 || lval >1){
8044: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8045: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8046: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8047: For example, for multinomial values like 1, 2 and 3,\n \
8048: build V1=0 V2=0 for the reference value (1),\n \
8049: V1=1 V2=0 for (2) \n \
1.223 brouard 8050: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8051: output of IMaCh is often meaningless.\n \
1.223 brouard 8052: Exiting.\n",lval,linei, i,line,j);
1.238 brouard 8053: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8054: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8055: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8056: For example, for multinomial values like 1, 2 and 3,\n \
8057: build V1=0 V2=0 for the reference value (1),\n \
8058: V1=1 V2=0 for (2) \n \
1.223 brouard 8059: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8060: output of IMaCh is often meaningless.\n \
1.223 brouard 8061: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.238 brouard 8062: return 1;
8063: }
8064: cotvar[j][iv][i]=(double)(lval);
8065: strcpy(line,stra);
1.223 brouard 8066: }/* end loop ntv */
1.225 brouard 8067:
1.223 brouard 8068: /* Statuses at wave */
1.137 brouard 8069: cutv(stra, strb, line, ' ');
1.223 brouard 8070: if(strb[0]=='.') { /* Missing value */
1.238 brouard 8071: lval=-1;
1.136 brouard 8072: }else{
1.238 brouard 8073: errno=0;
8074: lval=strtol(strb,&endptr,10);
8075: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8076: if( strb[0]=='\0' || (*endptr != '\0')){
8077: 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);
8078: 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);
8079: return 1;
8080: }
1.136 brouard 8081: }
1.225 brouard 8082:
1.136 brouard 8083: s[j][i]=lval;
1.225 brouard 8084:
1.223 brouard 8085: /* Date of Interview */
1.136 brouard 8086: strcpy(line,stra);
8087: cutv(stra, strb,line,' ');
1.169 brouard 8088: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8089: }
1.169 brouard 8090: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 8091: month=99;
8092: year=9999;
1.136 brouard 8093: }else{
1.225 brouard 8094: 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);
8095: 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);
8096: return 1;
1.136 brouard 8097: }
8098: anint[j][i]= (double) year;
8099: mint[j][i]= (double)month;
8100: strcpy(line,stra);
1.223 brouard 8101: } /* End loop on waves */
1.225 brouard 8102:
1.223 brouard 8103: /* Date of death */
1.136 brouard 8104: cutv(stra, strb,line,' ');
1.169 brouard 8105: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8106: }
1.169 brouard 8107: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 8108: month=99;
8109: year=9999;
8110: }else{
1.141 brouard 8111: 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 8112: 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);
8113: return 1;
1.136 brouard 8114: }
8115: andc[i]=(double) year;
8116: moisdc[i]=(double) month;
8117: strcpy(line,stra);
8118:
1.223 brouard 8119: /* Date of birth */
1.136 brouard 8120: cutv(stra, strb,line,' ');
1.169 brouard 8121: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8122: }
1.169 brouard 8123: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 8124: month=99;
8125: year=9999;
8126: }else{
1.141 brouard 8127: 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);
8128: 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 8129: return 1;
1.136 brouard 8130: }
8131: if (year==9999) {
1.141 brouard 8132: 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);
8133: 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 8134: return 1;
8135:
1.136 brouard 8136: }
8137: annais[i]=(double)(year);
8138: moisnais[i]=(double)(month);
8139: strcpy(line,stra);
1.225 brouard 8140:
1.223 brouard 8141: /* Sample weight */
1.136 brouard 8142: cutv(stra, strb,line,' ');
8143: errno=0;
8144: dval=strtod(strb,&endptr);
8145: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 8146: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
8147: 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 8148: fflush(ficlog);
8149: return 1;
8150: }
8151: weight[i]=dval;
8152: strcpy(line,stra);
1.225 brouard 8153:
1.223 brouard 8154: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
8155: cutv(stra, strb, line, ' ');
8156: if(strb[0]=='.') { /* Missing value */
1.225 brouard 8157: lval=-1;
1.223 brouard 8158: }else{
1.225 brouard 8159: errno=0;
8160: /* what_kind_of_number(strb); */
8161: dval=strtod(strb,&endptr);
8162: /* if(strb != endptr && *endptr == '\0') */
8163: /* dval=dlval; */
8164: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8165: if( strb[0]=='\0' || (*endptr != '\0')){
8166: 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);
8167: 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);
8168: return 1;
8169: }
8170: coqvar[iv][i]=dval;
1.226 brouard 8171: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 8172: }
8173: strcpy(line,stra);
8174: }/* end loop nqv */
1.136 brouard 8175:
1.223 brouard 8176: /* Covariate values */
1.136 brouard 8177: for (j=ncovcol;j>=1;j--){
8178: cutv(stra, strb,line,' ');
1.223 brouard 8179: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 8180: lval=-1;
1.136 brouard 8181: }else{
1.225 brouard 8182: errno=0;
8183: lval=strtol(strb,&endptr,10);
8184: if( strb[0]=='\0' || (*endptr != '\0')){
8185: 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);
8186: 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);
8187: return 1;
8188: }
1.136 brouard 8189: }
8190: if(lval <-1 || lval >1){
1.225 brouard 8191: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8192: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8193: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8194: For example, for multinomial values like 1, 2 and 3,\n \
8195: build V1=0 V2=0 for the reference value (1),\n \
8196: V1=1 V2=0 for (2) \n \
1.136 brouard 8197: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8198: output of IMaCh is often meaningless.\n \
1.136 brouard 8199: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 8200: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8201: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8202: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8203: For example, for multinomial values like 1, 2 and 3,\n \
8204: build V1=0 V2=0 for the reference value (1),\n \
8205: V1=1 V2=0 for (2) \n \
1.136 brouard 8206: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8207: output of IMaCh is often meaningless.\n \
1.136 brouard 8208: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 8209: return 1;
1.136 brouard 8210: }
8211: covar[j][i]=(double)(lval);
8212: strcpy(line,stra);
8213: }
8214: lstra=strlen(stra);
1.225 brouard 8215:
1.136 brouard 8216: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
8217: stratrunc = &(stra[lstra-9]);
8218: num[i]=atol(stratrunc);
8219: }
8220: else
8221: num[i]=atol(stra);
8222: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
8223: 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;}*/
8224:
8225: i=i+1;
8226: } /* End loop reading data */
1.225 brouard 8227:
1.136 brouard 8228: *imax=i-1; /* Number of individuals */
8229: fclose(fic);
1.225 brouard 8230:
1.136 brouard 8231: return (0);
1.164 brouard 8232: /* endread: */
1.225 brouard 8233: printf("Exiting readdata: ");
8234: fclose(fic);
8235: return (1);
1.223 brouard 8236: }
1.126 brouard 8237:
1.234 brouard 8238: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 8239: char *p1 = *stri, *p2 = *stri;
1.235 brouard 8240: while (*p2 == ' ')
1.234 brouard 8241: p2++;
8242: /* while ((*p1++ = *p2++) !=0) */
8243: /* ; */
8244: /* do */
8245: /* while (*p2 == ' ') */
8246: /* p2++; */
8247: /* while (*p1++ == *p2++); */
8248: *stri=p2;
1.145 brouard 8249: }
8250:
1.235 brouard 8251: int decoderesult ( char resultline[], int nres)
1.230 brouard 8252: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
8253: {
1.235 brouard 8254: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 8255: char resultsav[MAXLINE];
1.234 brouard 8256: int resultmodel[MAXLINE];
8257: int modelresult[MAXLINE];
1.230 brouard 8258: char stra[80], strb[80], strc[80], strd[80],stre[80];
8259:
1.234 brouard 8260: removefirstspace(&resultline);
1.233 brouard 8261: printf("decoderesult:%s\n",resultline);
1.230 brouard 8262:
8263: if (strstr(resultline,"v") !=0){
8264: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
8265: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
8266: return 1;
8267: }
8268: trimbb(resultsav, resultline);
8269: if (strlen(resultsav) >1){
8270: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
8271: }
1.234 brouard 8272: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
8273: 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);
8274: 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);
8275: }
8276: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8277: if(nbocc(resultsav,'=') >1){
8278: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8279: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8280: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8281: }else
8282: cutl(strc,strd,resultsav,'=');
1.230 brouard 8283: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8284:
1.230 brouard 8285: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8286: Tvarsel[k]=atoi(strc);
8287: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8288: /* cptcovsel++; */
8289: if (nbocc(stra,'=') >0)
8290: strcpy(resultsav,stra); /* and analyzes it */
8291: }
1.235 brouard 8292: /* Checking for missing or useless values in comparison of current model needs */
1.236 brouard 8293: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8294: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8295: match=0;
1.236 brouard 8296: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.237 brouard 8297: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */
1.236 brouard 8298: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8299: match=1;
8300: break;
8301: }
8302: }
8303: if(match == 0){
8304: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8305: }
8306: }
8307: }
1.235 brouard 8308: /* Checking for missing or useless values in comparison of current model needs */
8309: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8310: match=0;
1.235 brouard 8311: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8312: if(Typevar[k1]==0){ /* Single */
1.237 brouard 8313: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */
1.235 brouard 8314: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8315: ++match;
8316: }
8317: }
8318: }
8319: if(match == 0){
8320: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8321: }else if(match > 1){
8322: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8323: }
8324: }
1.235 brouard 8325:
1.234 brouard 8326: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8327: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8328: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8329: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8330: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8331: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8332: /* 1 0 0 0 */
8333: /* 2 1 0 0 */
8334: /* 3 0 1 0 */
8335: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8336: /* 5 0 0 1 */
8337: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8338: /* 7 0 1 1 */
8339: /* 8 1 1 1 */
1.237 brouard 8340: /* V(Tvresult)=Tresult V4=1 V3=0 V1=1 Tresult[nres=1][2]=0 */
8341: /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */
8342: /* V5*age V5 known which value for nres? */
8343: /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */
1.235 brouard 8344: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8345: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
1.237 brouard 8346: k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */
1.235 brouard 8347: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8348: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
1.237 brouard 8349: Tresult[nres][k4+1]=Tvalsel[k3];/* Tresult[nres][1]=1(V4=1) Tresult[nres][2]=0(V3=0) */
8350: Tvresult[nres][k4+1]=(int)Tvarsel[k3];/* Tvresult[nres][1]=4 Tvresult[nres][3]=1 */
8351: Tinvresult[nres][(int)Tvarsel[k3]]=Tvalsel[k3]; /* Tinvresult[nres][4]=1 */
1.235 brouard 8352: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8353: k4++;;
8354: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8355: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8356: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
1.237 brouard 8357: Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */
8358: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */
8359: Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */
1.235 brouard 8360: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8361: k4q++;;
8362: }
8363: }
1.234 brouard 8364:
1.235 brouard 8365: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8366: return (0);
8367: }
1.235 brouard 8368:
1.230 brouard 8369: int decodemodel( char model[], int lastobs)
8370: /**< This routine decodes the model and returns:
1.224 brouard 8371: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8372: * - nagesqr = 1 if age*age in the model, otherwise 0.
8373: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8374: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8375: * - cptcovage number of covariates with age*products =2
8376: * - cptcovs number of simple covariates
8377: * - 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
8378: * which is a new column after the 9 (ncovcol) variables.
8379: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8380: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8381: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8382: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8383: */
1.136 brouard 8384: {
1.238 brouard 8385: int i, j, k, ks, v;
1.227 brouard 8386: int j1, k1, k2, k3, k4;
1.136 brouard 8387: char modelsav[80];
1.145 brouard 8388: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8389: char *strpt;
1.136 brouard 8390:
1.145 brouard 8391: /*removespace(model);*/
1.136 brouard 8392: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8393: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8394: if (strstr(model,"AGE") !=0){
1.192 brouard 8395: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8396: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8397: return 1;
8398: }
1.141 brouard 8399: if (strstr(model,"v") !=0){
8400: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8401: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8402: return 1;
8403: }
1.187 brouard 8404: strcpy(modelsav,model);
8405: if ((strpt=strstr(model,"age*age")) !=0){
8406: printf(" strpt=%s, model=%s\n",strpt, model);
8407: if(strpt != model){
1.234 brouard 8408: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8409: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8410: corresponding column of parameters.\n",model);
1.234 brouard 8411: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8412: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8413: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8414: return 1;
1.225 brouard 8415: }
1.187 brouard 8416: nagesqr=1;
8417: if (strstr(model,"+age*age") !=0)
1.234 brouard 8418: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8419: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8420: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8421: else
1.234 brouard 8422: substrchaine(modelsav, model, "age*age");
1.187 brouard 8423: }else
8424: nagesqr=0;
8425: if (strlen(modelsav) >1){
8426: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8427: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8428: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8429: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8430: * cst, age and age*age
8431: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8432: /* including age products which are counted in cptcovage.
8433: * but the covariates which are products must be treated
8434: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8435: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8436: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8437:
8438:
1.187 brouard 8439: /* Design
8440: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8441: * < ncovcol=8 >
8442: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8443: * k= 1 2 3 4 5 6 7 8
8444: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8445: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8446: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8447: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8448: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8449: * Tage[++cptcovage]=k
8450: * if products, new covar are created after ncovcol with k1
8451: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8452: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8453: * 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
8454: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8455: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8456: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8457: * < ncovcol=8 >
8458: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8459: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8460: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8461: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8462: * p Tprod[1]@2={ 6, 5}
8463: *p Tvard[1][1]@4= {7, 8, 5, 6}
8464: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8465: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8466: *How to reorganize?
8467: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8468: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8469: * {2, 1, 4, 8, 5, 6, 3, 7}
8470: * Struct []
8471: */
1.225 brouard 8472:
1.187 brouard 8473: /* This loop fills the array Tvar from the string 'model'.*/
8474: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8475: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8476: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8477: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8478: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8479: /* k=1 Tvar[1]=2 (from V2) */
8480: /* k=5 Tvar[5] */
8481: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8482: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8483: /* } */
1.198 brouard 8484: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8485: /*
8486: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8487: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8488: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8489: }
1.187 brouard 8490: cptcovage=0;
8491: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8492: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8493: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8494: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8495: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8496: /*scanf("%d",i);*/
8497: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8498: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8499: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8500: /* covar is not filled and then is empty */
8501: cptcovprod--;
8502: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8503: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8504: Typevar[k]=1; /* 1 for age product */
8505: cptcovage++; /* Sums the number of covariates which include age as a product */
8506: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8507: /*printf("stre=%s ", stre);*/
8508: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8509: cptcovprod--;
8510: cutl(stre,strb,strc,'V');
8511: Tvar[k]=atoi(stre);
8512: Typevar[k]=1; /* 1 for age product */
8513: cptcovage++;
8514: Tage[cptcovage]=k;
8515: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8516: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8517: cptcovn++;
8518: cptcovprodnoage++;k1++;
8519: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8520: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8521: because this model-covariate is a construction we invent a new column
8522: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8523: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8524: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8525: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8526: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8527: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8528: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8529: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8530: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8531: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8532: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8533: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8534: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8535: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8536: for (i=1; i<=lastobs;i++){
8537: /* Computes the new covariate which is a product of
8538: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8539: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8540: }
8541: } /* End age is not in the model */
8542: } /* End if model includes a product */
8543: else { /* no more sum */
8544: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8545: /* scanf("%d",i);*/
8546: cutl(strd,strc,strb,'V');
8547: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8548: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8549: Tvar[k]=atoi(strd);
8550: Typevar[k]=0; /* 0 for simple covariates */
8551: }
8552: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8553: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8554: scanf("%d",i);*/
1.187 brouard 8555: } /* end of loop + on total covariates */
8556: } /* end if strlen(modelsave == 0) age*age might exist */
8557: } /* end if strlen(model == 0) */
1.136 brouard 8558:
8559: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8560: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8561:
1.136 brouard 8562: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8563: printf("cptcovprod=%d ", cptcovprod);
8564: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8565: scanf("%d ",i);*/
8566:
8567:
1.230 brouard 8568: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8569: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8570: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8571: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8572: k = 1 2 3 4 5 6 7 8 9
8573: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8574: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8575: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8576: Dummy[k] 1 0 0 0 3 1 1 2 3
8577: Tmodelind[combination of covar]=k;
1.225 brouard 8578: */
8579: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8580: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8581: /* 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 8582: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8583: printf("Model=%s\n\
8584: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8585: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8586: 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);
8587: fprintf(ficlog,"Model=%s\n\
8588: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8589: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8590: 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 8591: for(k=1;k<=cptcovt; k++){ Fixed[k]=0; Dummy[k]=0;}
1.234 brouard 8592: 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 */
8593: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 8594: Fixed[k]= 0;
8595: Dummy[k]= 0;
1.225 brouard 8596: ncoveff++;
1.232 brouard 8597: ncovf++;
1.234 brouard 8598: nsd++;
8599: modell[k].maintype= FTYPE;
8600: TvarsD[nsd]=Tvar[k];
8601: TvarsDind[nsd]=k;
8602: TvarF[ncovf]=Tvar[k];
8603: TvarFind[ncovf]=k;
8604: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8605: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8606: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
8607: Fixed[k]= 0;
8608: Dummy[k]= 0;
8609: ncoveff++;
8610: ncovf++;
8611: modell[k].maintype= FTYPE;
8612: TvarF[ncovf]=Tvar[k];
8613: TvarFind[ncovf]=k;
1.230 brouard 8614: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8615: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.240 brouard 8616: }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 8617: Fixed[k]= 0;
8618: Dummy[k]= 1;
1.230 brouard 8619: nqfveff++;
1.234 brouard 8620: modell[k].maintype= FTYPE;
8621: modell[k].subtype= FQ;
8622: nsq++;
8623: TvarsQ[nsq]=Tvar[k];
8624: TvarsQind[nsq]=k;
1.232 brouard 8625: ncovf++;
1.234 brouard 8626: TvarF[ncovf]=Tvar[k];
8627: TvarFind[ncovf]=k;
1.231 brouard 8628: 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 8629: 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 8630: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying dummy variables */
1.227 brouard 8631: Fixed[k]= 1;
8632: Dummy[k]= 0;
1.225 brouard 8633: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 8634: modell[k].maintype= VTYPE;
8635: modell[k].subtype= VD;
8636: nsd++;
8637: TvarsD[nsd]=Tvar[k];
8638: TvarsDind[nsd]=k;
8639: ncovv++; /* Only simple time varying variables */
8640: TvarV[ncovv]=Tvar[k];
1.242 brouard 8641: 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 8642: 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 */
8643: 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 8644: 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);
8645: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8646: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 8647: Fixed[k]= 1;
8648: Dummy[k]= 1;
8649: nqtveff++;
8650: modell[k].maintype= VTYPE;
8651: modell[k].subtype= VQ;
8652: ncovv++; /* Only simple time varying variables */
8653: nsq++;
8654: TvarsQ[nsq]=Tvar[k];
8655: TvarsQind[nsq]=k;
8656: TvarV[ncovv]=Tvar[k];
1.242 brouard 8657: 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 8658: 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 */
8659: 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 8660: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8661: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8662: 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 8663: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8664: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 8665: ncova++;
8666: TvarA[ncova]=Tvar[k];
8667: TvarAind[ncova]=k;
1.231 brouard 8668: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.240 brouard 8669: Fixed[k]= 2;
8670: Dummy[k]= 2;
8671: modell[k].maintype= ATYPE;
8672: modell[k].subtype= APFD;
8673: /* ncoveff++; */
1.227 brouard 8674: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.240 brouard 8675: Fixed[k]= 2;
8676: Dummy[k]= 3;
8677: modell[k].maintype= ATYPE;
8678: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8679: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8680: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.240 brouard 8681: Fixed[k]= 3;
8682: Dummy[k]= 2;
8683: modell[k].maintype= ATYPE;
8684: modell[k].subtype= APVD; /* Product age * varying dummy */
8685: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8686: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8687: Fixed[k]= 3;
8688: Dummy[k]= 3;
8689: modell[k].maintype= ATYPE;
8690: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8691: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8692: }
8693: }else if (Typevar[k] == 2) { /* product without age */
8694: k1=Tposprod[k];
8695: if(Tvard[k1][1] <=ncovcol){
1.240 brouard 8696: if(Tvard[k1][2] <=ncovcol){
8697: Fixed[k]= 1;
8698: Dummy[k]= 0;
8699: modell[k].maintype= FTYPE;
8700: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
8701: ncovf++; /* Fixed variables without age */
8702: TvarF[ncovf]=Tvar[k];
8703: TvarFind[ncovf]=k;
8704: }else if(Tvard[k1][2] <=ncovcol+nqv){
8705: Fixed[k]= 0; /* or 2 ?*/
8706: Dummy[k]= 1;
8707: modell[k].maintype= FTYPE;
8708: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
8709: ncovf++; /* Varying variables without age */
8710: TvarF[ncovf]=Tvar[k];
8711: TvarFind[ncovf]=k;
8712: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8713: Fixed[k]= 1;
8714: Dummy[k]= 0;
8715: modell[k].maintype= VTYPE;
8716: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
8717: ncovv++; /* Varying variables without age */
8718: TvarV[ncovv]=Tvar[k];
8719: TvarVind[ncovv]=k;
8720: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8721: Fixed[k]= 1;
8722: Dummy[k]= 1;
8723: modell[k].maintype= VTYPE;
8724: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
8725: ncovv++; /* Varying variables without age */
8726: TvarV[ncovv]=Tvar[k];
8727: TvarVind[ncovv]=k;
8728: }
1.227 brouard 8729: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.240 brouard 8730: if(Tvard[k1][2] <=ncovcol){
8731: Fixed[k]= 0; /* or 2 ?*/
8732: Dummy[k]= 1;
8733: modell[k].maintype= FTYPE;
8734: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
8735: ncovf++; /* Fixed variables without age */
8736: TvarF[ncovf]=Tvar[k];
8737: TvarFind[ncovf]=k;
8738: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8739: Fixed[k]= 1;
8740: Dummy[k]= 1;
8741: modell[k].maintype= VTYPE;
8742: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
8743: ncovv++; /* Varying variables without age */
8744: TvarV[ncovv]=Tvar[k];
8745: TvarVind[ncovv]=k;
8746: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8747: Fixed[k]= 1;
8748: Dummy[k]= 1;
8749: modell[k].maintype= VTYPE;
8750: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
8751: ncovv++; /* Varying variables without age */
8752: TvarV[ncovv]=Tvar[k];
8753: TvarVind[ncovv]=k;
8754: ncovv++; /* Varying variables without age */
8755: TvarV[ncovv]=Tvar[k];
8756: TvarVind[ncovv]=k;
8757: }
1.227 brouard 8758: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.240 brouard 8759: if(Tvard[k1][2] <=ncovcol){
8760: Fixed[k]= 1;
8761: Dummy[k]= 1;
8762: modell[k].maintype= VTYPE;
8763: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
8764: ncovv++; /* Varying variables without age */
8765: TvarV[ncovv]=Tvar[k];
8766: TvarVind[ncovv]=k;
8767: }else if(Tvard[k1][2] <=ncovcol+nqv){
8768: Fixed[k]= 1;
8769: Dummy[k]= 1;
8770: modell[k].maintype= VTYPE;
8771: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
8772: ncovv++; /* Varying variables without age */
8773: TvarV[ncovv]=Tvar[k];
8774: TvarVind[ncovv]=k;
8775: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8776: Fixed[k]= 1;
8777: Dummy[k]= 0;
8778: modell[k].maintype= VTYPE;
8779: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
8780: ncovv++; /* Varying variables without age */
8781: TvarV[ncovv]=Tvar[k];
8782: TvarVind[ncovv]=k;
8783: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8784: Fixed[k]= 1;
8785: Dummy[k]= 1;
8786: modell[k].maintype= VTYPE;
8787: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
8788: ncovv++; /* Varying variables without age */
8789: TvarV[ncovv]=Tvar[k];
8790: TvarVind[ncovv]=k;
8791: }
1.227 brouard 8792: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8793: if(Tvard[k1][2] <=ncovcol){
8794: Fixed[k]= 1;
8795: Dummy[k]= 1;
8796: modell[k].maintype= VTYPE;
8797: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
8798: ncovv++; /* Varying variables without age */
8799: TvarV[ncovv]=Tvar[k];
8800: TvarVind[ncovv]=k;
8801: }else if(Tvard[k1][2] <=ncovcol+nqv){
8802: Fixed[k]= 1;
8803: Dummy[k]= 1;
8804: modell[k].maintype= VTYPE;
8805: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
8806: ncovv++; /* Varying variables without age */
8807: TvarV[ncovv]=Tvar[k];
8808: TvarVind[ncovv]=k;
8809: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8810: Fixed[k]= 1;
8811: Dummy[k]= 1;
8812: modell[k].maintype= VTYPE;
8813: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
8814: ncovv++; /* Varying variables without age */
8815: TvarV[ncovv]=Tvar[k];
8816: TvarVind[ncovv]=k;
8817: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8818: Fixed[k]= 1;
8819: Dummy[k]= 1;
8820: modell[k].maintype= VTYPE;
8821: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
8822: ncovv++; /* Varying variables without age */
8823: TvarV[ncovv]=Tvar[k];
8824: TvarVind[ncovv]=k;
8825: }
1.227 brouard 8826: }else{
1.240 brouard 8827: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8828: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8829: } /*end k1*/
1.225 brouard 8830: }else{
1.226 brouard 8831: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
8832: 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 8833: }
1.227 brouard 8834: 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 8835: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 8836: 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]);
8837: }
8838: /* Searching for doublons in the model */
8839: for(k1=1; k1<= cptcovt;k1++){
8840: for(k2=1; k2 <k1;k2++){
8841: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 8842: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
8843: if(Tvar[k1]==Tvar[k2]){
8844: 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]]);
8845: 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);
8846: return(1);
8847: }
8848: }else if (Typevar[k1] ==2){
8849: k3=Tposprod[k1];
8850: k4=Tposprod[k2];
8851: 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])) ){
8852: 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]]);
8853: 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);
8854: return(1);
8855: }
8856: }
1.227 brouard 8857: }
8858: }
1.225 brouard 8859: }
8860: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
8861: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 8862: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
8863: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 8864: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 8865: /*endread:*/
1.225 brouard 8866: printf("Exiting decodemodel: ");
8867: return (1);
1.136 brouard 8868: }
8869:
1.169 brouard 8870: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.248 brouard 8871: {/* Check ages at death */
1.136 brouard 8872: int i, m;
1.218 brouard 8873: int firstone=0;
8874:
1.136 brouard 8875: for (i=1; i<=imx; i++) {
8876: for(m=2; (m<= maxwav); m++) {
8877: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
8878: anint[m][i]=9999;
1.216 brouard 8879: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
8880: s[m][i]=-1;
1.136 brouard 8881: }
8882: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.169 brouard 8883: *nberr = *nberr + 1;
1.218 brouard 8884: if(firstone == 0){
8885: firstone=1;
8886: 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);
8887: }
8888: 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 8889: s[m][i]=-1;
8890: }
8891: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 8892: (*nberr)++;
1.136 brouard 8893: 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]);
8894: 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]);
8895: s[m][i]=-1; /* We prefer to skip it (and to skip it in version 0.8a1 too */
8896: }
8897: }
8898: }
8899:
8900: for (i=1; i<=imx; i++) {
8901: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
8902: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 8903: 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 8904: if (s[m][i] >= nlstate+1) {
1.169 brouard 8905: if(agedc[i]>0){
8906: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 8907: agev[m][i]=agedc[i];
1.214 brouard 8908: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 8909: }else {
1.136 brouard 8910: if ((int)andc[i]!=9999){
8911: nbwarn++;
8912: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
8913: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
8914: agev[m][i]=-1;
8915: }
8916: }
1.169 brouard 8917: } /* agedc > 0 */
1.214 brouard 8918: } /* end if */
1.136 brouard 8919: else if(s[m][i] !=9){ /* Standard case, age in fractional
8920: years but with the precision of a month */
8921: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
8922: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
8923: agev[m][i]=1;
8924: else if(agev[m][i] < *agemin){
8925: *agemin=agev[m][i];
8926: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
8927: }
8928: else if(agev[m][i] >*agemax){
8929: *agemax=agev[m][i];
1.156 brouard 8930: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 8931: }
8932: /*agev[m][i]=anint[m][i]-annais[i];*/
8933: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 8934: } /* en if 9*/
1.136 brouard 8935: else { /* =9 */
1.214 brouard 8936: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 8937: agev[m][i]=1;
8938: s[m][i]=-1;
8939: }
8940: }
1.214 brouard 8941: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 8942: agev[m][i]=1;
1.214 brouard 8943: else{
8944: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8945: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8946: agev[m][i]=0;
8947: }
8948: } /* End for lastpass */
8949: }
1.136 brouard 8950:
8951: for (i=1; i<=imx; i++) {
8952: for(m=firstpass; (m<=lastpass); m++){
8953: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 8954: (*nberr)++;
1.136 brouard 8955: 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);
8956: 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);
8957: return 1;
8958: }
8959: }
8960: }
8961:
8962: /*for (i=1; i<=imx; i++){
8963: for (m=firstpass; (m<lastpass); m++){
8964: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
8965: }
8966:
8967: }*/
8968:
8969:
1.139 brouard 8970: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
8971: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 8972:
8973: return (0);
1.164 brouard 8974: /* endread:*/
1.136 brouard 8975: printf("Exiting calandcheckages: ");
8976: return (1);
8977: }
8978:
1.172 brouard 8979: #if defined(_MSC_VER)
8980: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8981: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8982: //#include "stdafx.h"
8983: //#include <stdio.h>
8984: //#include <tchar.h>
8985: //#include <windows.h>
8986: //#include <iostream>
8987: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
8988:
8989: LPFN_ISWOW64PROCESS fnIsWow64Process;
8990:
8991: BOOL IsWow64()
8992: {
8993: BOOL bIsWow64 = FALSE;
8994:
8995: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
8996: // (HANDLE, PBOOL);
8997:
8998: //LPFN_ISWOW64PROCESS fnIsWow64Process;
8999:
9000: HMODULE module = GetModuleHandle(_T("kernel32"));
9001: const char funcName[] = "IsWow64Process";
9002: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
9003: GetProcAddress(module, funcName);
9004:
9005: if (NULL != fnIsWow64Process)
9006: {
9007: if (!fnIsWow64Process(GetCurrentProcess(),
9008: &bIsWow64))
9009: //throw std::exception("Unknown error");
9010: printf("Unknown error\n");
9011: }
9012: return bIsWow64 != FALSE;
9013: }
9014: #endif
1.177 brouard 9015:
1.191 brouard 9016: void syscompilerinfo(int logged)
1.167 brouard 9017: {
9018: /* #include "syscompilerinfo.h"*/
1.185 brouard 9019: /* command line Intel compiler 32bit windows, XP compatible:*/
9020: /* /GS /W3 /Gy
9021: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
9022: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
9023: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 9024: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
9025: */
9026: /* 64 bits */
1.185 brouard 9027: /*
9028: /GS /W3 /Gy
9029: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
9030: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
9031: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
9032: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
9033: /* Optimization are useless and O3 is slower than O2 */
9034: /*
9035: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
9036: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
9037: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
9038: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
9039: */
1.186 brouard 9040: /* Link is */ /* /OUT:"visual studio
1.185 brouard 9041: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
9042: /PDB:"visual studio
9043: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
9044: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
9045: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
9046: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
9047: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
9048: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
9049: uiAccess='false'"
9050: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
9051: /NOLOGO /TLBID:1
9052: */
1.177 brouard 9053: #if defined __INTEL_COMPILER
1.178 brouard 9054: #if defined(__GNUC__)
9055: struct utsname sysInfo; /* For Intel on Linux and OS/X */
9056: #endif
1.177 brouard 9057: #elif defined(__GNUC__)
1.179 brouard 9058: #ifndef __APPLE__
1.174 brouard 9059: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 9060: #endif
1.177 brouard 9061: struct utsname sysInfo;
1.178 brouard 9062: int cross = CROSS;
9063: if (cross){
9064: printf("Cross-");
1.191 brouard 9065: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 9066: }
1.174 brouard 9067: #endif
9068:
1.171 brouard 9069: #include <stdint.h>
1.178 brouard 9070:
1.191 brouard 9071: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 9072: #if defined(__clang__)
1.191 brouard 9073: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 9074: #endif
9075: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 9076: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 9077: #endif
9078: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 9079: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 9080: #endif
9081: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 9082: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 9083: #endif
9084: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 9085: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 9086: #endif
9087: #if defined(_MSC_VER)
1.191 brouard 9088: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 9089: #endif
9090: #if defined(__PGI)
1.191 brouard 9091: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 9092: #endif
9093: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 9094: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 9095: #endif
1.191 brouard 9096: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 9097:
1.167 brouard 9098: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
9099: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
9100: // Windows (x64 and x86)
1.191 brouard 9101: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 9102: #elif __unix__ // all unices, not all compilers
9103: // Unix
1.191 brouard 9104: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 9105: #elif __linux__
9106: // linux
1.191 brouard 9107: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 9108: #elif __APPLE__
1.174 brouard 9109: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 9110: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 9111: #endif
9112:
9113: /* __MINGW32__ */
9114: /* __CYGWIN__ */
9115: /* __MINGW64__ */
9116: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
9117: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
9118: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
9119: /* _WIN64 // Defined for applications for Win64. */
9120: /* _M_X64 // Defined for compilations that target x64 processors. */
9121: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 9122:
1.167 brouard 9123: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 9124: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 9125: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 9126: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 9127: #else
1.191 brouard 9128: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 9129: #endif
9130:
1.169 brouard 9131: #if defined(__GNUC__)
9132: # if defined(__GNUC_PATCHLEVEL__)
9133: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9134: + __GNUC_MINOR__ * 100 \
9135: + __GNUC_PATCHLEVEL__)
9136: # else
9137: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9138: + __GNUC_MINOR__ * 100)
9139: # endif
1.174 brouard 9140: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 9141: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 9142:
9143: if (uname(&sysInfo) != -1) {
9144: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 9145: 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 9146: }
9147: else
9148: perror("uname() error");
1.179 brouard 9149: //#ifndef __INTEL_COMPILER
9150: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 9151: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 9152: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 9153: #endif
1.169 brouard 9154: #endif
1.172 brouard 9155:
9156: // void main()
9157: // {
1.169 brouard 9158: #if defined(_MSC_VER)
1.174 brouard 9159: if (IsWow64()){
1.191 brouard 9160: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
9161: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 9162: }
9163: else{
1.191 brouard 9164: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
9165: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 9166: }
1.172 brouard 9167: // printf("\nPress Enter to continue...");
9168: // getchar();
9169: // }
9170:
1.169 brouard 9171: #endif
9172:
1.167 brouard 9173:
1.219 brouard 9174: }
1.136 brouard 9175:
1.219 brouard 9176: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 9177: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 9178: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 9179: /* double ftolpl = 1.e-10; */
1.180 brouard 9180: double age, agebase, agelim;
1.203 brouard 9181: double tot;
1.180 brouard 9182:
1.202 brouard 9183: strcpy(filerespl,"PL_");
9184: strcat(filerespl,fileresu);
9185: if((ficrespl=fopen(filerespl,"w"))==NULL) {
9186: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9187: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9188: }
1.227 brouard 9189: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
9190: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 9191: pstamp(ficrespl);
1.203 brouard 9192: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 9193: fprintf(ficrespl,"#Age ");
9194: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
9195: fprintf(ficrespl,"\n");
1.180 brouard 9196:
1.219 brouard 9197: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 9198:
1.219 brouard 9199: agebase=ageminpar;
9200: agelim=agemaxpar;
1.180 brouard 9201:
1.227 brouard 9202: /* i1=pow(2,ncoveff); */
1.234 brouard 9203: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 9204: if (cptcovn < 1){i1=1;}
1.180 brouard 9205:
1.238 brouard 9206: for(k=1; k<=i1;k++){ /* For each combination k of dummy covariates in the model */
9207: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9208: if(TKresult[nres]!= k)
9209: continue;
1.235 brouard 9210:
1.238 brouard 9211: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9212: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
9213: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
9214: /* k=k+1; */
9215: /* to clean */
9216: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9217: fprintf(ficrespl,"#******");
9218: printf("#******");
9219: fprintf(ficlog,"#******");
9220: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9221: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
9222: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9223: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9224: }
9225: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9226: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9227: fprintf(ficrespl," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9228: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9229: }
9230: fprintf(ficrespl,"******\n");
9231: printf("******\n");
9232: fprintf(ficlog,"******\n");
9233: if(invalidvarcomb[k]){
9234: printf("\nCombination (%d) ignored because no case \n",k);
9235: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
9236: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
9237: continue;
9238: }
1.219 brouard 9239:
1.238 brouard 9240: fprintf(ficrespl,"#Age ");
9241: for(j=1;j<=cptcoveff;j++) {
9242: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9243: }
9244: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
9245: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 9246:
1.238 brouard 9247: for (age=agebase; age<=agelim; age++){
9248: /* for (age=agebase; age<=agebase; age++){ */
9249: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
9250: fprintf(ficrespl,"%.0f ",age );
9251: for(j=1;j<=cptcoveff;j++)
9252: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9253: tot=0.;
9254: for(i=1; i<=nlstate;i++){
9255: tot += prlim[i][i];
9256: fprintf(ficrespl," %.5f", prlim[i][i]);
9257: }
9258: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
9259: } /* Age */
9260: /* was end of cptcod */
9261: } /* cptcov */
9262: } /* nres */
1.219 brouard 9263: return 0;
1.180 brouard 9264: }
9265:
1.218 brouard 9266: 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){
9267: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
9268:
9269: /* Computes the back prevalence limit for any combination of covariate values
9270: * at any age between ageminpar and agemaxpar
9271: */
1.235 brouard 9272: int i, j, k, i1, nres=0 ;
1.217 brouard 9273: /* double ftolpl = 1.e-10; */
9274: double age, agebase, agelim;
9275: double tot;
1.218 brouard 9276: /* double ***mobaverage; */
9277: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 9278:
9279: strcpy(fileresplb,"PLB_");
9280: strcat(fileresplb,fileresu);
9281: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
9282: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9283: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9284: }
9285: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9286: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9287: pstamp(ficresplb);
9288: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9289: fprintf(ficresplb,"#Age ");
9290: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9291: fprintf(ficresplb,"\n");
9292:
1.218 brouard 9293:
9294: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9295:
9296: agebase=ageminpar;
9297: agelim=agemaxpar;
9298:
9299:
1.227 brouard 9300: i1=pow(2,cptcoveff);
1.218 brouard 9301: if (cptcovn < 1){i1=1;}
1.227 brouard 9302:
1.238 brouard 9303: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9304: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
9305: if(TKresult[nres]!= k)
9306: continue;
9307: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9308: fprintf(ficresplb,"#******");
9309: printf("#******");
9310: fprintf(ficlog,"#******");
9311: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9312: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9313: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9314: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9315: }
9316: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9317: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9318: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9319: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9320: }
9321: fprintf(ficresplb,"******\n");
9322: printf("******\n");
9323: fprintf(ficlog,"******\n");
9324: if(invalidvarcomb[k]){
9325: printf("\nCombination (%d) ignored because no cases \n",k);
9326: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9327: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9328: continue;
9329: }
1.218 brouard 9330:
1.238 brouard 9331: fprintf(ficresplb,"#Age ");
9332: for(j=1;j<=cptcoveff;j++) {
9333: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9334: }
9335: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9336: fprintf(ficresplb,"Total Years_to_converge\n");
1.218 brouard 9337:
9338:
1.238 brouard 9339: for (age=agebase; age<=agelim; age++){
9340: /* for (age=agebase; age<=agebase; age++){ */
9341: if(mobilavproj > 0){
9342: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9343: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9344: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k, nres);
1.238 brouard 9345: }else if (mobilavproj == 0){
9346: 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);
9347: 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);
9348: exit(1);
9349: }else{
9350: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9351: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k,nres);
1.238 brouard 9352: }
9353: fprintf(ficresplb,"%.0f ",age );
9354: for(j=1;j<=cptcoveff;j++)
9355: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9356: tot=0.;
9357: for(i=1; i<=nlstate;i++){
9358: tot += bprlim[i][i];
9359: fprintf(ficresplb," %.5f", bprlim[i][i]);
9360: }
9361: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9362: } /* Age */
9363: /* was end of cptcod */
9364: } /* end of any combination */
9365: } /* end of nres */
1.218 brouard 9366: /* hBijx(p, bage, fage); */
9367: /* fclose(ficrespijb); */
9368:
9369: return 0;
1.217 brouard 9370: }
1.218 brouard 9371:
1.180 brouard 9372: int hPijx(double *p, int bage, int fage){
9373: /*------------- h Pij x at various ages ------------*/
9374:
9375: int stepsize;
9376: int agelim;
9377: int hstepm;
9378: int nhstepm;
1.235 brouard 9379: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9380:
9381: double agedeb;
9382: double ***p3mat;
9383:
1.201 brouard 9384: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9385: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9386: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9387: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9388: }
9389: printf("Computing pij: result on file '%s' \n", filerespij);
9390: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9391:
9392: stepsize=(int) (stepm+YEARM-1)/YEARM;
9393: /*if (stepm<=24) stepsize=2;*/
9394:
9395: agelim=AGESUP;
9396: hstepm=stepsize*YEARM; /* Every year of age */
9397: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9398:
1.180 brouard 9399: /* hstepm=1; aff par mois*/
9400: pstamp(ficrespij);
9401: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9402: i1= pow(2,cptcoveff);
1.218 brouard 9403: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9404: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9405: /* k=k+1; */
1.235 brouard 9406: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9407: for(k=1; k<=i1;k++){
9408: if(TKresult[nres]!= k)
9409: continue;
1.183 brouard 9410: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9411: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9412: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9413: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9414: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9415: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9416: }
1.183 brouard 9417: fprintf(ficrespij,"******\n");
9418:
9419: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9420: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9421: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9422:
9423: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9424:
1.183 brouard 9425: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9426: oldm=oldms;savm=savms;
1.235 brouard 9427: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9428: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9429: for(i=1; i<=nlstate;i++)
9430: for(j=1; j<=nlstate+ndeath;j++)
9431: fprintf(ficrespij," %1d-%1d",i,j);
9432: fprintf(ficrespij,"\n");
9433: for (h=0; h<=nhstepm; h++){
9434: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9435: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9436: for(i=1; i<=nlstate;i++)
9437: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9438: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9439: fprintf(ficrespij,"\n");
9440: }
1.183 brouard 9441: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9442: fprintf(ficrespij,"\n");
9443: }
1.180 brouard 9444: /*}*/
9445: }
1.218 brouard 9446: return 0;
1.180 brouard 9447: }
1.218 brouard 9448:
9449: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9450: /*------------- h Bij x at various ages ------------*/
9451:
9452: int stepsize;
1.218 brouard 9453: /* int agelim; */
9454: int ageminl;
1.217 brouard 9455: int hstepm;
9456: int nhstepm;
1.238 brouard 9457: int h, i, i1, j, k, nres;
1.218 brouard 9458:
1.217 brouard 9459: double agedeb;
9460: double ***p3mat;
1.218 brouard 9461:
9462: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9463: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9464: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9465: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9466: }
9467: printf("Computing pij back: result on file '%s' \n", filerespijb);
9468: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9469:
9470: stepsize=(int) (stepm+YEARM-1)/YEARM;
9471: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9472:
1.218 brouard 9473: /* agelim=AGESUP; */
9474: ageminl=30;
9475: hstepm=stepsize*YEARM; /* Every year of age */
9476: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9477:
9478: /* hstepm=1; aff par mois*/
9479: pstamp(ficrespijb);
9480: fprintf(ficrespijb,"#****** h Pij x Back Probability to be in state i at age x-h being in j at x ");
1.227 brouard 9481: i1= pow(2,cptcoveff);
1.218 brouard 9482: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9483: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9484: /* k=k+1; */
1.238 brouard 9485: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9486: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
9487: if(TKresult[nres]!= k)
9488: continue;
9489: fprintf(ficrespijb,"\n#****** ");
9490: for(j=1;j<=cptcoveff;j++)
9491: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9492: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9493: fprintf(ficrespijb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9494: }
9495: fprintf(ficrespijb,"******\n");
9496: if(invalidvarcomb[k]){
9497: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9498: continue;
9499: }
9500:
9501: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9502: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9503: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9504: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9505: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9506:
9507: /* nhstepm=nhstepm*YEARM; aff par mois*/
9508:
9509: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9510: /* oldm=oldms;savm=savms; */
9511: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9512: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9513: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
9514: fprintf(ficrespijb,"# Cov Agex agex-h hpijx with i,j=");
1.217 brouard 9515: for(i=1; i<=nlstate;i++)
9516: for(j=1; j<=nlstate+ndeath;j++)
1.238 brouard 9517: fprintf(ficrespijb," %1d-%1d",i,j);
1.217 brouard 9518: fprintf(ficrespijb,"\n");
1.238 brouard 9519: for (h=0; h<=nhstepm; h++){
9520: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9521: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9522: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
9523: for(i=1; i<=nlstate;i++)
9524: for(j=1; j<=nlstate+ndeath;j++)
9525: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
9526: fprintf(ficrespijb,"\n");
9527: }
9528: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9529: fprintf(ficrespijb,"\n");
9530: } /* end age deb */
9531: } /* end combination */
9532: } /* end nres */
1.218 brouard 9533: return 0;
9534: } /* hBijx */
1.217 brouard 9535:
1.180 brouard 9536:
1.136 brouard 9537: /***********************************************/
9538: /**************** Main Program *****************/
9539: /***********************************************/
9540:
9541: int main(int argc, char *argv[])
9542: {
9543: #ifdef GSL
9544: const gsl_multimin_fminimizer_type *T;
9545: size_t iteri = 0, it;
9546: int rval = GSL_CONTINUE;
9547: int status = GSL_SUCCESS;
9548: double ssval;
9549: #endif
9550: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9551: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9552: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9553: int jj, ll, li, lj, lk;
1.136 brouard 9554: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9555: int num_filled;
1.136 brouard 9556: int itimes;
9557: int NDIM=2;
9558: int vpopbased=0;
1.235 brouard 9559: int nres=0;
1.136 brouard 9560:
1.164 brouard 9561: char ca[32], cb[32];
1.136 brouard 9562: /* FILE *fichtm; *//* Html File */
9563: /* FILE *ficgp;*/ /*Gnuplot File */
9564: struct stat info;
1.191 brouard 9565: double agedeb=0.;
1.194 brouard 9566:
9567: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9568: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9569:
1.165 brouard 9570: double fret;
1.191 brouard 9571: double dum=0.; /* Dummy variable */
1.136 brouard 9572: double ***p3mat;
1.218 brouard 9573: /* double ***mobaverage; */
1.164 brouard 9574:
9575: char line[MAXLINE];
1.197 brouard 9576: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9577:
1.234 brouard 9578: char modeltemp[MAXLINE];
1.230 brouard 9579: char resultline[MAXLINE];
9580:
1.136 brouard 9581: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9582: char *tok, *val; /* pathtot */
1.136 brouard 9583: int firstobs=1, lastobs=10;
1.195 brouard 9584: int c, h , cpt, c2;
1.191 brouard 9585: int jl=0;
9586: int i1, j1, jk, stepsize=0;
1.194 brouard 9587: int count=0;
9588:
1.164 brouard 9589: int *tab;
1.136 brouard 9590: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9591: int backcast=0;
1.136 brouard 9592: int mobilav=0,popforecast=0;
1.191 brouard 9593: int hstepm=0, nhstepm=0;
1.136 brouard 9594: int agemortsup;
9595: float sumlpop=0.;
9596: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9597: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9598:
1.191 brouard 9599: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9600: double ftolpl=FTOL;
9601: double **prlim;
1.217 brouard 9602: double **bprlim;
1.136 brouard 9603: double ***param; /* Matrix of parameters */
9604: double *p;
9605: double **matcov; /* Matrix of covariance */
1.203 brouard 9606: double **hess; /* Hessian matrix */
1.136 brouard 9607: double ***delti3; /* Scale */
9608: double *delti; /* Scale */
9609: double ***eij, ***vareij;
9610: double **varpl; /* Variances of prevalence limits by age */
9611: double *epj, vepp;
1.164 brouard 9612:
1.136 brouard 9613: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9614: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9615:
1.136 brouard 9616: double **ximort;
1.145 brouard 9617: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9618: int *dcwave;
9619:
1.164 brouard 9620: char z[1]="c";
1.136 brouard 9621:
9622: /*char *strt;*/
9623: char strtend[80];
1.126 brouard 9624:
1.164 brouard 9625:
1.126 brouard 9626: /* setlocale (LC_ALL, ""); */
9627: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9628: /* textdomain (PACKAGE); */
9629: /* setlocale (LC_CTYPE, ""); */
9630: /* setlocale (LC_MESSAGES, ""); */
9631:
9632: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9633: rstart_time = time(NULL);
9634: /* (void) gettimeofday(&start_time,&tzp);*/
9635: start_time = *localtime(&rstart_time);
1.126 brouard 9636: curr_time=start_time;
1.157 brouard 9637: /*tml = *localtime(&start_time.tm_sec);*/
9638: /* strcpy(strstart,asctime(&tml)); */
9639: strcpy(strstart,asctime(&start_time));
1.126 brouard 9640:
9641: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9642: /* tp.tm_sec = tp.tm_sec +86400; */
9643: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9644: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9645: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9646: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9647: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9648: /* strt=asctime(&tmg); */
9649: /* printf("Time(after) =%s",strstart); */
9650: /* (void) time (&time_value);
9651: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9652: * tm = *localtime(&time_value);
9653: * strstart=asctime(&tm);
9654: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9655: */
9656:
9657: nberr=0; /* Number of errors and warnings */
9658: nbwarn=0;
1.184 brouard 9659: #ifdef WIN32
9660: _getcwd(pathcd, size);
9661: #else
1.126 brouard 9662: getcwd(pathcd, size);
1.184 brouard 9663: #endif
1.191 brouard 9664: syscompilerinfo(0);
1.196 brouard 9665: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9666: if(argc <=1){
9667: printf("\nEnter the parameter file name: ");
1.205 brouard 9668: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9669: printf("ERROR Empty parameter file name\n");
9670: goto end;
9671: }
1.126 brouard 9672: i=strlen(pathr);
9673: if(pathr[i-1]=='\n')
9674: pathr[i-1]='\0';
1.156 brouard 9675: i=strlen(pathr);
1.205 brouard 9676: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9677: pathr[i-1]='\0';
1.205 brouard 9678: }
9679: i=strlen(pathr);
9680: if( i==0 ){
9681: printf("ERROR Empty parameter file name\n");
9682: goto end;
9683: }
9684: for (tok = pathr; tok != NULL; ){
1.126 brouard 9685: printf("Pathr |%s|\n",pathr);
9686: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9687: printf("val= |%s| pathr=%s\n",val,pathr);
9688: strcpy (pathtot, val);
9689: if(pathr[0] == '\0') break; /* Dirty */
9690: }
9691: }
9692: else{
9693: strcpy(pathtot,argv[1]);
9694: }
9695: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9696: /*cygwin_split_path(pathtot,path,optionfile);
9697: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9698: /* cutv(path,optionfile,pathtot,'\\');*/
9699:
9700: /* Split argv[0], imach program to get pathimach */
9701: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9702: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9703: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9704: /* strcpy(pathimach,argv[0]); */
9705: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9706: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9707: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9708: #ifdef WIN32
9709: _chdir(path); /* Can be a relative path */
9710: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9711: #else
1.126 brouard 9712: chdir(path); /* Can be a relative path */
1.184 brouard 9713: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9714: #endif
9715: printf("Current directory %s!\n",pathcd);
1.126 brouard 9716: strcpy(command,"mkdir ");
9717: strcat(command,optionfilefiname);
9718: if((outcmd=system(command)) != 0){
1.169 brouard 9719: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9720: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9721: /* fclose(ficlog); */
9722: /* exit(1); */
9723: }
9724: /* if((imk=mkdir(optionfilefiname))<0){ */
9725: /* perror("mkdir"); */
9726: /* } */
9727:
9728: /*-------- arguments in the command line --------*/
9729:
1.186 brouard 9730: /* Main Log file */
1.126 brouard 9731: strcat(filelog, optionfilefiname);
9732: strcat(filelog,".log"); /* */
9733: if((ficlog=fopen(filelog,"w"))==NULL) {
9734: printf("Problem with logfile %s\n",filelog);
9735: goto end;
9736: }
9737: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9738: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9739: fprintf(ficlog,"\nEnter the parameter file name: \n");
9740: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9741: path=%s \n\
9742: optionfile=%s\n\
9743: optionfilext=%s\n\
1.156 brouard 9744: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9745:
1.197 brouard 9746: syscompilerinfo(1);
1.167 brouard 9747:
1.126 brouard 9748: printf("Local time (at start):%s",strstart);
9749: fprintf(ficlog,"Local time (at start): %s",strstart);
9750: fflush(ficlog);
9751: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9752: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9753:
9754: /* */
9755: strcpy(fileres,"r");
9756: strcat(fileres, optionfilefiname);
1.201 brouard 9757: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9758: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9759: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 9760:
1.186 brouard 9761: /* Main ---------arguments file --------*/
1.126 brouard 9762:
9763: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 9764: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
9765: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 9766: fflush(ficlog);
1.149 brouard 9767: /* goto end; */
9768: exit(70);
1.126 brouard 9769: }
9770:
9771:
9772:
9773: strcpy(filereso,"o");
1.201 brouard 9774: strcat(filereso,fileresu);
1.126 brouard 9775: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
9776: printf("Problem with Output resultfile: %s\n", filereso);
9777: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
9778: fflush(ficlog);
9779: goto end;
9780: }
9781:
9782: /* Reads comments: lines beginning with '#' */
9783: numlinepar=0;
1.197 brouard 9784:
9785: /* First parameter line */
9786: while(fgets(line, MAXLINE, ficpar)) {
9787: /* If line starts with a # it is a comment */
9788: if (line[0] == '#') {
9789: numlinepar++;
9790: fputs(line,stdout);
9791: fputs(line,ficparo);
9792: fputs(line,ficlog);
9793: continue;
9794: }else
9795: break;
9796: }
9797: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
9798: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
9799: if (num_filled != 5) {
9800: printf("Should be 5 parameters\n");
9801: }
1.126 brouard 9802: numlinepar++;
1.197 brouard 9803: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
9804: }
9805: /* Second parameter line */
9806: while(fgets(line, MAXLINE, ficpar)) {
9807: /* If line starts with a # it is a comment */
9808: if (line[0] == '#') {
9809: numlinepar++;
9810: fputs(line,stdout);
9811: fputs(line,ficparo);
9812: fputs(line,ficlog);
9813: continue;
9814: }else
9815: break;
9816: }
1.223 brouard 9817: 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", \
9818: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
9819: if (num_filled != 11) {
9820: 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 9821: printf("but line=%s\n",line);
1.197 brouard 9822: }
1.223 brouard 9823: 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 9824: }
1.203 brouard 9825: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 9826: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 9827: /* Third parameter line */
9828: while(fgets(line, MAXLINE, ficpar)) {
9829: /* If line starts with a # it is a comment */
9830: if (line[0] == '#') {
9831: numlinepar++;
9832: fputs(line,stdout);
9833: fputs(line,ficparo);
9834: fputs(line,ficlog);
9835: continue;
9836: }else
9837: break;
9838: }
1.201 brouard 9839: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
9840: if (num_filled == 0)
9841: model[0]='\0';
9842: else if (num_filled != 1){
1.197 brouard 9843: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9844: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9845: model[0]='\0';
9846: goto end;
9847: }
9848: else{
9849: if (model[0]=='+'){
9850: for(i=1; i<=strlen(model);i++)
9851: modeltemp[i-1]=model[i];
1.201 brouard 9852: strcpy(model,modeltemp);
1.197 brouard 9853: }
9854: }
1.199 brouard 9855: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 9856: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 9857: }
9858: /* 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); */
9859: /* numlinepar=numlinepar+3; /\* In general *\/ */
9860: /* 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 9861: 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);
9862: 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 9863: fflush(ficlog);
1.190 brouard 9864: /* if(model[0]=='#'|| model[0]== '\0'){ */
9865: if(model[0]=='#'){
1.187 brouard 9866: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
9867: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
9868: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
9869: if(mle != -1){
9870: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
9871: exit(1);
9872: }
9873: }
1.126 brouard 9874: while((c=getc(ficpar))=='#' && c!= EOF){
9875: ungetc(c,ficpar);
9876: fgets(line, MAXLINE, ficpar);
9877: numlinepar++;
1.195 brouard 9878: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
9879: z[0]=line[1];
9880: }
9881: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 9882: fputs(line, stdout);
9883: //puts(line);
1.126 brouard 9884: fputs(line,ficparo);
9885: fputs(line,ficlog);
9886: }
9887: ungetc(c,ficpar);
9888:
9889:
1.145 brouard 9890: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 9891: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 9892: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 9893: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 9894: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
9895: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
9896: v1+v2*age+v2*v3 makes cptcovn = 3
9897: */
9898: if (strlen(model)>1)
1.187 brouard 9899: 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 9900: else
1.187 brouard 9901: ncovmodel=2; /* Constant and age */
1.133 brouard 9902: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
9903: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 9904: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
9905: 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);
9906: 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);
9907: fflush(stdout);
9908: fclose (ficlog);
9909: goto end;
9910: }
1.126 brouard 9911: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9912: delti=delti3[1][1];
9913: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
9914: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
1.247 brouard 9915: /* We could also provide initial parameters values giving by simple logistic regression
9916: * only one way, that is without matrix product. We will have nlstate maximizations */
9917: /* for(i=1;i<nlstate;i++){ */
9918: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
9919: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
9920: /* } */
1.126 brouard 9921: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 9922: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
9923: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9924: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
9925: fclose (ficparo);
9926: fclose (ficlog);
9927: goto end;
9928: exit(0);
1.248 brouard 9929: } else if(mle==-2) { /* Guessing from means */
9930: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
9931: printf(" You chose mle=-2, look at file %s for a template of covariance matrix \n",filereso);
9932: fprintf(ficlog," You chose mle=-2, look at file %s for a template of covariance matrix \n",filereso);
9933:
1.220 brouard 9934: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 9935: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 9936: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
9937: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9938: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9939: matcov=matrix(1,npar,1,npar);
1.203 brouard 9940: hess=matrix(1,npar,1,npar);
1.220 brouard 9941: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 9942: /* Read guessed parameters */
1.126 brouard 9943: /* Reads comments: lines beginning with '#' */
9944: while((c=getc(ficpar))=='#' && c!= EOF){
9945: ungetc(c,ficpar);
9946: fgets(line, MAXLINE, ficpar);
9947: numlinepar++;
1.141 brouard 9948: fputs(line,stdout);
1.126 brouard 9949: fputs(line,ficparo);
9950: fputs(line,ficlog);
9951: }
9952: ungetc(c,ficpar);
9953:
9954: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9955: for(i=1; i <=nlstate; i++){
1.234 brouard 9956: j=0;
1.126 brouard 9957: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 9958: if(jj==i) continue;
9959: j++;
9960: fscanf(ficpar,"%1d%1d",&i1,&j1);
9961: if ((i1 != i) || (j1 != jj)){
9962: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 9963: It might be a problem of design; if ncovcol and the model are correct\n \
9964: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 9965: exit(1);
9966: }
9967: fprintf(ficparo,"%1d%1d",i1,j1);
9968: if(mle==1)
9969: printf("%1d%1d",i,jj);
9970: fprintf(ficlog,"%1d%1d",i,jj);
9971: for(k=1; k<=ncovmodel;k++){
9972: fscanf(ficpar," %lf",¶m[i][j][k]);
9973: if(mle==1){
9974: printf(" %lf",param[i][j][k]);
9975: fprintf(ficlog," %lf",param[i][j][k]);
9976: }
9977: else
9978: fprintf(ficlog," %lf",param[i][j][k]);
9979: fprintf(ficparo," %lf",param[i][j][k]);
9980: }
9981: fscanf(ficpar,"\n");
9982: numlinepar++;
9983: if(mle==1)
9984: printf("\n");
9985: fprintf(ficlog,"\n");
9986: fprintf(ficparo,"\n");
1.126 brouard 9987: }
9988: }
9989: fflush(ficlog);
1.234 brouard 9990:
1.145 brouard 9991: /* Reads scales values */
1.126 brouard 9992: p=param[1][1];
9993:
9994: /* Reads comments: lines beginning with '#' */
9995: while((c=getc(ficpar))=='#' && c!= EOF){
9996: ungetc(c,ficpar);
9997: fgets(line, MAXLINE, ficpar);
9998: numlinepar++;
1.141 brouard 9999: fputs(line,stdout);
1.126 brouard 10000: fputs(line,ficparo);
10001: fputs(line,ficlog);
10002: }
10003: ungetc(c,ficpar);
10004:
10005: for(i=1; i <=nlstate; i++){
10006: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 10007: fscanf(ficpar,"%1d%1d",&i1,&j1);
10008: if ( (i1-i) * (j1-j) != 0){
10009: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
10010: exit(1);
10011: }
10012: printf("%1d%1d",i,j);
10013: fprintf(ficparo,"%1d%1d",i1,j1);
10014: fprintf(ficlog,"%1d%1d",i1,j1);
10015: for(k=1; k<=ncovmodel;k++){
10016: fscanf(ficpar,"%le",&delti3[i][j][k]);
10017: printf(" %le",delti3[i][j][k]);
10018: fprintf(ficparo," %le",delti3[i][j][k]);
10019: fprintf(ficlog," %le",delti3[i][j][k]);
10020: }
10021: fscanf(ficpar,"\n");
10022: numlinepar++;
10023: printf("\n");
10024: fprintf(ficparo,"\n");
10025: fprintf(ficlog,"\n");
1.126 brouard 10026: }
10027: }
10028: fflush(ficlog);
1.234 brouard 10029:
1.145 brouard 10030: /* Reads covariance matrix */
1.126 brouard 10031: delti=delti3[1][1];
1.220 brouard 10032:
10033:
1.126 brouard 10034: /* 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 10035:
1.126 brouard 10036: /* Reads comments: lines beginning with '#' */
10037: while((c=getc(ficpar))=='#' && c!= EOF){
10038: ungetc(c,ficpar);
10039: fgets(line, MAXLINE, ficpar);
10040: numlinepar++;
1.141 brouard 10041: fputs(line,stdout);
1.126 brouard 10042: fputs(line,ficparo);
10043: fputs(line,ficlog);
10044: }
10045: ungetc(c,ficpar);
1.220 brouard 10046:
1.126 brouard 10047: matcov=matrix(1,npar,1,npar);
1.203 brouard 10048: hess=matrix(1,npar,1,npar);
1.131 brouard 10049: for(i=1; i <=npar; i++)
10050: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 10051:
1.194 brouard 10052: /* Scans npar lines */
1.126 brouard 10053: for(i=1; i <=npar; i++){
1.226 brouard 10054: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 10055: if(count != 3){
1.226 brouard 10056: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10057: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10058: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10059: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10060: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10061: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10062: exit(1);
1.220 brouard 10063: }else{
1.226 brouard 10064: if(mle==1)
10065: printf("%1d%1d%d",i1,j1,jk);
10066: }
10067: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
10068: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 10069: for(j=1; j <=i; j++){
1.226 brouard 10070: fscanf(ficpar," %le",&matcov[i][j]);
10071: if(mle==1){
10072: printf(" %.5le",matcov[i][j]);
10073: }
10074: fprintf(ficlog," %.5le",matcov[i][j]);
10075: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 10076: }
10077: fscanf(ficpar,"\n");
10078: numlinepar++;
10079: if(mle==1)
1.220 brouard 10080: printf("\n");
1.126 brouard 10081: fprintf(ficlog,"\n");
10082: fprintf(ficparo,"\n");
10083: }
1.194 brouard 10084: /* End of read covariance matrix npar lines */
1.126 brouard 10085: for(i=1; i <=npar; i++)
10086: for(j=i+1;j<=npar;j++)
1.226 brouard 10087: matcov[i][j]=matcov[j][i];
1.126 brouard 10088:
10089: if(mle==1)
10090: printf("\n");
10091: fprintf(ficlog,"\n");
10092:
10093: fflush(ficlog);
10094:
10095: /*-------- Rewriting parameter file ----------*/
10096: strcpy(rfileres,"r"); /* "Rparameterfile */
10097: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
10098: strcat(rfileres,"."); /* */
10099: strcat(rfileres,optionfilext); /* Other files have txt extension */
10100: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 10101: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
10102: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 10103: }
10104: fprintf(ficres,"#%s\n",version);
10105: } /* End of mle != -3 */
1.218 brouard 10106:
1.186 brouard 10107: /* Main data
10108: */
1.126 brouard 10109: n= lastobs;
10110: num=lvector(1,n);
10111: moisnais=vector(1,n);
10112: annais=vector(1,n);
10113: moisdc=vector(1,n);
10114: andc=vector(1,n);
1.220 brouard 10115: weight=vector(1,n);
1.126 brouard 10116: agedc=vector(1,n);
10117: cod=ivector(1,n);
1.220 brouard 10118: for(i=1;i<=n;i++){
1.234 brouard 10119: num[i]=0;
10120: moisnais[i]=0;
10121: annais[i]=0;
10122: moisdc[i]=0;
10123: andc[i]=0;
10124: agedc[i]=0;
10125: cod[i]=0;
10126: weight[i]=1.0; /* Equal weights, 1 by default */
10127: }
1.126 brouard 10128: mint=matrix(1,maxwav,1,n);
10129: anint=matrix(1,maxwav,1,n);
1.131 brouard 10130: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 10131: tab=ivector(1,NCOVMAX);
1.144 brouard 10132: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 10133: 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 10134:
1.136 brouard 10135: /* Reads data from file datafile */
10136: if (readdata(datafile, firstobs, lastobs, &imx)==1)
10137: goto end;
10138:
10139: /* Calculation of the number of parameters from char model */
1.234 brouard 10140: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 10141: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
10142: k=3 V4 Tvar[k=3]= 4 (from V4)
10143: k=2 V1 Tvar[k=2]= 1 (from V1)
10144: k=1 Tvar[1]=2 (from V2)
1.234 brouard 10145: */
10146:
10147: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
10148: TvarsDind=ivector(1,NCOVMAX); /* */
10149: TvarsD=ivector(1,NCOVMAX); /* */
10150: TvarsQind=ivector(1,NCOVMAX); /* */
10151: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 10152: TvarF=ivector(1,NCOVMAX); /* */
10153: TvarFind=ivector(1,NCOVMAX); /* */
10154: TvarV=ivector(1,NCOVMAX); /* */
10155: TvarVind=ivector(1,NCOVMAX); /* */
10156: TvarA=ivector(1,NCOVMAX); /* */
10157: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 10158: TvarFD=ivector(1,NCOVMAX); /* */
10159: TvarFDind=ivector(1,NCOVMAX); /* */
10160: TvarFQ=ivector(1,NCOVMAX); /* */
10161: TvarFQind=ivector(1,NCOVMAX); /* */
10162: TvarVD=ivector(1,NCOVMAX); /* */
10163: TvarVDind=ivector(1,NCOVMAX); /* */
10164: TvarVQ=ivector(1,NCOVMAX); /* */
10165: TvarVQind=ivector(1,NCOVMAX); /* */
10166:
1.230 brouard 10167: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 10168: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 10169: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
10170: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
10171: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 10172: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
10173: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
10174: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
10175: */
10176: /* For model-covariate k tells which data-covariate to use but
10177: because this model-covariate is a construction we invent a new column
10178: ncovcol + k1
10179: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
10180: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 10181: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
10182: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 10183: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
10184: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 10185: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 10186: */
1.145 brouard 10187: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
10188: 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 10189: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
10190: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 10191: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 10192: 4 covariates (3 plus signs)
10193: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
10194: */
1.230 brouard 10195: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 10196: * individual dummy, fixed or varying:
10197: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
10198: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 10199: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
10200: * V1 df, V2 qf, V3 & V4 dv, V5 qv
10201: * Tmodelind[1]@9={9,0,3,2,}*/
10202: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
10203: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 10204: * individual quantitative, fixed or varying:
10205: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
10206: * 3, 1, 0, 0, 0, 0, 0, 0},
10207: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 10208: /* Main decodemodel */
10209:
1.187 brouard 10210:
1.223 brouard 10211: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 10212: goto end;
10213:
1.137 brouard 10214: if((double)(lastobs-imx)/(double)imx > 1.10){
10215: nbwarn++;
10216: 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);
10217: 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);
10218: }
1.136 brouard 10219: /* if(mle==1){*/
1.137 brouard 10220: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
10221: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 10222: }
10223:
10224: /*-calculation of age at interview from date of interview and age at death -*/
10225: agev=matrix(1,maxwav,1,imx);
10226:
10227: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
10228: goto end;
10229:
1.126 brouard 10230:
1.136 brouard 10231: agegomp=(int)agemin;
10232: free_vector(moisnais,1,n);
10233: free_vector(annais,1,n);
1.126 brouard 10234: /* free_matrix(mint,1,maxwav,1,n);
10235: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 10236: /* free_vector(moisdc,1,n); */
10237: /* free_vector(andc,1,n); */
1.145 brouard 10238: /* */
10239:
1.126 brouard 10240: wav=ivector(1,imx);
1.214 brouard 10241: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
10242: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
10243: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
10244: 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.*/
10245: bh=imatrix(1,lastpass-firstpass+2,1,imx);
10246: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 10247:
10248: /* Concatenates waves */
1.214 brouard 10249: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
10250: Death is a valid wave (if date is known).
10251: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
10252: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
10253: and mw[mi+1][i]. dh depends on stepm.
10254: */
10255:
1.126 brouard 10256: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.248 brouard 10257: /* Concatenates waves */
1.145 brouard 10258:
1.215 brouard 10259: free_vector(moisdc,1,n);
10260: free_vector(andc,1,n);
10261:
1.126 brouard 10262: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
10263: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
10264: ncodemax[1]=1;
1.145 brouard 10265: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 10266: cptcoveff=0;
1.220 brouard 10267: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
10268: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 10269: }
10270:
10271: ncovcombmax=pow(2,cptcoveff);
10272: invalidvarcomb=ivector(1, ncovcombmax);
10273: for(i=1;i<ncovcombmax;i++)
10274: invalidvarcomb[i]=0;
10275:
1.211 brouard 10276: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 10277: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 10278: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 10279:
1.200 brouard 10280: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 10281: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 10282: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 10283: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
10284: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
10285: * (currently 0 or 1) in the data.
10286: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
10287: * corresponding modality (h,j).
10288: */
10289:
1.145 brouard 10290: h=0;
10291: /*if (cptcovn > 0) */
1.126 brouard 10292: m=pow(2,cptcoveff);
10293:
1.144 brouard 10294: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 10295: * For k=4 covariates, h goes from 1 to m=2**k
10296: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
10297: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 10298: * h\k 1 2 3 4
1.143 brouard 10299: *______________________________
10300: * 1 i=1 1 i=1 1 i=1 1 i=1 1
10301: * 2 2 1 1 1
10302: * 3 i=2 1 2 1 1
10303: * 4 2 2 1 1
10304: * 5 i=3 1 i=2 1 2 1
10305: * 6 2 1 2 1
10306: * 7 i=4 1 2 2 1
10307: * 8 2 2 2 1
1.197 brouard 10308: * 9 i=5 1 i=3 1 i=2 1 2
10309: * 10 2 1 1 2
10310: * 11 i=6 1 2 1 2
10311: * 12 2 2 1 2
10312: * 13 i=7 1 i=4 1 2 2
10313: * 14 2 1 2 2
10314: * 15 i=8 1 2 2 2
10315: * 16 2 2 2 2
1.143 brouard 10316: */
1.212 brouard 10317: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10318: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10319: * and the value of each covariate?
10320: * V1=1, V2=1, V3=2, V4=1 ?
10321: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10322: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10323: * In order to get the real value in the data, we use nbcode
10324: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10325: * We are keeping this crazy system in order to be able (in the future?)
10326: * to have more than 2 values (0 or 1) for a covariate.
10327: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10328: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10329: * bbbbbbbb
10330: * 76543210
10331: * h-1 00000101 (6-1=5)
1.219 brouard 10332: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10333: * &
10334: * 1 00000001 (1)
1.219 brouard 10335: * 00000000 = 1 & ((h-1) >> (k-1))
10336: * +1= 00000001 =1
1.211 brouard 10337: *
10338: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10339: * h' 1101 =2^3+2^2+0x2^1+2^0
10340: * >>k' 11
10341: * & 00000001
10342: * = 00000001
10343: * +1 = 00000010=2 = codtabm(14,3)
10344: * Reverse h=6 and m=16?
10345: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10346: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10347: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10348: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10349: * V3=decodtabm(14,3,2**4)=2
10350: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10351: *(h-1) >> (j-1) 0011 =13 >> 2
10352: * &1 000000001
10353: * = 000000001
10354: * +1= 000000010 =2
10355: * 2211
10356: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10357: * V3=2
1.220 brouard 10358: * codtabm and decodtabm are identical
1.211 brouard 10359: */
10360:
1.145 brouard 10361:
10362: free_ivector(Ndum,-1,NCOVMAX);
10363:
10364:
1.126 brouard 10365:
1.186 brouard 10366: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10367: strcpy(optionfilegnuplot,optionfilefiname);
10368: if(mle==-3)
1.201 brouard 10369: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10370: strcat(optionfilegnuplot,".gp");
10371:
10372: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10373: printf("Problem with file %s",optionfilegnuplot);
10374: }
10375: else{
1.204 brouard 10376: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10377: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10378: //fprintf(ficgp,"set missing 'NaNq'\n");
10379: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10380: }
10381: /* fclose(ficgp);*/
1.186 brouard 10382:
10383:
10384: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10385:
10386: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10387: if(mle==-3)
1.201 brouard 10388: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10389: strcat(optionfilehtm,".htm");
10390: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10391: printf("Problem with %s \n",optionfilehtm);
10392: exit(0);
1.126 brouard 10393: }
10394:
10395: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10396: strcat(optionfilehtmcov,"-cov.htm");
10397: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10398: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10399: }
10400: else{
10401: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10402: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10403: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10404: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10405: }
10406:
1.213 brouard 10407: 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 10408: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10409: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10410: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10411: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10412: \n\
10413: <hr size=\"2\" color=\"#EC5E5E\">\
10414: <ul><li><h4>Parameter files</h4>\n\
10415: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10416: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10417: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10418: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10419: - Date and time at start: %s</ul>\n",\
10420: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10421: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10422: fileres,fileres,\
10423: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10424: fflush(fichtm);
10425:
10426: strcpy(pathr,path);
10427: strcat(pathr,optionfilefiname);
1.184 brouard 10428: #ifdef WIN32
10429: _chdir(optionfilefiname); /* Move to directory named optionfile */
10430: #else
1.126 brouard 10431: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10432: #endif
10433:
1.126 brouard 10434:
1.220 brouard 10435: /* Calculates basic frequencies. Computes observed prevalence at single age
10436: and for any valid combination of covariates
1.126 brouard 10437: and prints on file fileres'p'. */
1.227 brouard 10438: freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
10439: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10440:
10441: fprintf(fichtm,"\n");
10442: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10443: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10444: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10445: imx,agemin,agemax,jmin,jmax,jmean);
10446: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10447: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10448: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10449: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10450: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10451:
1.126 brouard 10452: /* For Powell, parameters are in a vector p[] starting at p[1]
10453: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10454: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10455:
10456: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10457: /* For mortality only */
1.126 brouard 10458: if (mle==-3){
1.136 brouard 10459: ximort=matrix(1,NDIM,1,NDIM);
1.248 brouard 10460: for(i=1;i<=NDIM;i++)
10461: for(j=1;j<=NDIM;j++)
10462: ximort[i][j]=0.;
1.186 brouard 10463: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10464: cens=ivector(1,n);
10465: ageexmed=vector(1,n);
10466: agecens=vector(1,n);
10467: dcwave=ivector(1,n);
1.223 brouard 10468:
1.126 brouard 10469: for (i=1; i<=imx; i++){
10470: dcwave[i]=-1;
10471: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10472: if (s[m][i]>nlstate) {
10473: dcwave[i]=m;
10474: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10475: break;
10476: }
1.126 brouard 10477: }
1.226 brouard 10478:
1.126 brouard 10479: for (i=1; i<=imx; i++) {
10480: if (wav[i]>0){
1.226 brouard 10481: ageexmed[i]=agev[mw[1][i]][i];
10482: j=wav[i];
10483: agecens[i]=1.;
10484:
10485: if (ageexmed[i]> 1 && wav[i] > 0){
10486: agecens[i]=agev[mw[j][i]][i];
10487: cens[i]= 1;
10488: }else if (ageexmed[i]< 1)
10489: cens[i]= -1;
10490: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10491: cens[i]=0 ;
1.126 brouard 10492: }
10493: else cens[i]=-1;
10494: }
10495:
10496: for (i=1;i<=NDIM;i++) {
10497: for (j=1;j<=NDIM;j++)
1.226 brouard 10498: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10499: }
10500:
1.145 brouard 10501: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10502: /*printf("%lf %lf", p[1], p[2]);*/
10503:
10504:
1.136 brouard 10505: #ifdef GSL
10506: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10507: #else
1.126 brouard 10508: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10509: #endif
1.201 brouard 10510: strcpy(filerespow,"POW-MORT_");
10511: strcat(filerespow,fileresu);
1.126 brouard 10512: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10513: printf("Problem with resultfile: %s\n", filerespow);
10514: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10515: }
1.136 brouard 10516: #ifdef GSL
10517: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10518: #else
1.126 brouard 10519: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10520: #endif
1.126 brouard 10521: /* for (i=1;i<=nlstate;i++)
10522: for(j=1;j<=nlstate+ndeath;j++)
10523: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10524: */
10525: fprintf(ficrespow,"\n");
1.136 brouard 10526: #ifdef GSL
10527: /* gsl starts here */
10528: T = gsl_multimin_fminimizer_nmsimplex;
10529: gsl_multimin_fminimizer *sfm = NULL;
10530: gsl_vector *ss, *x;
10531: gsl_multimin_function minex_func;
10532:
10533: /* Initial vertex size vector */
10534: ss = gsl_vector_alloc (NDIM);
10535:
10536: if (ss == NULL){
10537: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10538: }
10539: /* Set all step sizes to 1 */
10540: gsl_vector_set_all (ss, 0.001);
10541:
10542: /* Starting point */
1.126 brouard 10543:
1.136 brouard 10544: x = gsl_vector_alloc (NDIM);
10545:
10546: if (x == NULL){
10547: gsl_vector_free(ss);
10548: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10549: }
10550:
10551: /* Initialize method and iterate */
10552: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10553: /* gsl_vector_set(x, 0, 0.0268); */
10554: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10555: gsl_vector_set(x, 0, p[1]);
10556: gsl_vector_set(x, 1, p[2]);
10557:
10558: minex_func.f = &gompertz_f;
10559: minex_func.n = NDIM;
10560: minex_func.params = (void *)&p; /* ??? */
10561:
10562: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10563: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10564:
10565: printf("Iterations beginning .....\n\n");
10566: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10567:
10568: iteri=0;
10569: while (rval == GSL_CONTINUE){
10570: iteri++;
10571: status = gsl_multimin_fminimizer_iterate(sfm);
10572:
10573: if (status) printf("error: %s\n", gsl_strerror (status));
10574: fflush(0);
10575:
10576: if (status)
10577: break;
10578:
10579: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10580: ssval = gsl_multimin_fminimizer_size (sfm);
10581:
10582: if (rval == GSL_SUCCESS)
10583: printf ("converged to a local maximum at\n");
10584:
10585: printf("%5d ", iteri);
10586: for (it = 0; it < NDIM; it++){
10587: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10588: }
10589: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10590: }
10591:
10592: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10593:
10594: gsl_vector_free(x); /* initial values */
10595: gsl_vector_free(ss); /* inital step size */
10596: for (it=0; it<NDIM; it++){
10597: p[it+1]=gsl_vector_get(sfm->x,it);
10598: fprintf(ficrespow," %.12lf", p[it]);
10599: }
10600: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10601: #endif
10602: #ifdef POWELL
10603: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10604: #endif
1.126 brouard 10605: fclose(ficrespow);
10606:
1.203 brouard 10607: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10608:
10609: for(i=1; i <=NDIM; i++)
10610: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10611: matcov[i][j]=matcov[j][i];
1.126 brouard 10612:
10613: printf("\nCovariance matrix\n ");
1.203 brouard 10614: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10615: for(i=1; i <=NDIM; i++) {
10616: for(j=1;j<=NDIM;j++){
1.220 brouard 10617: printf("%f ",matcov[i][j]);
10618: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10619: }
1.203 brouard 10620: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10621: }
10622:
10623: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10624: for (i=1;i<=NDIM;i++) {
1.126 brouard 10625: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10626: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10627: }
1.126 brouard 10628: lsurv=vector(1,AGESUP);
10629: lpop=vector(1,AGESUP);
10630: tpop=vector(1,AGESUP);
10631: lsurv[agegomp]=100000;
10632:
10633: for (k=agegomp;k<=AGESUP;k++) {
10634: agemortsup=k;
10635: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10636: }
10637:
10638: for (k=agegomp;k<agemortsup;k++)
10639: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10640:
10641: for (k=agegomp;k<agemortsup;k++){
10642: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10643: sumlpop=sumlpop+lpop[k];
10644: }
10645:
10646: tpop[agegomp]=sumlpop;
10647: for (k=agegomp;k<(agemortsup-3);k++){
10648: /* tpop[k+1]=2;*/
10649: tpop[k+1]=tpop[k]-lpop[k];
10650: }
10651:
10652:
10653: printf("\nAge lx qx dx Lx Tx e(x)\n");
10654: for (k=agegomp;k<(agemortsup-2);k++)
10655: 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]);
10656:
10657:
10658: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10659: ageminpar=50;
10660: agemaxpar=100;
1.194 brouard 10661: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10662: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10663: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10664: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10665: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10666: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10667: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10668: }else{
10669: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10670: 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 10671: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10672: }
1.201 brouard 10673: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10674: stepm, weightopt,\
10675: model,imx,p,matcov,agemortsup);
10676:
10677: free_vector(lsurv,1,AGESUP);
10678: free_vector(lpop,1,AGESUP);
10679: free_vector(tpop,1,AGESUP);
1.220 brouard 10680: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10681: free_ivector(cens,1,n);
10682: free_vector(agecens,1,n);
10683: free_ivector(dcwave,1,n);
1.220 brouard 10684: #ifdef GSL
1.136 brouard 10685: #endif
1.186 brouard 10686: } /* Endof if mle==-3 mortality only */
1.205 brouard 10687: /* Standard */
10688: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10689: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10690: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10691: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10692: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10693: for (k=1; k<=npar;k++)
10694: printf(" %d %8.5f",k,p[k]);
10695: printf("\n");
1.205 brouard 10696: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10697: /* mlikeli uses func not funcone */
1.247 brouard 10698: /* for(i=1;i<nlstate;i++){ */
10699: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10700: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10701: /* } */
1.205 brouard 10702: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10703: }
10704: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10705: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10706: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10707: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10708: }
10709: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10710: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10711: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10712: for (k=1; k<=npar;k++)
10713: printf(" %d %8.5f",k,p[k]);
10714: printf("\n");
10715:
10716: /*--------- results files --------------*/
1.224 brouard 10717: 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 10718:
10719:
10720: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10721: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10722: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10723: for(i=1,jk=1; i <=nlstate; i++){
10724: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10725: if (k != i) {
10726: printf("%d%d ",i,k);
10727: fprintf(ficlog,"%d%d ",i,k);
10728: fprintf(ficres,"%1d%1d ",i,k);
10729: for(j=1; j <=ncovmodel; j++){
10730: printf("%12.7f ",p[jk]);
10731: fprintf(ficlog,"%12.7f ",p[jk]);
10732: fprintf(ficres,"%12.7f ",p[jk]);
10733: jk++;
10734: }
10735: printf("\n");
10736: fprintf(ficlog,"\n");
10737: fprintf(ficres,"\n");
10738: }
1.126 brouard 10739: }
10740: }
1.203 brouard 10741: if(mle != 0){
10742: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10743: ftolhess=ftol; /* Usually correct */
1.203 brouard 10744: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10745: 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");
10746: 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");
10747: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10748: for(k=1; k <=(nlstate+ndeath); k++){
10749: if (k != i) {
10750: printf("%d%d ",i,k);
10751: fprintf(ficlog,"%d%d ",i,k);
10752: for(j=1; j <=ncovmodel; j++){
10753: 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]));
10754: 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]));
10755: jk++;
10756: }
10757: printf("\n");
10758: fprintf(ficlog,"\n");
10759: }
10760: }
1.193 brouard 10761: }
1.203 brouard 10762: } /* end of hesscov and Wald tests */
1.225 brouard 10763:
1.203 brouard 10764: /* */
1.126 brouard 10765: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
10766: printf("# Scales (for hessian or gradient estimation)\n");
10767: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
10768: for(i=1,jk=1; i <=nlstate; i++){
10769: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 10770: if (j!=i) {
10771: fprintf(ficres,"%1d%1d",i,j);
10772: printf("%1d%1d",i,j);
10773: fprintf(ficlog,"%1d%1d",i,j);
10774: for(k=1; k<=ncovmodel;k++){
10775: printf(" %.5e",delti[jk]);
10776: fprintf(ficlog," %.5e",delti[jk]);
10777: fprintf(ficres," %.5e",delti[jk]);
10778: jk++;
10779: }
10780: printf("\n");
10781: fprintf(ficlog,"\n");
10782: fprintf(ficres,"\n");
10783: }
1.126 brouard 10784: }
10785: }
10786:
10787: 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 10788: if(mle >= 1) /* To big for the screen */
1.126 brouard 10789: 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");
10790: 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");
10791: /* # 121 Var(a12)\n\ */
10792: /* # 122 Cov(b12,a12) Var(b12)\n\ */
10793: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
10794: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
10795: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
10796: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
10797: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
10798: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
10799:
10800:
10801: /* Just to have a covariance matrix which will be more understandable
10802: even is we still don't want to manage dictionary of variables
10803: */
10804: for(itimes=1;itimes<=2;itimes++){
10805: jj=0;
10806: for(i=1; i <=nlstate; i++){
1.225 brouard 10807: for(j=1; j <=nlstate+ndeath; j++){
10808: if(j==i) continue;
10809: for(k=1; k<=ncovmodel;k++){
10810: jj++;
10811: ca[0]= k+'a'-1;ca[1]='\0';
10812: if(itimes==1){
10813: if(mle>=1)
10814: printf("#%1d%1d%d",i,j,k);
10815: fprintf(ficlog,"#%1d%1d%d",i,j,k);
10816: fprintf(ficres,"#%1d%1d%d",i,j,k);
10817: }else{
10818: if(mle>=1)
10819: printf("%1d%1d%d",i,j,k);
10820: fprintf(ficlog,"%1d%1d%d",i,j,k);
10821: fprintf(ficres,"%1d%1d%d",i,j,k);
10822: }
10823: ll=0;
10824: for(li=1;li <=nlstate; li++){
10825: for(lj=1;lj <=nlstate+ndeath; lj++){
10826: if(lj==li) continue;
10827: for(lk=1;lk<=ncovmodel;lk++){
10828: ll++;
10829: if(ll<=jj){
10830: cb[0]= lk +'a'-1;cb[1]='\0';
10831: if(ll<jj){
10832: if(itimes==1){
10833: if(mle>=1)
10834: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10835: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10836: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10837: }else{
10838: if(mle>=1)
10839: printf(" %.5e",matcov[jj][ll]);
10840: fprintf(ficlog," %.5e",matcov[jj][ll]);
10841: fprintf(ficres," %.5e",matcov[jj][ll]);
10842: }
10843: }else{
10844: if(itimes==1){
10845: if(mle>=1)
10846: printf(" Var(%s%1d%1d)",ca,i,j);
10847: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
10848: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
10849: }else{
10850: if(mle>=1)
10851: printf(" %.7e",matcov[jj][ll]);
10852: fprintf(ficlog," %.7e",matcov[jj][ll]);
10853: fprintf(ficres," %.7e",matcov[jj][ll]);
10854: }
10855: }
10856: }
10857: } /* end lk */
10858: } /* end lj */
10859: } /* end li */
10860: if(mle>=1)
10861: printf("\n");
10862: fprintf(ficlog,"\n");
10863: fprintf(ficres,"\n");
10864: numlinepar++;
10865: } /* end k*/
10866: } /*end j */
1.126 brouard 10867: } /* end i */
10868: } /* end itimes */
10869:
10870: fflush(ficlog);
10871: fflush(ficres);
1.225 brouard 10872: while(fgets(line, MAXLINE, ficpar)) {
10873: /* If line starts with a # it is a comment */
10874: if (line[0] == '#') {
10875: numlinepar++;
10876: fputs(line,stdout);
10877: fputs(line,ficparo);
10878: fputs(line,ficlog);
10879: continue;
10880: }else
10881: break;
10882: }
10883:
1.209 brouard 10884: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
10885: /* ungetc(c,ficpar); */
10886: /* fgets(line, MAXLINE, ficpar); */
10887: /* fputs(line,stdout); */
10888: /* fputs(line,ficparo); */
10889: /* } */
10890: /* ungetc(c,ficpar); */
1.126 brouard 10891:
10892: estepm=0;
1.209 brouard 10893: 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 10894:
10895: if (num_filled != 6) {
10896: 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);
10897: 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);
10898: goto end;
10899: }
10900: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
10901: }
10902: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
10903: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
10904:
1.209 brouard 10905: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 10906: if (estepm==0 || estepm < stepm) estepm=stepm;
10907: if (fage <= 2) {
10908: bage = ageminpar;
10909: fage = agemaxpar;
10910: }
10911:
10912: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 10913: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
10914: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 10915:
1.186 brouard 10916: /* Other stuffs, more or less useful */
1.126 brouard 10917: while((c=getc(ficpar))=='#' && c!= EOF){
10918: ungetc(c,ficpar);
10919: fgets(line, MAXLINE, ficpar);
1.141 brouard 10920: fputs(line,stdout);
1.126 brouard 10921: fputs(line,ficparo);
10922: }
10923: ungetc(c,ficpar);
10924:
10925: 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);
10926: 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);
10927: 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);
10928: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
10929: 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);
10930:
10931: while((c=getc(ficpar))=='#' && c!= EOF){
10932: ungetc(c,ficpar);
10933: fgets(line, MAXLINE, ficpar);
1.141 brouard 10934: fputs(line,stdout);
1.126 brouard 10935: fputs(line,ficparo);
10936: }
10937: ungetc(c,ficpar);
10938:
10939:
10940: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
10941: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
10942:
10943: fscanf(ficpar,"pop_based=%d\n",&popbased);
1.193 brouard 10944: fprintf(ficlog,"pop_based=%d\n",popbased);
1.126 brouard 10945: fprintf(ficparo,"pop_based=%d\n",popbased);
10946: fprintf(ficres,"pop_based=%d\n",popbased);
10947:
10948: while((c=getc(ficpar))=='#' && c!= EOF){
10949: ungetc(c,ficpar);
10950: fgets(line, MAXLINE, ficpar);
1.141 brouard 10951: fputs(line,stdout);
1.238 brouard 10952: fputs(line,ficres);
1.126 brouard 10953: fputs(line,ficparo);
10954: }
10955: ungetc(c,ficpar);
10956:
10957: 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);
10958: 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);
10959: 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);
10960: 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);
10961: 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);
10962: /* day and month of proj2 are not used but only year anproj2.*/
10963:
1.217 brouard 10964: while((c=getc(ficpar))=='#' && c!= EOF){
10965: ungetc(c,ficpar);
10966: fgets(line, MAXLINE, ficpar);
10967: fputs(line,stdout);
10968: fputs(line,ficparo);
1.238 brouard 10969: fputs(line,ficres);
1.217 brouard 10970: }
10971: ungetc(c,ficpar);
10972:
10973: 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);
1.223 brouard 10974: 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);
10975: 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);
10976: 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 10977: /* day and month of proj2 are not used but only year anproj2.*/
1.126 brouard 10978:
1.230 brouard 10979: /* Results */
1.235 brouard 10980: nresult=0;
1.230 brouard 10981: while(fgets(line, MAXLINE, ficpar)) {
10982: /* If line starts with a # it is a comment */
10983: if (line[0] == '#') {
10984: numlinepar++;
10985: fputs(line,stdout);
10986: fputs(line,ficparo);
10987: fputs(line,ficlog);
1.238 brouard 10988: fputs(line,ficres);
1.230 brouard 10989: continue;
10990: }else
10991: break;
10992: }
1.240 brouard 10993: if (!feof(ficpar))
1.230 brouard 10994: while((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
1.240 brouard 10995: if (num_filled == 0){
1.230 brouard 10996: resultline[0]='\0';
1.240 brouard 10997: break;
10998: } else if (num_filled != 1){
1.230 brouard 10999: printf("ERROR %d: result line should be at minimum 'result=' %s\n",num_filled, line);
11000: }
1.235 brouard 11001: nresult++; /* Sum of resultlines */
11002: printf("Result %d: result=%s\n",nresult, resultline);
11003: if(nresult > MAXRESULTLINES){
11004: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11005: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11006: goto end;
11007: }
11008: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
1.238 brouard 11009: fprintf(ficparo,"result: %s\n",resultline);
11010: fprintf(ficres,"result: %s\n",resultline);
11011: fprintf(ficlog,"result: %s\n",resultline);
1.230 brouard 11012: while(fgets(line, MAXLINE, ficpar)) {
11013: /* If line starts with a # it is a comment */
11014: if (line[0] == '#') {
11015: numlinepar++;
11016: fputs(line,stdout);
11017: fputs(line,ficparo);
1.238 brouard 11018: fputs(line,ficres);
1.230 brouard 11019: fputs(line,ficlog);
11020: continue;
11021: }else
11022: break;
11023: }
11024: if (feof(ficpar))
11025: break;
11026: else{ /* Processess output results for this combination of covariate values */
11027: }
1.240 brouard 11028: } /* end while */
1.230 brouard 11029:
11030:
1.126 brouard 11031:
1.230 brouard 11032: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 11033: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 11034:
11035: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 11036: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 11037: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11038: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11039: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 11040: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11041: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11042: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11043: }else{
1.218 brouard 11044: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 11045: }
11046: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.225 brouard 11047: model,imx,jmin,jmax,jmean,rfileres,popforecast,prevfcast,backcast, estepm, \
11048: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 11049:
1.225 brouard 11050: /*------------ free_vector -------------*/
11051: /* chdir(path); */
1.220 brouard 11052:
1.215 brouard 11053: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
11054: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
11055: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
11056: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 11057: free_lvector(num,1,n);
11058: free_vector(agedc,1,n);
11059: /*free_matrix(covar,0,NCOVMAX,1,n);*/
11060: /*free_matrix(covar,1,NCOVMAX,1,n);*/
11061: fclose(ficparo);
11062: fclose(ficres);
1.220 brouard 11063:
11064:
1.186 brouard 11065: /* Other results (useful)*/
1.220 brouard 11066:
11067:
1.126 brouard 11068: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 11069: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
11070: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 11071: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 11072: fclose(ficrespl);
11073:
11074: /*------------- h Pij x at various ages ------------*/
1.180 brouard 11075: /*#include "hpijx.h"*/
11076: hPijx(p, bage, fage);
1.145 brouard 11077: fclose(ficrespij);
1.227 brouard 11078:
1.220 brouard 11079: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 11080: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 11081: k=1;
1.126 brouard 11082: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 11083:
1.219 brouard 11084: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 11085: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 11086: for(i=1;i<=AGESUP;i++)
1.219 brouard 11087: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 11088: for(k=1;k<=ncovcombmax;k++)
11089: probs[i][j][k]=0.;
1.219 brouard 11090: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
11091: if (mobilav!=0 ||mobilavproj !=0 ) {
11092: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 11093: for(i=1;i<=AGESUP;i++)
11094: for(j=1;j<=nlstate;j++)
11095: for(k=1;k<=ncovcombmax;k++)
11096: mobaverages[i][j][k]=0.;
1.219 brouard 11097: mobaverage=mobaverages;
11098: if (mobilav!=0) {
1.235 brouard 11099: printf("Movingaveraging observed prevalence\n");
1.227 brouard 11100: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
11101: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
11102: printf(" Error in movingaverage mobilav=%d\n",mobilav);
11103: }
1.219 brouard 11104: }
11105: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
11106: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
11107: else if (mobilavproj !=0) {
1.235 brouard 11108: printf("Movingaveraging projected observed prevalence\n");
1.227 brouard 11109: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
11110: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
11111: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
11112: }
1.219 brouard 11113: }
11114: }/* end if moving average */
1.227 brouard 11115:
1.126 brouard 11116: /*---------- Forecasting ------------------*/
11117: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
11118: if(prevfcast==1){
11119: /* if(stepm ==1){*/
1.225 brouard 11120: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 11121: }
1.217 brouard 11122: if(backcast==1){
1.219 brouard 11123: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11124: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11125: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11126:
11127: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
11128:
11129: bprlim=matrix(1,nlstate,1,nlstate);
11130: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
11131: fclose(ficresplb);
11132:
1.222 brouard 11133: hBijx(p, bage, fage, mobaverage);
11134: fclose(ficrespijb);
1.219 brouard 11135: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
11136:
11137: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 11138: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 11139: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11140: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11141: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11142: }
1.217 brouard 11143:
1.186 brouard 11144:
11145: /* ------ Other prevalence ratios------------ */
1.126 brouard 11146:
1.215 brouard 11147: free_ivector(wav,1,imx);
11148: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
11149: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
11150: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 11151:
11152:
1.127 brouard 11153: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 11154:
1.201 brouard 11155: strcpy(filerese,"E_");
11156: strcat(filerese,fileresu);
1.126 brouard 11157: if((ficreseij=fopen(filerese,"w"))==NULL) {
11158: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11159: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11160: }
1.208 brouard 11161: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
11162: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.238 brouard 11163:
11164: pstamp(ficreseij);
1.219 brouard 11165:
1.235 brouard 11166: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11167: if (cptcovn < 1){i1=1;}
11168:
11169: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11170: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
11171: if(TKresult[nres]!= k)
11172: continue;
1.219 brouard 11173: fprintf(ficreseij,"\n#****** ");
1.235 brouard 11174: printf("\n#****** ");
1.225 brouard 11175: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11176: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11177: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11178: }
11179: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11180: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11181: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 11182: }
11183: fprintf(ficreseij,"******\n");
1.235 brouard 11184: printf("******\n");
1.219 brouard 11185:
11186: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11187: oldm=oldms;savm=savms;
1.235 brouard 11188: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 11189:
1.219 brouard 11190: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 11191: }
11192: fclose(ficreseij);
1.208 brouard 11193: printf("done evsij\n");fflush(stdout);
11194: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 11195:
1.227 brouard 11196: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 11197:
11198:
1.201 brouard 11199: strcpy(filerest,"T_");
11200: strcat(filerest,fileresu);
1.127 brouard 11201: if((ficrest=fopen(filerest,"w"))==NULL) {
11202: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
11203: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
11204: }
1.208 brouard 11205: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
11206: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 11207:
1.126 brouard 11208:
1.201 brouard 11209: strcpy(fileresstde,"STDE_");
11210: strcat(fileresstde,fileresu);
1.126 brouard 11211: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 11212: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
11213: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 11214: }
1.227 brouard 11215: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
11216: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 11217:
1.201 brouard 11218: strcpy(filerescve,"CVE_");
11219: strcat(filerescve,fileresu);
1.126 brouard 11220: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 11221: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
11222: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 11223: }
1.227 brouard 11224: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
11225: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 11226:
1.201 brouard 11227: strcpy(fileresv,"V_");
11228: strcat(fileresv,fileresu);
1.126 brouard 11229: if((ficresvij=fopen(fileresv,"w"))==NULL) {
11230: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
11231: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
11232: }
1.227 brouard 11233: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
11234: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 11235:
1.145 brouard 11236: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11237: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
11238:
1.235 brouard 11239: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11240: if (cptcovn < 1){i1=1;}
11241:
11242: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11243: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
11244: if(TKresult[nres]!= k)
11245: continue;
1.242 brouard 11246: printf("\n#****** Result for:");
11247: fprintf(ficrest,"\n#****** Result for:");
11248: fprintf(ficlog,"\n#****** Result for:");
1.227 brouard 11249: for(j=1;j<=cptcoveff;j++){
11250: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11251: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11252: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11253: }
1.235 brouard 11254: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11255: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11256: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11257: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11258: }
1.208 brouard 11259: fprintf(ficrest,"******\n");
1.227 brouard 11260: fprintf(ficlog,"******\n");
11261: printf("******\n");
1.208 brouard 11262:
11263: fprintf(ficresstdeij,"\n#****** ");
11264: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 11265: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11266: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11267: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 11268: }
1.235 brouard 11269: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11270: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11271: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11272: }
1.208 brouard 11273: fprintf(ficresstdeij,"******\n");
11274: fprintf(ficrescveij,"******\n");
11275:
11276: fprintf(ficresvij,"\n#****** ");
1.238 brouard 11277: /* pstamp(ficresvij); */
1.225 brouard 11278: for(j=1;j<=cptcoveff;j++)
1.227 brouard 11279: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11280: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11281: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11282: }
1.208 brouard 11283: fprintf(ficresvij,"******\n");
11284:
11285: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11286: oldm=oldms;savm=savms;
1.235 brouard 11287: printf(" cvevsij ");
11288: fprintf(ficlog, " cvevsij ");
11289: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 11290: printf(" end cvevsij \n ");
11291: fprintf(ficlog, " end cvevsij \n ");
11292:
11293: /*
11294: */
11295: /* goto endfree; */
11296:
11297: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11298: pstamp(ficrest);
11299:
11300:
11301: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 11302: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
11303: cptcod= 0; /* To be deleted */
11304: printf("varevsij vpopbased=%d \n",vpopbased);
11305: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 11306: 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 11307: 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 ");
11308: if(vpopbased==1)
11309: 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);
11310: else
11311: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
11312: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
11313: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
11314: fprintf(ficrest,"\n");
11315: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
11316: epj=vector(1,nlstate+1);
11317: printf("Computing age specific period (stable) prevalences in each health state \n");
11318: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11319: for(age=bage; age <=fage ;age++){
1.235 brouard 11320: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11321: if (vpopbased==1) {
11322: if(mobilav ==0){
11323: for(i=1; i<=nlstate;i++)
11324: prlim[i][i]=probs[(int)age][i][k];
11325: }else{ /* mobilav */
11326: for(i=1; i<=nlstate;i++)
11327: prlim[i][i]=mobaverage[(int)age][i][k];
11328: }
11329: }
1.219 brouard 11330:
1.227 brouard 11331: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11332: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11333: /* printf(" age %4.0f ",age); */
11334: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11335: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11336: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11337: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11338: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11339: }
11340: epj[nlstate+1] +=epj[j];
11341: }
11342: /* printf(" age %4.0f \n",age); */
1.219 brouard 11343:
1.227 brouard 11344: for(i=1, vepp=0.;i <=nlstate;i++)
11345: for(j=1;j <=nlstate;j++)
11346: vepp += vareij[i][j][(int)age];
11347: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11348: for(j=1;j <=nlstate;j++){
11349: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11350: }
11351: fprintf(ficrest,"\n");
11352: }
1.208 brouard 11353: } /* End vpopbased */
11354: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11355: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11356: free_vector(epj,1,nlstate+1);
1.235 brouard 11357: printf("done selection\n");fflush(stdout);
11358: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11359:
1.145 brouard 11360: /*}*/
1.235 brouard 11361: } /* End k selection */
1.227 brouard 11362:
11363: printf("done State-specific expectancies\n");fflush(stdout);
11364: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11365:
1.126 brouard 11366: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11367:
1.201 brouard 11368: strcpy(fileresvpl,"VPL_");
11369: strcat(fileresvpl,fileresu);
1.126 brouard 11370: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11371: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11372: exit(0);
11373: }
1.208 brouard 11374: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11375: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11376:
1.145 brouard 11377: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11378: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11379:
1.235 brouard 11380: i1=pow(2,cptcoveff);
11381: if (cptcovn < 1){i1=1;}
11382:
11383: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11384: for(k=1; k<=i1;k++){
11385: if(TKresult[nres]!= k)
11386: continue;
1.227 brouard 11387: fprintf(ficresvpl,"\n#****** ");
11388: printf("\n#****** ");
11389: fprintf(ficlog,"\n#****** ");
11390: for(j=1;j<=cptcoveff;j++) {
11391: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11392: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11393: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11394: }
1.235 brouard 11395: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11396: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11397: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11398: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11399: }
1.227 brouard 11400: fprintf(ficresvpl,"******\n");
11401: printf("******\n");
11402: fprintf(ficlog,"******\n");
11403:
11404: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11405: oldm=oldms;savm=savms;
1.235 brouard 11406: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11407: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11408: /*}*/
1.126 brouard 11409: }
1.227 brouard 11410:
1.126 brouard 11411: fclose(ficresvpl);
1.208 brouard 11412: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11413: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11414:
11415: free_vector(weight,1,n);
11416: free_imatrix(Tvard,1,NCOVMAX,1,2);
11417: free_imatrix(s,1,maxwav+1,1,n);
11418: free_matrix(anint,1,maxwav,1,n);
11419: free_matrix(mint,1,maxwav,1,n);
11420: free_ivector(cod,1,n);
11421: free_ivector(tab,1,NCOVMAX);
11422: fclose(ficresstdeij);
11423: fclose(ficrescveij);
11424: fclose(ficresvij);
11425: fclose(ficrest);
11426: fclose(ficpar);
11427:
11428:
1.126 brouard 11429: /*---------- End : free ----------------*/
1.219 brouard 11430: if (mobilav!=0 ||mobilavproj !=0)
11431: 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 11432: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11433: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11434: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11435: } /* mle==-3 arrives here for freeing */
1.227 brouard 11436: /* endfree:*/
11437: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11438: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11439: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11440: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11441: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11442: free_matrix(coqvar,1,maxwav,1,n);
11443: free_matrix(covar,0,NCOVMAX,1,n);
11444: free_matrix(matcov,1,npar,1,npar);
11445: free_matrix(hess,1,npar,1,npar);
11446: /*free_vector(delti,1,npar);*/
11447: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11448: free_matrix(agev,1,maxwav,1,imx);
11449: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11450:
11451: free_ivector(ncodemax,1,NCOVMAX);
11452: free_ivector(ncodemaxwundef,1,NCOVMAX);
11453: free_ivector(Dummy,-1,NCOVMAX);
11454: free_ivector(Fixed,-1,NCOVMAX);
1.238 brouard 11455: free_ivector(DummyV,1,NCOVMAX);
11456: free_ivector(FixedV,1,NCOVMAX);
1.227 brouard 11457: free_ivector(Typevar,-1,NCOVMAX);
11458: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11459: free_ivector(TvarsQ,1,NCOVMAX);
11460: free_ivector(TvarsQind,1,NCOVMAX);
11461: free_ivector(TvarsD,1,NCOVMAX);
11462: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11463: free_ivector(TvarFD,1,NCOVMAX);
11464: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11465: free_ivector(TvarF,1,NCOVMAX);
11466: free_ivector(TvarFind,1,NCOVMAX);
11467: free_ivector(TvarV,1,NCOVMAX);
11468: free_ivector(TvarVind,1,NCOVMAX);
11469: free_ivector(TvarA,1,NCOVMAX);
11470: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11471: free_ivector(TvarFQ,1,NCOVMAX);
11472: free_ivector(TvarFQind,1,NCOVMAX);
11473: free_ivector(TvarVD,1,NCOVMAX);
11474: free_ivector(TvarVDind,1,NCOVMAX);
11475: free_ivector(TvarVQ,1,NCOVMAX);
11476: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11477: free_ivector(Tvarsel,1,NCOVMAX);
11478: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11479: free_ivector(Tposprod,1,NCOVMAX);
11480: free_ivector(Tprod,1,NCOVMAX);
11481: free_ivector(Tvaraff,1,NCOVMAX);
11482: free_ivector(invalidvarcomb,1,ncovcombmax);
11483: free_ivector(Tage,1,NCOVMAX);
11484: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11485: free_ivector(TmodelInvind,1,NCOVMAX);
11486: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11487:
11488: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11489: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11490: fflush(fichtm);
11491: fflush(ficgp);
11492:
1.227 brouard 11493:
1.126 brouard 11494: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11495: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11496: 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 11497: }else{
11498: printf("End of Imach\n");
11499: fprintf(ficlog,"End of Imach\n");
11500: }
11501: printf("See log file on %s\n",filelog);
11502: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11503: /*(void) gettimeofday(&end_time,&tzp);*/
11504: rend_time = time(NULL);
11505: end_time = *localtime(&rend_time);
11506: /* tml = *localtime(&end_time.tm_sec); */
11507: strcpy(strtend,asctime(&end_time));
1.126 brouard 11508: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11509: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11510: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11511:
1.157 brouard 11512: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11513: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11514: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11515: /* printf("Total time was %d uSec.\n", total_usecs);*/
11516: /* if(fileappend(fichtm,optionfilehtm)){ */
11517: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11518: fclose(fichtm);
11519: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11520: fclose(fichtmcov);
11521: fclose(ficgp);
11522: fclose(ficlog);
11523: /*------ End -----------*/
1.227 brouard 11524:
11525:
11526: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11527: #ifdef WIN32
1.227 brouard 11528: if (_chdir(pathcd) != 0)
11529: printf("Can't move to directory %s!\n",path);
11530: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11531: #else
1.227 brouard 11532: if(chdir(pathcd) != 0)
11533: printf("Can't move to directory %s!\n", path);
11534: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11535: #endif
1.126 brouard 11536: printf("Current directory %s!\n",pathcd);
11537: /*strcat(plotcmd,CHARSEPARATOR);*/
11538: sprintf(plotcmd,"gnuplot");
1.157 brouard 11539: #ifdef _WIN32
1.126 brouard 11540: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11541: #endif
11542: if(!stat(plotcmd,&info)){
1.158 brouard 11543: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11544: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11545: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11546: }else
11547: strcpy(pplotcmd,plotcmd);
1.157 brouard 11548: #ifdef __unix
1.126 brouard 11549: strcpy(plotcmd,GNUPLOTPROGRAM);
11550: if(!stat(plotcmd,&info)){
1.158 brouard 11551: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11552: }else
11553: strcpy(pplotcmd,plotcmd);
11554: #endif
11555: }else
11556: strcpy(pplotcmd,plotcmd);
11557:
11558: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 11559: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 11560:
1.126 brouard 11561: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 11562: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 11563: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 11564: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 11565: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 11566: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 11567: }
1.158 brouard 11568: printf(" Successful, please wait...");
1.126 brouard 11569: while (z[0] != 'q') {
11570: /* chdir(path); */
1.154 brouard 11571: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 11572: scanf("%s",z);
11573: /* if (z[0] == 'c') system("./imach"); */
11574: if (z[0] == 'e') {
1.158 brouard 11575: #ifdef __APPLE__
1.152 brouard 11576: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 11577: #elif __linux
11578: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 11579: #else
1.152 brouard 11580: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 11581: #endif
11582: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
11583: system(pplotcmd);
1.126 brouard 11584: }
11585: else if (z[0] == 'g') system(plotcmd);
11586: else if (z[0] == 'q') exit(0);
11587: }
1.227 brouard 11588: end:
1.126 brouard 11589: while (z[0] != 'q') {
1.195 brouard 11590: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 11591: scanf("%s",z);
11592: }
11593: }
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