Annotation of imach/src/imach.c, revision 1.236
1.236 ! brouard 1: /* $Id: imach.c,v 1.235 2016/08/25 06:59:23 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.236 ! brouard 4: Revision 1.235 2016/08/25 06:59:23 brouard
! 5: *** empty log message ***
! 6:
1.235 brouard 7: Revision 1.234 2016/08/23 16:51:20 brouard
8: *** empty log message ***
9:
1.234 brouard 10: Revision 1.233 2016/08/23 07:40:50 brouard
11: Summary: not working
12:
1.233 brouard 13: Revision 1.232 2016/08/22 14:20:21 brouard
14: Summary: not working
15:
1.232 brouard 16: Revision 1.231 2016/08/22 07:17:15 brouard
17: Summary: not working
18:
1.231 brouard 19: Revision 1.230 2016/08/22 06:55:53 brouard
20: Summary: Not working
21:
1.230 brouard 22: Revision 1.229 2016/07/23 09:45:53 brouard
23: Summary: Completing for func too
24:
1.229 brouard 25: Revision 1.228 2016/07/22 17:45:30 brouard
26: Summary: Fixing some arrays, still debugging
27:
1.227 brouard 28: Revision 1.226 2016/07/12 18:42:34 brouard
29: Summary: temp
30:
1.226 brouard 31: Revision 1.225 2016/07/12 08:40:03 brouard
32: Summary: saving but not running
33:
1.225 brouard 34: Revision 1.224 2016/07/01 13:16:01 brouard
35: Summary: Fixes
36:
1.224 brouard 37: Revision 1.223 2016/02/19 09:23:35 brouard
38: Summary: temporary
39:
1.223 brouard 40: Revision 1.222 2016/02/17 08:14:50 brouard
41: Summary: Probably last 0.98 stable version 0.98r6
42:
1.222 brouard 43: Revision 1.221 2016/02/15 23:35:36 brouard
44: Summary: minor bug
45:
1.220 brouard 46: Revision 1.219 2016/02/15 00:48:12 brouard
47: *** empty log message ***
48:
1.219 brouard 49: Revision 1.218 2016/02/12 11:29:23 brouard
50: Summary: 0.99 Back projections
51:
1.218 brouard 52: Revision 1.217 2015/12/23 17:18:31 brouard
53: Summary: Experimental backcast
54:
1.217 brouard 55: Revision 1.216 2015/12/18 17:32:11 brouard
56: Summary: 0.98r4 Warning and status=-2
57:
58: Version 0.98r4 is now:
59: - displaying an error when status is -1, date of interview unknown and date of death known;
60: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
61: Older changes concerning s=-2, dating from 2005 have been supersed.
62:
1.216 brouard 63: Revision 1.215 2015/12/16 08:52:24 brouard
64: Summary: 0.98r4 working
65:
1.215 brouard 66: Revision 1.214 2015/12/16 06:57:54 brouard
67: Summary: temporary not working
68:
1.214 brouard 69: Revision 1.213 2015/12/11 18:22:17 brouard
70: Summary: 0.98r4
71:
1.213 brouard 72: Revision 1.212 2015/11/21 12:47:24 brouard
73: Summary: minor typo
74:
1.212 brouard 75: Revision 1.211 2015/11/21 12:41:11 brouard
76: Summary: 0.98r3 with some graph of projected cross-sectional
77:
78: Author: Nicolas Brouard
79:
1.211 brouard 80: Revision 1.210 2015/11/18 17:41:20 brouard
81: Summary: Start working on projected prevalences
82:
1.210 brouard 83: Revision 1.209 2015/11/17 22:12:03 brouard
84: Summary: Adding ftolpl parameter
85: Author: N Brouard
86:
87: We had difficulties to get smoothed confidence intervals. It was due
88: to the period prevalence which wasn't computed accurately. The inner
89: parameter ftolpl is now an outer parameter of the .imach parameter
90: file after estepm. If ftolpl is small 1.e-4 and estepm too,
91: computation are long.
92:
1.209 brouard 93: Revision 1.208 2015/11/17 14:31:57 brouard
94: Summary: temporary
95:
1.208 brouard 96: Revision 1.207 2015/10/27 17:36:57 brouard
97: *** empty log message ***
98:
1.207 brouard 99: Revision 1.206 2015/10/24 07:14:11 brouard
100: *** empty log message ***
101:
1.206 brouard 102: Revision 1.205 2015/10/23 15:50:53 brouard
103: Summary: 0.98r3 some clarification for graphs on likelihood contributions
104:
1.205 brouard 105: Revision 1.204 2015/10/01 16:20:26 brouard
106: Summary: Some new graphs of contribution to likelihood
107:
1.204 brouard 108: Revision 1.203 2015/09/30 17:45:14 brouard
109: Summary: looking at better estimation of the hessian
110:
111: Also a better criteria for convergence to the period prevalence And
112: therefore adding the number of years needed to converge. (The
113: prevalence in any alive state shold sum to one
114:
1.203 brouard 115: Revision 1.202 2015/09/22 19:45:16 brouard
116: Summary: Adding some overall graph on contribution to likelihood. Might change
117:
1.202 brouard 118: Revision 1.201 2015/09/15 17:34:58 brouard
119: Summary: 0.98r0
120:
121: - Some new graphs like suvival functions
122: - Some bugs fixed like model=1+age+V2.
123:
1.201 brouard 124: Revision 1.200 2015/09/09 16:53:55 brouard
125: Summary: Big bug thanks to Flavia
126:
127: Even model=1+age+V2. did not work anymore
128:
1.200 brouard 129: Revision 1.199 2015/09/07 14:09:23 brouard
130: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
131:
1.199 brouard 132: Revision 1.198 2015/09/03 07:14:39 brouard
133: Summary: 0.98q5 Flavia
134:
1.198 brouard 135: Revision 1.197 2015/09/01 18:24:39 brouard
136: *** empty log message ***
137:
1.197 brouard 138: Revision 1.196 2015/08/18 23:17:52 brouard
139: Summary: 0.98q5
140:
1.196 brouard 141: Revision 1.195 2015/08/18 16:28:39 brouard
142: Summary: Adding a hack for testing purpose
143:
144: After reading the title, ftol and model lines, if the comment line has
145: a q, starting with #q, the answer at the end of the run is quit. It
146: permits to run test files in batch with ctest. The former workaround was
147: $ echo q | imach foo.imach
148:
1.195 brouard 149: Revision 1.194 2015/08/18 13:32:00 brouard
150: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
151:
1.194 brouard 152: Revision 1.193 2015/08/04 07:17:42 brouard
153: Summary: 0.98q4
154:
1.193 brouard 155: Revision 1.192 2015/07/16 16:49:02 brouard
156: Summary: Fixing some outputs
157:
1.192 brouard 158: Revision 1.191 2015/07/14 10:00:33 brouard
159: Summary: Some fixes
160:
1.191 brouard 161: Revision 1.190 2015/05/05 08:51:13 brouard
162: Summary: Adding digits in output parameters (7 digits instead of 6)
163:
164: Fix 1+age+.
165:
1.190 brouard 166: Revision 1.189 2015/04/30 14:45:16 brouard
167: Summary: 0.98q2
168:
1.189 brouard 169: Revision 1.188 2015/04/30 08:27:53 brouard
170: *** empty log message ***
171:
1.188 brouard 172: Revision 1.187 2015/04/29 09:11:15 brouard
173: *** empty log message ***
174:
1.187 brouard 175: Revision 1.186 2015/04/23 12:01:52 brouard
176: Summary: V1*age is working now, version 0.98q1
177:
178: Some codes had been disabled in order to simplify and Vn*age was
179: working in the optimization phase, ie, giving correct MLE parameters,
180: but, as usual, outputs were not correct and program core dumped.
181:
1.186 brouard 182: Revision 1.185 2015/03/11 13:26:42 brouard
183: Summary: Inclusion of compile and links command line for Intel Compiler
184:
1.185 brouard 185: Revision 1.184 2015/03/11 11:52:39 brouard
186: Summary: Back from Windows 8. Intel Compiler
187:
1.184 brouard 188: Revision 1.183 2015/03/10 20:34:32 brouard
189: Summary: 0.98q0, trying with directest, mnbrak fixed
190:
191: We use directest instead of original Powell test; probably no
192: incidence on the results, but better justifications;
193: We fixed Numerical Recipes mnbrak routine which was wrong and gave
194: wrong results.
195:
1.183 brouard 196: Revision 1.182 2015/02/12 08:19:57 brouard
197: Summary: Trying to keep directest which seems simpler and more general
198: Author: Nicolas Brouard
199:
1.182 brouard 200: Revision 1.181 2015/02/11 23:22:24 brouard
201: Summary: Comments on Powell added
202:
203: Author:
204:
1.181 brouard 205: Revision 1.180 2015/02/11 17:33:45 brouard
206: Summary: Finishing move from main to function (hpijx and prevalence_limit)
207:
1.180 brouard 208: Revision 1.179 2015/01/04 09:57:06 brouard
209: Summary: back to OS/X
210:
1.179 brouard 211: Revision 1.178 2015/01/04 09:35:48 brouard
212: *** empty log message ***
213:
1.178 brouard 214: Revision 1.177 2015/01/03 18:40:56 brouard
215: Summary: Still testing ilc32 on OSX
216:
1.177 brouard 217: Revision 1.176 2015/01/03 16:45:04 brouard
218: *** empty log message ***
219:
1.176 brouard 220: Revision 1.175 2015/01/03 16:33:42 brouard
221: *** empty log message ***
222:
1.175 brouard 223: Revision 1.174 2015/01/03 16:15:49 brouard
224: Summary: Still in cross-compilation
225:
1.174 brouard 226: Revision 1.173 2015/01/03 12:06:26 brouard
227: Summary: trying to detect cross-compilation
228:
1.173 brouard 229: Revision 1.172 2014/12/27 12:07:47 brouard
230: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
231:
1.172 brouard 232: Revision 1.171 2014/12/23 13:26:59 brouard
233: Summary: Back from Visual C
234:
235: Still problem with utsname.h on Windows
236:
1.171 brouard 237: Revision 1.170 2014/12/23 11:17:12 brouard
238: Summary: Cleaning some \%% back to %%
239:
240: The escape was mandatory for a specific compiler (which one?), but too many warnings.
241:
1.170 brouard 242: Revision 1.169 2014/12/22 23:08:31 brouard
243: Summary: 0.98p
244:
245: Outputs some informations on compiler used, OS etc. Testing on different platforms.
246:
1.169 brouard 247: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 248: Summary: update
1.169 brouard 249:
1.168 brouard 250: Revision 1.167 2014/12/22 13:50:56 brouard
251: Summary: Testing uname and compiler version and if compiled 32 or 64
252:
253: Testing on Linux 64
254:
1.167 brouard 255: Revision 1.166 2014/12/22 11:40:47 brouard
256: *** empty log message ***
257:
1.166 brouard 258: Revision 1.165 2014/12/16 11:20:36 brouard
259: Summary: After compiling on Visual C
260:
261: * imach.c (Module): Merging 1.61 to 1.162
262:
1.165 brouard 263: Revision 1.164 2014/12/16 10:52:11 brouard
264: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
265:
266: * imach.c (Module): Merging 1.61 to 1.162
267:
1.164 brouard 268: Revision 1.163 2014/12/16 10:30:11 brouard
269: * imach.c (Module): Merging 1.61 to 1.162
270:
1.163 brouard 271: Revision 1.162 2014/09/25 11:43:39 brouard
272: Summary: temporary backup 0.99!
273:
1.162 brouard 274: Revision 1.1 2014/09/16 11:06:58 brouard
275: Summary: With some code (wrong) for nlopt
276:
277: Author:
278:
279: Revision 1.161 2014/09/15 20:41:41 brouard
280: Summary: Problem with macro SQR on Intel compiler
281:
1.161 brouard 282: Revision 1.160 2014/09/02 09:24:05 brouard
283: *** empty log message ***
284:
1.160 brouard 285: Revision 1.159 2014/09/01 10:34:10 brouard
286: Summary: WIN32
287: Author: Brouard
288:
1.159 brouard 289: Revision 1.158 2014/08/27 17:11:51 brouard
290: *** empty log message ***
291:
1.158 brouard 292: Revision 1.157 2014/08/27 16:26:55 brouard
293: Summary: Preparing windows Visual studio version
294: Author: Brouard
295:
296: In order to compile on Visual studio, time.h is now correct and time_t
297: and tm struct should be used. difftime should be used but sometimes I
298: just make the differences in raw time format (time(&now).
299: Trying to suppress #ifdef LINUX
300: Add xdg-open for __linux in order to open default browser.
301:
1.157 brouard 302: Revision 1.156 2014/08/25 20:10:10 brouard
303: *** empty log message ***
304:
1.156 brouard 305: Revision 1.155 2014/08/25 18:32:34 brouard
306: Summary: New compile, minor changes
307: Author: Brouard
308:
1.155 brouard 309: Revision 1.154 2014/06/20 17:32:08 brouard
310: Summary: Outputs now all graphs of convergence to period prevalence
311:
1.154 brouard 312: Revision 1.153 2014/06/20 16:45:46 brouard
313: Summary: If 3 live state, convergence to period prevalence on same graph
314: Author: Brouard
315:
1.153 brouard 316: Revision 1.152 2014/06/18 17:54:09 brouard
317: Summary: open browser, use gnuplot on same dir than imach if not found in the path
318:
1.152 brouard 319: Revision 1.151 2014/06/18 16:43:30 brouard
320: *** empty log message ***
321:
1.151 brouard 322: Revision 1.150 2014/06/18 16:42:35 brouard
323: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
324: Author: brouard
325:
1.150 brouard 326: Revision 1.149 2014/06/18 15:51:14 brouard
327: Summary: Some fixes in parameter files errors
328: Author: Nicolas Brouard
329:
1.149 brouard 330: Revision 1.148 2014/06/17 17:38:48 brouard
331: Summary: Nothing new
332: Author: Brouard
333:
334: Just a new packaging for OS/X version 0.98nS
335:
1.148 brouard 336: Revision 1.147 2014/06/16 10:33:11 brouard
337: *** empty log message ***
338:
1.147 brouard 339: Revision 1.146 2014/06/16 10:20:28 brouard
340: Summary: Merge
341: Author: Brouard
342:
343: Merge, before building revised version.
344:
1.146 brouard 345: Revision 1.145 2014/06/10 21:23:15 brouard
346: Summary: Debugging with valgrind
347: Author: Nicolas Brouard
348:
349: Lot of changes in order to output the results with some covariates
350: After the Edimburgh REVES conference 2014, it seems mandatory to
351: improve the code.
352: No more memory valgrind error but a lot has to be done in order to
353: continue the work of splitting the code into subroutines.
354: Also, decodemodel has been improved. Tricode is still not
355: optimal. nbcode should be improved. Documentation has been added in
356: the source code.
357:
1.144 brouard 358: Revision 1.143 2014/01/26 09:45:38 brouard
359: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
360:
361: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
362: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
363:
1.143 brouard 364: Revision 1.142 2014/01/26 03:57:36 brouard
365: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
366:
367: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
368:
1.142 brouard 369: Revision 1.141 2014/01/26 02:42:01 brouard
370: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
371:
1.141 brouard 372: Revision 1.140 2011/09/02 10:37:54 brouard
373: Summary: times.h is ok with mingw32 now.
374:
1.140 brouard 375: Revision 1.139 2010/06/14 07:50:17 brouard
376: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
377: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
378:
1.139 brouard 379: Revision 1.138 2010/04/30 18:19:40 brouard
380: *** empty log message ***
381:
1.138 brouard 382: Revision 1.137 2010/04/29 18:11:38 brouard
383: (Module): Checking covariates for more complex models
384: than V1+V2. A lot of change to be done. Unstable.
385:
1.137 brouard 386: Revision 1.136 2010/04/26 20:30:53 brouard
387: (Module): merging some libgsl code. Fixing computation
388: of likelione (using inter/intrapolation if mle = 0) in order to
389: get same likelihood as if mle=1.
390: Some cleaning of code and comments added.
391:
1.136 brouard 392: Revision 1.135 2009/10/29 15:33:14 brouard
393: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
394:
1.135 brouard 395: Revision 1.134 2009/10/29 13:18:53 brouard
396: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
397:
1.134 brouard 398: Revision 1.133 2009/07/06 10:21:25 brouard
399: just nforces
400:
1.133 brouard 401: Revision 1.132 2009/07/06 08:22:05 brouard
402: Many tings
403:
1.132 brouard 404: Revision 1.131 2009/06/20 16:22:47 brouard
405: Some dimensions resccaled
406:
1.131 brouard 407: Revision 1.130 2009/05/26 06:44:34 brouard
408: (Module): Max Covariate is now set to 20 instead of 8. A
409: lot of cleaning with variables initialized to 0. Trying to make
410: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
411:
1.130 brouard 412: Revision 1.129 2007/08/31 13:49:27 lievre
413: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
414:
1.129 lievre 415: Revision 1.128 2006/06/30 13:02:05 brouard
416: (Module): Clarifications on computing e.j
417:
1.128 brouard 418: Revision 1.127 2006/04/28 18:11:50 brouard
419: (Module): Yes the sum of survivors was wrong since
420: imach-114 because nhstepm was no more computed in the age
421: loop. Now we define nhstepma in the age loop.
422: (Module): In order to speed up (in case of numerous covariates) we
423: compute health expectancies (without variances) in a first step
424: and then all the health expectancies with variances or standard
425: deviation (needs data from the Hessian matrices) which slows the
426: computation.
427: In the future we should be able to stop the program is only health
428: expectancies and graph are needed without standard deviations.
429:
1.127 brouard 430: Revision 1.126 2006/04/28 17:23:28 brouard
431: (Module): Yes the sum of survivors was wrong since
432: imach-114 because nhstepm was no more computed in the age
433: loop. Now we define nhstepma in the age loop.
434: Version 0.98h
435:
1.126 brouard 436: Revision 1.125 2006/04/04 15:20:31 lievre
437: Errors in calculation of health expectancies. Age was not initialized.
438: Forecasting file added.
439:
440: Revision 1.124 2006/03/22 17:13:53 lievre
441: Parameters are printed with %lf instead of %f (more numbers after the comma).
442: The log-likelihood is printed in the log file
443:
444: Revision 1.123 2006/03/20 10:52:43 brouard
445: * imach.c (Module): <title> changed, corresponds to .htm file
446: name. <head> headers where missing.
447:
448: * imach.c (Module): Weights can have a decimal point as for
449: English (a comma might work with a correct LC_NUMERIC environment,
450: otherwise the weight is truncated).
451: Modification of warning when the covariates values are not 0 or
452: 1.
453: Version 0.98g
454:
455: Revision 1.122 2006/03/20 09:45:41 brouard
456: (Module): Weights can have a decimal point as for
457: English (a comma might work with a correct LC_NUMERIC environment,
458: otherwise the weight is truncated).
459: Modification of warning when the covariates values are not 0 or
460: 1.
461: Version 0.98g
462:
463: Revision 1.121 2006/03/16 17:45:01 lievre
464: * imach.c (Module): Comments concerning covariates added
465:
466: * imach.c (Module): refinements in the computation of lli if
467: status=-2 in order to have more reliable computation if stepm is
468: not 1 month. Version 0.98f
469:
470: Revision 1.120 2006/03/16 15:10:38 lievre
471: (Module): refinements in the computation of lli if
472: status=-2 in order to have more reliable computation if stepm is
473: not 1 month. Version 0.98f
474:
475: Revision 1.119 2006/03/15 17:42:26 brouard
476: (Module): Bug if status = -2, the loglikelihood was
477: computed as likelihood omitting the logarithm. Version O.98e
478:
479: Revision 1.118 2006/03/14 18:20:07 brouard
480: (Module): varevsij Comments added explaining the second
481: table of variances if popbased=1 .
482: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
483: (Module): Function pstamp added
484: (Module): Version 0.98d
485:
486: Revision 1.117 2006/03/14 17:16:22 brouard
487: (Module): varevsij Comments added explaining the second
488: table of variances if popbased=1 .
489: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
490: (Module): Function pstamp added
491: (Module): Version 0.98d
492:
493: Revision 1.116 2006/03/06 10:29:27 brouard
494: (Module): Variance-covariance wrong links and
495: varian-covariance of ej. is needed (Saito).
496:
497: Revision 1.115 2006/02/27 12:17:45 brouard
498: (Module): One freematrix added in mlikeli! 0.98c
499:
500: Revision 1.114 2006/02/26 12:57:58 brouard
501: (Module): Some improvements in processing parameter
502: filename with strsep.
503:
504: Revision 1.113 2006/02/24 14:20:24 brouard
505: (Module): Memory leaks checks with valgrind and:
506: datafile was not closed, some imatrix were not freed and on matrix
507: allocation too.
508:
509: Revision 1.112 2006/01/30 09:55:26 brouard
510: (Module): Back to gnuplot.exe instead of wgnuplot.exe
511:
512: Revision 1.111 2006/01/25 20:38:18 brouard
513: (Module): Lots of cleaning and bugs added (Gompertz)
514: (Module): Comments can be added in data file. Missing date values
515: can be a simple dot '.'.
516:
517: Revision 1.110 2006/01/25 00:51:50 brouard
518: (Module): Lots of cleaning and bugs added (Gompertz)
519:
520: Revision 1.109 2006/01/24 19:37:15 brouard
521: (Module): Comments (lines starting with a #) are allowed in data.
522:
523: Revision 1.108 2006/01/19 18:05:42 lievre
524: Gnuplot problem appeared...
525: To be fixed
526:
527: Revision 1.107 2006/01/19 16:20:37 brouard
528: Test existence of gnuplot in imach path
529:
530: Revision 1.106 2006/01/19 13:24:36 brouard
531: Some cleaning and links added in html output
532:
533: Revision 1.105 2006/01/05 20:23:19 lievre
534: *** empty log message ***
535:
536: Revision 1.104 2005/09/30 16:11:43 lievre
537: (Module): sump fixed, loop imx fixed, and simplifications.
538: (Module): If the status is missing at the last wave but we know
539: that the person is alive, then we can code his/her status as -2
540: (instead of missing=-1 in earlier versions) and his/her
541: contributions to the likelihood is 1 - Prob of dying from last
542: health status (= 1-p13= p11+p12 in the easiest case of somebody in
543: the healthy state at last known wave). Version is 0.98
544:
545: Revision 1.103 2005/09/30 15:54:49 lievre
546: (Module): sump fixed, loop imx fixed, and simplifications.
547:
548: Revision 1.102 2004/09/15 17:31:30 brouard
549: Add the possibility to read data file including tab characters.
550:
551: Revision 1.101 2004/09/15 10:38:38 brouard
552: Fix on curr_time
553:
554: Revision 1.100 2004/07/12 18:29:06 brouard
555: Add version for Mac OS X. Just define UNIX in Makefile
556:
557: Revision 1.99 2004/06/05 08:57:40 brouard
558: *** empty log message ***
559:
560: Revision 1.98 2004/05/16 15:05:56 brouard
561: New version 0.97 . First attempt to estimate force of mortality
562: directly from the data i.e. without the need of knowing the health
563: state at each age, but using a Gompertz model: log u =a + b*age .
564: This is the basic analysis of mortality and should be done before any
565: other analysis, in order to test if the mortality estimated from the
566: cross-longitudinal survey is different from the mortality estimated
567: from other sources like vital statistic data.
568:
569: The same imach parameter file can be used but the option for mle should be -3.
570:
1.133 brouard 571: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 572: former routines in order to include the new code within the former code.
573:
574: The output is very simple: only an estimate of the intercept and of
575: the slope with 95% confident intervals.
576:
577: Current limitations:
578: A) Even if you enter covariates, i.e. with the
579: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
580: B) There is no computation of Life Expectancy nor Life Table.
581:
582: Revision 1.97 2004/02/20 13:25:42 lievre
583: Version 0.96d. Population forecasting command line is (temporarily)
584: suppressed.
585:
586: Revision 1.96 2003/07/15 15:38:55 brouard
587: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
588: rewritten within the same printf. Workaround: many printfs.
589:
590: Revision 1.95 2003/07/08 07:54:34 brouard
591: * imach.c (Repository):
592: (Repository): Using imachwizard code to output a more meaningful covariance
593: matrix (cov(a12,c31) instead of numbers.
594:
595: Revision 1.94 2003/06/27 13:00:02 brouard
596: Just cleaning
597:
598: Revision 1.93 2003/06/25 16:33:55 brouard
599: (Module): On windows (cygwin) function asctime_r doesn't
600: exist so I changed back to asctime which exists.
601: (Module): Version 0.96b
602:
603: Revision 1.92 2003/06/25 16:30:45 brouard
604: (Module): On windows (cygwin) function asctime_r doesn't
605: exist so I changed back to asctime which exists.
606:
607: Revision 1.91 2003/06/25 15:30:29 brouard
608: * imach.c (Repository): Duplicated warning errors corrected.
609: (Repository): Elapsed time after each iteration is now output. It
610: helps to forecast when convergence will be reached. Elapsed time
611: is stamped in powell. We created a new html file for the graphs
612: concerning matrix of covariance. It has extension -cov.htm.
613:
614: Revision 1.90 2003/06/24 12:34:15 brouard
615: (Module): Some bugs corrected for windows. Also, when
616: mle=-1 a template is output in file "or"mypar.txt with the design
617: of the covariance matrix to be input.
618:
619: Revision 1.89 2003/06/24 12:30:52 brouard
620: (Module): Some bugs corrected for windows. Also, when
621: mle=-1 a template is output in file "or"mypar.txt with the design
622: of the covariance matrix to be input.
623:
624: Revision 1.88 2003/06/23 17:54:56 brouard
625: * 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.
626:
627: Revision 1.87 2003/06/18 12:26:01 brouard
628: Version 0.96
629:
630: Revision 1.86 2003/06/17 20:04:08 brouard
631: (Module): Change position of html and gnuplot routines and added
632: routine fileappend.
633:
634: Revision 1.85 2003/06/17 13:12:43 brouard
635: * imach.c (Repository): Check when date of death was earlier that
636: current date of interview. It may happen when the death was just
637: prior to the death. In this case, dh was negative and likelihood
638: was wrong (infinity). We still send an "Error" but patch by
639: assuming that the date of death was just one stepm after the
640: interview.
641: (Repository): Because some people have very long ID (first column)
642: we changed int to long in num[] and we added a new lvector for
643: memory allocation. But we also truncated to 8 characters (left
644: truncation)
645: (Repository): No more line truncation errors.
646:
647: Revision 1.84 2003/06/13 21:44:43 brouard
648: * imach.c (Repository): Replace "freqsummary" at a correct
649: place. It differs from routine "prevalence" which may be called
650: many times. Probs is memory consuming and must be used with
651: parcimony.
652: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
653:
654: Revision 1.83 2003/06/10 13:39:11 lievre
655: *** empty log message ***
656:
657: Revision 1.82 2003/06/05 15:57:20 brouard
658: Add log in imach.c and fullversion number is now printed.
659:
660: */
661: /*
662: Interpolated Markov Chain
663:
664: Short summary of the programme:
665:
1.227 brouard 666: This program computes Healthy Life Expectancies or State-specific
667: (if states aren't health statuses) Expectancies from
668: cross-longitudinal data. Cross-longitudinal data consist in:
669:
670: -1- a first survey ("cross") where individuals from different ages
671: are interviewed on their health status or degree of disability (in
672: the case of a health survey which is our main interest)
673:
674: -2- at least a second wave of interviews ("longitudinal") which
675: measure each change (if any) in individual health status. Health
676: expectancies are computed from the time spent in each health state
677: according to a model. More health states you consider, more time is
678: necessary to reach the Maximum Likelihood of the parameters involved
679: in the model. The simplest model is the multinomial logistic model
680: where pij is the probability to be observed in state j at the second
681: wave conditional to be observed in state i at the first
682: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
683: etc , where 'age' is age and 'sex' is a covariate. If you want to
684: have a more complex model than "constant and age", you should modify
685: the program where the markup *Covariates have to be included here
686: again* invites you to do it. More covariates you add, slower the
1.126 brouard 687: convergence.
688:
689: The advantage of this computer programme, compared to a simple
690: multinomial logistic model, is clear when the delay between waves is not
691: identical for each individual. Also, if a individual missed an
692: intermediate interview, the information is lost, but taken into
693: account using an interpolation or extrapolation.
694:
695: hPijx is the probability to be observed in state i at age x+h
696: conditional to the observed state i at age x. The delay 'h' can be
697: split into an exact number (nh*stepm) of unobserved intermediate
698: states. This elementary transition (by month, quarter,
699: semester or year) is modelled as a multinomial logistic. The hPx
700: matrix is simply the matrix product of nh*stepm elementary matrices
701: and the contribution of each individual to the likelihood is simply
702: hPijx.
703:
704: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 705: of the life expectancies. It also computes the period (stable) prevalence.
706:
707: Back prevalence and projections:
1.227 brouard 708:
709: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
710: double agemaxpar, double ftolpl, int *ncvyearp, double
711: dateprev1,double dateprev2, int firstpass, int lastpass, int
712: mobilavproj)
713:
714: Computes the back prevalence limit for any combination of
715: covariate values k at any age between ageminpar and agemaxpar and
716: returns it in **bprlim. In the loops,
717:
718: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
719: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
720:
721: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 722: Computes for any combination of covariates k and any age between bage and fage
723: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
724: oldm=oldms;savm=savms;
1.227 brouard 725:
726: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 727: Computes the transition matrix starting at age 'age' over
728: 'nhstepm*hstepm*stepm' months (i.e. until
729: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 730: nhstepm*hstepm matrices.
731:
732: Returns p3mat[i][j][h] after calling
733: p3mat[i][j][h]=matprod2(newm,
734: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
735: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
736: oldm);
1.226 brouard 737:
738: Important routines
739:
740: - func (or funcone), computes logit (pij) distinguishing
741: o fixed variables (single or product dummies or quantitative);
742: o varying variables by:
743: (1) wave (single, product dummies, quantitative),
744: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
745: % fixed dummy (treated) or quantitative (not done because time-consuming);
746: % varying dummy (not done) or quantitative (not done);
747: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
748: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
749: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
750: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
751: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 752:
1.226 brouard 753:
754:
1.133 brouard 755: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
756: Institut national d'études démographiques, Paris.
1.126 brouard 757: This software have been partly granted by Euro-REVES, a concerted action
758: from the European Union.
759: It is copyrighted identically to a GNU software product, ie programme and
760: software can be distributed freely for non commercial use. Latest version
761: can be accessed at http://euroreves.ined.fr/imach .
762:
763: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
764: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
765:
766: **********************************************************************/
767: /*
768: main
769: read parameterfile
770: read datafile
771: concatwav
772: freqsummary
773: if (mle >= 1)
774: mlikeli
775: print results files
776: if mle==1
777: computes hessian
778: read end of parameter file: agemin, agemax, bage, fage, estepm
779: begin-prev-date,...
780: open gnuplot file
781: open html file
1.145 brouard 782: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
783: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
784: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
785: freexexit2 possible for memory heap.
786:
787: h Pij x | pij_nom ficrestpij
788: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
789: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
790: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
791:
792: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
793: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
794: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
795: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
796: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
797:
1.126 brouard 798: forecasting if prevfcast==1 prevforecast call prevalence()
799: health expectancies
800: Variance-covariance of DFLE
801: prevalence()
802: movingaverage()
803: varevsij()
804: if popbased==1 varevsij(,popbased)
805: total life expectancies
806: Variance of period (stable) prevalence
807: end
808: */
809:
1.187 brouard 810: /* #define DEBUG */
811: /* #define DEBUGBRENT */
1.203 brouard 812: /* #define DEBUGLINMIN */
813: /* #define DEBUGHESS */
814: #define DEBUGHESSIJ
1.224 brouard 815: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 816: #define POWELL /* Instead of NLOPT */
1.224 brouard 817: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 818: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
819: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 820:
821: #include <math.h>
822: #include <stdio.h>
823: #include <stdlib.h>
824: #include <string.h>
1.226 brouard 825: #include <ctype.h>
1.159 brouard 826:
827: #ifdef _WIN32
828: #include <io.h>
1.172 brouard 829: #include <windows.h>
830: #include <tchar.h>
1.159 brouard 831: #else
1.126 brouard 832: #include <unistd.h>
1.159 brouard 833: #endif
1.126 brouard 834:
835: #include <limits.h>
836: #include <sys/types.h>
1.171 brouard 837:
838: #if defined(__GNUC__)
839: #include <sys/utsname.h> /* Doesn't work on Windows */
840: #endif
841:
1.126 brouard 842: #include <sys/stat.h>
843: #include <errno.h>
1.159 brouard 844: /* extern int errno; */
1.126 brouard 845:
1.157 brouard 846: /* #ifdef LINUX */
847: /* #include <time.h> */
848: /* #include "timeval.h" */
849: /* #else */
850: /* #include <sys/time.h> */
851: /* #endif */
852:
1.126 brouard 853: #include <time.h>
854:
1.136 brouard 855: #ifdef GSL
856: #include <gsl/gsl_errno.h>
857: #include <gsl/gsl_multimin.h>
858: #endif
859:
1.167 brouard 860:
1.162 brouard 861: #ifdef NLOPT
862: #include <nlopt.h>
863: typedef struct {
864: double (* function)(double [] );
865: } myfunc_data ;
866: #endif
867:
1.126 brouard 868: /* #include <libintl.h> */
869: /* #define _(String) gettext (String) */
870:
1.141 brouard 871: #define MAXLINE 1024 /* Was 256. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 872:
873: #define GNUPLOTPROGRAM "gnuplot"
874: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
875: #define FILENAMELENGTH 132
876:
877: #define GLOCK_ERROR_NOPATH -1 /* empty path */
878: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
879:
1.144 brouard 880: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
881: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 882:
883: #define NINTERVMAX 8
1.144 brouard 884: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
885: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
886: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 887: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 888: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
889: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 890: #define MAXN 20000
1.144 brouard 891: #define YEARM 12. /**< Number of months per year */
1.218 brouard 892: /* #define AGESUP 130 */
893: #define AGESUP 150
894: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 895: #define AGEBASE 40
1.194 brouard 896: #define AGEOVERFLOW 1.e20
1.164 brouard 897: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 898: #ifdef _WIN32
899: #define DIRSEPARATOR '\\'
900: #define CHARSEPARATOR "\\"
901: #define ODIRSEPARATOR '/'
902: #else
1.126 brouard 903: #define DIRSEPARATOR '/'
904: #define CHARSEPARATOR "/"
905: #define ODIRSEPARATOR '\\'
906: #endif
907:
1.236 ! brouard 908: /* $Id: imach.c,v 1.235 2016/08/25 06:59:23 brouard Exp $ */
1.126 brouard 909: /* $State: Exp $ */
1.196 brouard 910: #include "version.h"
911: char version[]=__IMACH_VERSION__;
1.224 brouard 912: 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";
1.236 ! brouard 913: char fullversion[]="$Revision: 1.235 $ $Date: 2016/08/25 06:59:23 $";
1.126 brouard 914: char strstart[80];
915: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 916: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 917: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 918: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
919: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
920: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 921: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
922: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 923: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
924: int cptcovprodnoage=0; /**< Number of covariate products without age */
925: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 926: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
927: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 928: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 929: int nsd=0; /**< Total number of single dummy variables (output) */
930: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 931: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 932: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 933: int ntveff=0; /**< ntveff number of effective time varying variables */
934: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 935: int cptcov=0; /* Working variable */
1.218 brouard 936: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 937: int npar=NPARMAX;
938: int nlstate=2; /* Number of live states */
939: int ndeath=1; /* Number of dead states */
1.130 brouard 940: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 941: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 942: int popbased=0;
943:
944: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 945: int maxwav=0; /* Maxim number of waves */
946: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
947: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
948: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 949: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 950: int mle=1, weightopt=0;
1.126 brouard 951: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
952: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
953: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
954: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 955: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 956: int selected(int kvar); /* Is covariate kvar selected for printing results */
957:
1.130 brouard 958: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 959: double **matprod2(); /* test */
1.126 brouard 960: double **oldm, **newm, **savm; /* Working pointers to matrices */
961: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 962: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
963:
1.136 brouard 964: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 965: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 966: FILE *ficlog, *ficrespow;
1.130 brouard 967: int globpr=0; /* Global variable for printing or not */
1.126 brouard 968: double fretone; /* Only one call to likelihood */
1.130 brouard 969: long ipmx=0; /* Number of contributions */
1.126 brouard 970: double sw; /* Sum of weights */
971: char filerespow[FILENAMELENGTH];
972: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
973: FILE *ficresilk;
974: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
975: FILE *ficresprobmorprev;
976: FILE *fichtm, *fichtmcov; /* Html File */
977: FILE *ficreseij;
978: char filerese[FILENAMELENGTH];
979: FILE *ficresstdeij;
980: char fileresstde[FILENAMELENGTH];
981: FILE *ficrescveij;
982: char filerescve[FILENAMELENGTH];
983: FILE *ficresvij;
984: char fileresv[FILENAMELENGTH];
985: FILE *ficresvpl;
986: char fileresvpl[FILENAMELENGTH];
987: char title[MAXLINE];
1.234 brouard 988: char model[MAXLINE]; /**< The model line */
1.217 brouard 989: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 990: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
991: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
992: char command[FILENAMELENGTH];
993: int outcmd=0;
994:
1.217 brouard 995: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 996: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 997: char filelog[FILENAMELENGTH]; /* Log file */
998: char filerest[FILENAMELENGTH];
999: char fileregp[FILENAMELENGTH];
1000: char popfile[FILENAMELENGTH];
1001:
1002: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1003:
1.157 brouard 1004: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1005: /* struct timezone tzp; */
1006: /* extern int gettimeofday(); */
1007: struct tm tml, *gmtime(), *localtime();
1008:
1009: extern time_t time();
1010:
1011: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1012: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1013: struct tm tm;
1014:
1.126 brouard 1015: char strcurr[80], strfor[80];
1016:
1017: char *endptr;
1018: long lval;
1019: double dval;
1020:
1021: #define NR_END 1
1022: #define FREE_ARG char*
1023: #define FTOL 1.0e-10
1024:
1025: #define NRANSI
1026: #define ITMAX 200
1027:
1028: #define TOL 2.0e-4
1029:
1030: #define CGOLD 0.3819660
1031: #define ZEPS 1.0e-10
1032: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1033:
1034: #define GOLD 1.618034
1035: #define GLIMIT 100.0
1036: #define TINY 1.0e-20
1037:
1038: static double maxarg1,maxarg2;
1039: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1040: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1041:
1042: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1043: #define rint(a) floor(a+0.5)
1.166 brouard 1044: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1045: #define mytinydouble 1.0e-16
1.166 brouard 1046: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1047: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1048: /* static double dsqrarg; */
1049: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1050: static double sqrarg;
1051: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1052: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1053: int agegomp= AGEGOMP;
1054:
1055: int imx;
1056: int stepm=1;
1057: /* Stepm, step in month: minimum step interpolation*/
1058:
1059: int estepm;
1060: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1061:
1062: int m,nb;
1063: long *num;
1.197 brouard 1064: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1065: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1066: covariate for which somebody answered excluding
1067: undefined. Usually 2: 0 and 1. */
1068: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1069: covariate for which somebody answered including
1070: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1071: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1072: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1073: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1074: double *ageexmed,*agecens;
1075: double dateintmean=0;
1076:
1077: double *weight;
1078: int **s; /* Status */
1.141 brouard 1079: double *agedc;
1.145 brouard 1080: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1081: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1082: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1083: double **coqvar; /* Fixed quantitative covariate iqv */
1084: double ***cotvar; /* Time varying covariate itv */
1085: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1086: double idx;
1087: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1088: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1089: /*k 1 2 3 4 5 6 7 8 9 */
1090: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1091: /* Tndvar[k] 1 2 3 4 5 */
1092: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1093: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1094: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1095: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1096: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1097: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1098: /* Tprod[i]=k 4 7 */
1099: /* Tage[i]=k 5 8 */
1100: /* */
1101: /* Type */
1102: /* V 1 2 3 4 5 */
1103: /* F F V V V */
1104: /* D Q D D Q */
1105: /* */
1106: int *TvarsD;
1107: int *TvarsDind;
1108: int *TvarsQ;
1109: int *TvarsQind;
1110:
1.235 brouard 1111: #define MAXRESULTLINES 10
1112: int nresult=0;
1113: int TKresult[MAXRESULTLINES];
1114: double Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1115: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1116: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1117: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1118:
1.234 brouard 1119: /* 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 1120: 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 */
1121: 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 */
1122: 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 */
1123: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1124: 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 */
1125: 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 1126: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1127: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1128: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1129: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1130: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1131: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1132: 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 */
1133: 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 */
1134:
1.230 brouard 1135: int *Tvarsel; /**< Selected covariates for output */
1136: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1137: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1138: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1139: 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 */
1.197 brouard 1140: int *Tage;
1.227 brouard 1141: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1142: 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 1143: 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*/
1144: 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 1145: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1146: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1147: int **Tvard;
1148: int *Tprod;/**< Gives the k position of the k1 product */
1149: int *Tposprod; /**< Gives the k1 product from the k position */
1150: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
1151: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1152: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1153: */
1154: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1155: double *lsurv, *lpop, *tpop;
1156:
1.231 brouard 1157: #define FD 1; /* Fixed dummy covariate */
1158: #define FQ 2; /* Fixed quantitative covariate */
1159: #define FP 3; /* Fixed product covariate */
1160: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1161: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1162: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1163: #define VD 10; /* Varying dummy covariate */
1164: #define VQ 11; /* Varying quantitative covariate */
1165: #define VP 12; /* Varying product covariate */
1166: #define VPDD 13; /* Varying product dummy*dummy covariate */
1167: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1168: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1169: #define APFD 16; /* Age product * fixed dummy covariate */
1170: #define APFQ 17; /* Age product * fixed quantitative covariate */
1171: #define APVD 18; /* Age product * varying dummy covariate */
1172: #define APVQ 19; /* Age product * varying quantitative covariate */
1173:
1174: #define FTYPE 1; /* Fixed covariate */
1175: #define VTYPE 2; /* Varying covariate (loop in wave) */
1176: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1177:
1178: struct kmodel{
1179: int maintype; /* main type */
1180: int subtype; /* subtype */
1181: };
1182: struct kmodel modell[NCOVMAX];
1183:
1.143 brouard 1184: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1185: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1186:
1187: /**************** split *************************/
1188: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1189: {
1190: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1191: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1192: */
1193: char *ss; /* pointer */
1.186 brouard 1194: int l1=0, l2=0; /* length counters */
1.126 brouard 1195:
1196: l1 = strlen(path ); /* length of path */
1197: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1198: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1199: if ( ss == NULL ) { /* no directory, so determine current directory */
1200: strcpy( name, path ); /* we got the fullname name because no directory */
1201: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1202: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1203: /* get current working directory */
1204: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1205: #ifdef WIN32
1206: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1207: #else
1208: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1209: #endif
1.126 brouard 1210: return( GLOCK_ERROR_GETCWD );
1211: }
1212: /* got dirc from getcwd*/
1213: printf(" DIRC = %s \n",dirc);
1.205 brouard 1214: } else { /* strip directory from path */
1.126 brouard 1215: ss++; /* after this, the filename */
1216: l2 = strlen( ss ); /* length of filename */
1217: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1218: strcpy( name, ss ); /* save file name */
1219: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1220: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1221: printf(" DIRC2 = %s \n",dirc);
1222: }
1223: /* We add a separator at the end of dirc if not exists */
1224: l1 = strlen( dirc ); /* length of directory */
1225: if( dirc[l1-1] != DIRSEPARATOR ){
1226: dirc[l1] = DIRSEPARATOR;
1227: dirc[l1+1] = 0;
1228: printf(" DIRC3 = %s \n",dirc);
1229: }
1230: ss = strrchr( name, '.' ); /* find last / */
1231: if (ss >0){
1232: ss++;
1233: strcpy(ext,ss); /* save extension */
1234: l1= strlen( name);
1235: l2= strlen(ss)+1;
1236: strncpy( finame, name, l1-l2);
1237: finame[l1-l2]= 0;
1238: }
1239:
1240: return( 0 ); /* we're done */
1241: }
1242:
1243:
1244: /******************************************/
1245:
1246: void replace_back_to_slash(char *s, char*t)
1247: {
1248: int i;
1249: int lg=0;
1250: i=0;
1251: lg=strlen(t);
1252: for(i=0; i<= lg; i++) {
1253: (s[i] = t[i]);
1254: if (t[i]== '\\') s[i]='/';
1255: }
1256: }
1257:
1.132 brouard 1258: char *trimbb(char *out, char *in)
1.137 brouard 1259: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1260: char *s;
1261: s=out;
1262: while (*in != '\0'){
1.137 brouard 1263: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1264: in++;
1265: }
1266: *out++ = *in++;
1267: }
1268: *out='\0';
1269: return s;
1270: }
1271:
1.187 brouard 1272: /* char *substrchaine(char *out, char *in, char *chain) */
1273: /* { */
1274: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1275: /* char *s, *t; */
1276: /* t=in;s=out; */
1277: /* while ((*in != *chain) && (*in != '\0')){ */
1278: /* *out++ = *in++; */
1279: /* } */
1280:
1281: /* /\* *in matches *chain *\/ */
1282: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1283: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1284: /* } */
1285: /* in--; chain--; */
1286: /* while ( (*in != '\0')){ */
1287: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1288: /* *out++ = *in++; */
1289: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1290: /* } */
1291: /* *out='\0'; */
1292: /* out=s; */
1293: /* return out; */
1294: /* } */
1295: char *substrchaine(char *out, char *in, char *chain)
1296: {
1297: /* Substract chain 'chain' from 'in', return and output 'out' */
1298: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1299:
1300: char *strloc;
1301:
1302: strcpy (out, in);
1303: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1304: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1305: if(strloc != NULL){
1306: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1307: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1308: /* strcpy (strloc, strloc +strlen(chain));*/
1309: }
1310: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1311: return out;
1312: }
1313:
1314:
1.145 brouard 1315: char *cutl(char *blocc, char *alocc, char *in, char occ)
1316: {
1.187 brouard 1317: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1318: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1319: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1320: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1321: */
1.160 brouard 1322: char *s, *t;
1.145 brouard 1323: t=in;s=in;
1324: while ((*in != occ) && (*in != '\0')){
1325: *alocc++ = *in++;
1326: }
1327: if( *in == occ){
1328: *(alocc)='\0';
1329: s=++in;
1330: }
1331:
1332: if (s == t) {/* occ not found */
1333: *(alocc-(in-s))='\0';
1334: in=s;
1335: }
1336: while ( *in != '\0'){
1337: *blocc++ = *in++;
1338: }
1339:
1340: *blocc='\0';
1341: return t;
1342: }
1.137 brouard 1343: char *cutv(char *blocc, char *alocc, char *in, char occ)
1344: {
1.187 brouard 1345: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1346: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1347: gives blocc="abcdef2ghi" and alocc="j".
1348: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1349: */
1350: char *s, *t;
1351: t=in;s=in;
1352: while (*in != '\0'){
1353: while( *in == occ){
1354: *blocc++ = *in++;
1355: s=in;
1356: }
1357: *blocc++ = *in++;
1358: }
1359: if (s == t) /* occ not found */
1360: *(blocc-(in-s))='\0';
1361: else
1362: *(blocc-(in-s)-1)='\0';
1363: in=s;
1364: while ( *in != '\0'){
1365: *alocc++ = *in++;
1366: }
1367:
1368: *alocc='\0';
1369: return s;
1370: }
1371:
1.126 brouard 1372: int nbocc(char *s, char occ)
1373: {
1374: int i,j=0;
1375: int lg=20;
1376: i=0;
1377: lg=strlen(s);
1378: for(i=0; i<= lg; i++) {
1.234 brouard 1379: if (s[i] == occ ) j++;
1.126 brouard 1380: }
1381: return j;
1382: }
1383:
1.137 brouard 1384: /* void cutv(char *u,char *v, char*t, char occ) */
1385: /* { */
1386: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1387: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1388: /* gives u="abcdef2ghi" and v="j" *\/ */
1389: /* int i,lg,j,p=0; */
1390: /* i=0; */
1391: /* lg=strlen(t); */
1392: /* for(j=0; j<=lg-1; j++) { */
1393: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1394: /* } */
1.126 brouard 1395:
1.137 brouard 1396: /* for(j=0; j<p; j++) { */
1397: /* (u[j] = t[j]); */
1398: /* } */
1399: /* u[p]='\0'; */
1.126 brouard 1400:
1.137 brouard 1401: /* for(j=0; j<= lg; j++) { */
1402: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1403: /* } */
1404: /* } */
1.126 brouard 1405:
1.160 brouard 1406: #ifdef _WIN32
1407: char * strsep(char **pp, const char *delim)
1408: {
1409: char *p, *q;
1410:
1411: if ((p = *pp) == NULL)
1412: return 0;
1413: if ((q = strpbrk (p, delim)) != NULL)
1414: {
1415: *pp = q + 1;
1416: *q = '\0';
1417: }
1418: else
1419: *pp = 0;
1420: return p;
1421: }
1422: #endif
1423:
1.126 brouard 1424: /********************** nrerror ********************/
1425:
1426: void nrerror(char error_text[])
1427: {
1428: fprintf(stderr,"ERREUR ...\n");
1429: fprintf(stderr,"%s\n",error_text);
1430: exit(EXIT_FAILURE);
1431: }
1432: /*********************** vector *******************/
1433: double *vector(int nl, int nh)
1434: {
1435: double *v;
1436: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1437: if (!v) nrerror("allocation failure in vector");
1438: return v-nl+NR_END;
1439: }
1440:
1441: /************************ free vector ******************/
1442: void free_vector(double*v, int nl, int nh)
1443: {
1444: free((FREE_ARG)(v+nl-NR_END));
1445: }
1446:
1447: /************************ivector *******************************/
1448: int *ivector(long nl,long nh)
1449: {
1450: int *v;
1451: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1452: if (!v) nrerror("allocation failure in ivector");
1453: return v-nl+NR_END;
1454: }
1455:
1456: /******************free ivector **************************/
1457: void free_ivector(int *v, long nl, long nh)
1458: {
1459: free((FREE_ARG)(v+nl-NR_END));
1460: }
1461:
1462: /************************lvector *******************************/
1463: long *lvector(long nl,long nh)
1464: {
1465: long *v;
1466: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1467: if (!v) nrerror("allocation failure in ivector");
1468: return v-nl+NR_END;
1469: }
1470:
1471: /******************free lvector **************************/
1472: void free_lvector(long *v, long nl, long nh)
1473: {
1474: free((FREE_ARG)(v+nl-NR_END));
1475: }
1476:
1477: /******************* imatrix *******************************/
1478: int **imatrix(long nrl, long nrh, long ncl, long nch)
1479: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1480: {
1481: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1482: int **m;
1483:
1484: /* allocate pointers to rows */
1485: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1486: if (!m) nrerror("allocation failure 1 in matrix()");
1487: m += NR_END;
1488: m -= nrl;
1489:
1490:
1491: /* allocate rows and set pointers to them */
1492: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1493: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1494: m[nrl] += NR_END;
1495: m[nrl] -= ncl;
1496:
1497: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1498:
1499: /* return pointer to array of pointers to rows */
1500: return m;
1501: }
1502:
1503: /****************** free_imatrix *************************/
1504: void free_imatrix(m,nrl,nrh,ncl,nch)
1505: int **m;
1506: long nch,ncl,nrh,nrl;
1507: /* free an int matrix allocated by imatrix() */
1508: {
1509: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1510: free((FREE_ARG) (m+nrl-NR_END));
1511: }
1512:
1513: /******************* matrix *******************************/
1514: double **matrix(long nrl, long nrh, long ncl, long nch)
1515: {
1516: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1517: double **m;
1518:
1519: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1520: if (!m) nrerror("allocation failure 1 in matrix()");
1521: m += NR_END;
1522: m -= nrl;
1523:
1524: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1525: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1526: m[nrl] += NR_END;
1527: m[nrl] -= ncl;
1528:
1529: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1530: return m;
1.145 brouard 1531: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1532: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1533: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1534: */
1535: }
1536:
1537: /*************************free matrix ************************/
1538: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1539: {
1540: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1541: free((FREE_ARG)(m+nrl-NR_END));
1542: }
1543:
1544: /******************* ma3x *******************************/
1545: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1546: {
1547: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1548: double ***m;
1549:
1550: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1551: if (!m) nrerror("allocation failure 1 in matrix()");
1552: m += NR_END;
1553: m -= nrl;
1554:
1555: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1556: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1557: m[nrl] += NR_END;
1558: m[nrl] -= ncl;
1559:
1560: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1561:
1562: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1563: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1564: m[nrl][ncl] += NR_END;
1565: m[nrl][ncl] -= nll;
1566: for (j=ncl+1; j<=nch; j++)
1567: m[nrl][j]=m[nrl][j-1]+nlay;
1568:
1569: for (i=nrl+1; i<=nrh; i++) {
1570: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1571: for (j=ncl+1; j<=nch; j++)
1572: m[i][j]=m[i][j-1]+nlay;
1573: }
1574: return m;
1575: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1576: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1577: */
1578: }
1579:
1580: /*************************free ma3x ************************/
1581: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1582: {
1583: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1584: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1585: free((FREE_ARG)(m+nrl-NR_END));
1586: }
1587:
1588: /*************** function subdirf ***********/
1589: char *subdirf(char fileres[])
1590: {
1591: /* Caution optionfilefiname is hidden */
1592: strcpy(tmpout,optionfilefiname);
1593: strcat(tmpout,"/"); /* Add to the right */
1594: strcat(tmpout,fileres);
1595: return tmpout;
1596: }
1597:
1598: /*************** function subdirf2 ***********/
1599: char *subdirf2(char fileres[], char *preop)
1600: {
1601:
1602: /* Caution optionfilefiname is hidden */
1603: strcpy(tmpout,optionfilefiname);
1604: strcat(tmpout,"/");
1605: strcat(tmpout,preop);
1606: strcat(tmpout,fileres);
1607: return tmpout;
1608: }
1609:
1610: /*************** function subdirf3 ***********/
1611: char *subdirf3(char fileres[], char *preop, char *preop2)
1612: {
1613:
1614: /* Caution optionfilefiname is hidden */
1615: strcpy(tmpout,optionfilefiname);
1616: strcat(tmpout,"/");
1617: strcat(tmpout,preop);
1618: strcat(tmpout,preop2);
1619: strcat(tmpout,fileres);
1620: return tmpout;
1621: }
1.213 brouard 1622:
1623: /*************** function subdirfext ***********/
1624: char *subdirfext(char fileres[], char *preop, char *postop)
1625: {
1626:
1627: strcpy(tmpout,preop);
1628: strcat(tmpout,fileres);
1629: strcat(tmpout,postop);
1630: return tmpout;
1631: }
1.126 brouard 1632:
1.213 brouard 1633: /*************** function subdirfext3 ***********/
1634: char *subdirfext3(char fileres[], char *preop, char *postop)
1635: {
1636:
1637: /* Caution optionfilefiname is hidden */
1638: strcpy(tmpout,optionfilefiname);
1639: strcat(tmpout,"/");
1640: strcat(tmpout,preop);
1641: strcat(tmpout,fileres);
1642: strcat(tmpout,postop);
1643: return tmpout;
1644: }
1645:
1.162 brouard 1646: char *asc_diff_time(long time_sec, char ascdiff[])
1647: {
1648: long sec_left, days, hours, minutes;
1649: days = (time_sec) / (60*60*24);
1650: sec_left = (time_sec) % (60*60*24);
1651: hours = (sec_left) / (60*60) ;
1652: sec_left = (sec_left) %(60*60);
1653: minutes = (sec_left) /60;
1654: sec_left = (sec_left) % (60);
1655: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1656: return ascdiff;
1657: }
1658:
1.126 brouard 1659: /***************** f1dim *************************/
1660: extern int ncom;
1661: extern double *pcom,*xicom;
1662: extern double (*nrfunc)(double []);
1663:
1664: double f1dim(double x)
1665: {
1666: int j;
1667: double f;
1668: double *xt;
1669:
1670: xt=vector(1,ncom);
1671: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1672: f=(*nrfunc)(xt);
1673: free_vector(xt,1,ncom);
1674: return f;
1675: }
1676:
1677: /*****************brent *************************/
1678: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1679: {
1680: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1681: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1682: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1683: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1684: * returned function value.
1685: */
1.126 brouard 1686: int iter;
1687: double a,b,d,etemp;
1.159 brouard 1688: double fu=0,fv,fw,fx;
1.164 brouard 1689: double ftemp=0.;
1.126 brouard 1690: double p,q,r,tol1,tol2,u,v,w,x,xm;
1691: double e=0.0;
1692:
1693: a=(ax < cx ? ax : cx);
1694: b=(ax > cx ? ax : cx);
1695: x=w=v=bx;
1696: fw=fv=fx=(*f)(x);
1697: for (iter=1;iter<=ITMAX;iter++) {
1698: xm=0.5*(a+b);
1699: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1700: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1701: printf(".");fflush(stdout);
1702: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1703: #ifdef DEBUGBRENT
1.126 brouard 1704: 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);
1705: 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);
1706: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1707: #endif
1708: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1709: *xmin=x;
1710: return fx;
1711: }
1712: ftemp=fu;
1713: if (fabs(e) > tol1) {
1714: r=(x-w)*(fx-fv);
1715: q=(x-v)*(fx-fw);
1716: p=(x-v)*q-(x-w)*r;
1717: q=2.0*(q-r);
1718: if (q > 0.0) p = -p;
1719: q=fabs(q);
1720: etemp=e;
1721: e=d;
1722: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1723: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1724: else {
1.224 brouard 1725: d=p/q;
1726: u=x+d;
1727: if (u-a < tol2 || b-u < tol2)
1728: d=SIGN(tol1,xm-x);
1.126 brouard 1729: }
1730: } else {
1731: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1732: }
1733: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1734: fu=(*f)(u);
1735: if (fu <= fx) {
1736: if (u >= x) a=x; else b=x;
1737: SHFT(v,w,x,u)
1.183 brouard 1738: SHFT(fv,fw,fx,fu)
1739: } else {
1740: if (u < x) a=u; else b=u;
1741: if (fu <= fw || w == x) {
1.224 brouard 1742: v=w;
1743: w=u;
1744: fv=fw;
1745: fw=fu;
1.183 brouard 1746: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1747: v=u;
1748: fv=fu;
1.183 brouard 1749: }
1750: }
1.126 brouard 1751: }
1752: nrerror("Too many iterations in brent");
1753: *xmin=x;
1754: return fx;
1755: }
1756:
1757: /****************** mnbrak ***********************/
1758:
1759: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1760: double (*func)(double))
1.183 brouard 1761: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1762: the downhill direction (defined by the function as evaluated at the initial points) and returns
1763: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1764: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1765: */
1.126 brouard 1766: double ulim,u,r,q, dum;
1767: double fu;
1.187 brouard 1768:
1769: double scale=10.;
1770: int iterscale=0;
1771:
1772: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1773: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1774:
1775:
1776: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1777: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1778: /* *bx = *ax - (*ax - *bx)/scale; */
1779: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1780: /* } */
1781:
1.126 brouard 1782: if (*fb > *fa) {
1783: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1784: SHFT(dum,*fb,*fa,dum)
1785: }
1.126 brouard 1786: *cx=(*bx)+GOLD*(*bx-*ax);
1787: *fc=(*func)(*cx);
1.183 brouard 1788: #ifdef DEBUG
1.224 brouard 1789: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1790: 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 1791: #endif
1.224 brouard 1792: 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 1793: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1794: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1795: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1796: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1797: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1798: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1799: fu=(*func)(u);
1.163 brouard 1800: #ifdef DEBUG
1801: /* f(x)=A(x-u)**2+f(u) */
1802: double A, fparabu;
1803: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1804: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1805: 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);
1806: 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 1807: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1808: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1809: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1810: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1811: #endif
1.184 brouard 1812: #ifdef MNBRAKORIGINAL
1.183 brouard 1813: #else
1.191 brouard 1814: /* if (fu > *fc) { */
1815: /* #ifdef DEBUG */
1816: /* printf("mnbrak4 fu > fc \n"); */
1817: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1818: /* #endif */
1819: /* /\* 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 *\\/ *\/ */
1820: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1821: /* dum=u; /\* Shifting c and u *\/ */
1822: /* u = *cx; */
1823: /* *cx = dum; */
1824: /* dum = fu; */
1825: /* fu = *fc; */
1826: /* *fc =dum; */
1827: /* } else { /\* end *\/ */
1828: /* #ifdef DEBUG */
1829: /* printf("mnbrak3 fu < fc \n"); */
1830: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1831: /* #endif */
1832: /* dum=u; /\* Shifting c and u *\/ */
1833: /* u = *cx; */
1834: /* *cx = dum; */
1835: /* dum = fu; */
1836: /* fu = *fc; */
1837: /* *fc =dum; */
1838: /* } */
1.224 brouard 1839: #ifdef DEBUGMNBRAK
1840: double A, fparabu;
1841: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1842: fparabu= *fa - A*(*ax-u)*(*ax-u);
1843: 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);
1844: 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 1845: #endif
1.191 brouard 1846: dum=u; /* Shifting c and u */
1847: u = *cx;
1848: *cx = dum;
1849: dum = fu;
1850: fu = *fc;
1851: *fc =dum;
1.183 brouard 1852: #endif
1.162 brouard 1853: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1854: #ifdef DEBUG
1.224 brouard 1855: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1856: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1857: #endif
1.126 brouard 1858: fu=(*func)(u);
1859: if (fu < *fc) {
1.183 brouard 1860: #ifdef DEBUG
1.224 brouard 1861: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1862: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1863: #endif
1864: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1865: SHFT(*fb,*fc,fu,(*func)(u))
1866: #ifdef DEBUG
1867: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1868: #endif
1869: }
1.162 brouard 1870: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1871: #ifdef DEBUG
1.224 brouard 1872: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1873: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1874: #endif
1.126 brouard 1875: u=ulim;
1876: fu=(*func)(u);
1.183 brouard 1877: } else { /* u could be left to b (if r > q parabola has a maximum) */
1878: #ifdef DEBUG
1.224 brouard 1879: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1880: 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 1881: #endif
1.126 brouard 1882: u=(*cx)+GOLD*(*cx-*bx);
1883: fu=(*func)(u);
1.224 brouard 1884: #ifdef DEBUG
1885: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1886: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1887: #endif
1.183 brouard 1888: } /* end tests */
1.126 brouard 1889: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1890: SHFT(*fa,*fb,*fc,fu)
1891: #ifdef DEBUG
1.224 brouard 1892: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1893: 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 1894: #endif
1895: } /* 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 1896: }
1897:
1898: /*************** linmin ************************/
1.162 brouard 1899: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1900: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1901: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1902: the value of func at the returned location p . This is actually all accomplished by calling the
1903: routines mnbrak and brent .*/
1.126 brouard 1904: int ncom;
1905: double *pcom,*xicom;
1906: double (*nrfunc)(double []);
1907:
1.224 brouard 1908: #ifdef LINMINORIGINAL
1.126 brouard 1909: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1910: #else
1911: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1912: #endif
1.126 brouard 1913: {
1914: double brent(double ax, double bx, double cx,
1915: double (*f)(double), double tol, double *xmin);
1916: double f1dim(double x);
1917: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
1918: double *fc, double (*func)(double));
1919: int j;
1920: double xx,xmin,bx,ax;
1921: double fx,fb,fa;
1.187 brouard 1922:
1.203 brouard 1923: #ifdef LINMINORIGINAL
1924: #else
1925: double scale=10., axs, xxs; /* Scale added for infinity */
1926: #endif
1927:
1.126 brouard 1928: ncom=n;
1929: pcom=vector(1,n);
1930: xicom=vector(1,n);
1931: nrfunc=func;
1932: for (j=1;j<=n;j++) {
1933: pcom[j]=p[j];
1.202 brouard 1934: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 1935: }
1.187 brouard 1936:
1.203 brouard 1937: #ifdef LINMINORIGINAL
1938: xx=1.;
1939: #else
1940: axs=0.0;
1941: xxs=1.;
1942: do{
1943: xx= xxs;
1944: #endif
1.187 brouard 1945: ax=0.;
1946: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
1947: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
1948: /* 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)) */
1949: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
1950: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
1951: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
1952: /* 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 1953: #ifdef LINMINORIGINAL
1954: #else
1955: if (fx != fx){
1.224 brouard 1956: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
1957: printf("|");
1958: fprintf(ficlog,"|");
1.203 brouard 1959: #ifdef DEBUGLINMIN
1.224 brouard 1960: 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 1961: #endif
1962: }
1.224 brouard 1963: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 1964: #endif
1965:
1.191 brouard 1966: #ifdef DEBUGLINMIN
1967: 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 1968: 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 1969: #endif
1.224 brouard 1970: #ifdef LINMINORIGINAL
1971: #else
1972: if(fb == fx){ /* Flat function in the direction */
1973: xmin=xx;
1974: *flat=1;
1975: }else{
1976: *flat=0;
1977: #endif
1978: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 1979: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
1980: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
1981: /* fmin = f(p[j] + xmin * xi[j]) */
1982: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
1983: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 1984: #ifdef DEBUG
1.224 brouard 1985: 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);
1986: 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);
1987: #endif
1988: #ifdef LINMINORIGINAL
1989: #else
1990: }
1.126 brouard 1991: #endif
1.191 brouard 1992: #ifdef DEBUGLINMIN
1993: printf("linmin end ");
1.202 brouard 1994: fprintf(ficlog,"linmin end ");
1.191 brouard 1995: #endif
1.126 brouard 1996: for (j=1;j<=n;j++) {
1.203 brouard 1997: #ifdef LINMINORIGINAL
1998: xi[j] *= xmin;
1999: #else
2000: #ifdef DEBUGLINMIN
2001: if(xxs <1.0)
2002: printf(" before xi[%d]=%12.8f", j,xi[j]);
2003: #endif
2004: 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) */
2005: #ifdef DEBUGLINMIN
2006: if(xxs <1.0)
2007: 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 );
2008: #endif
2009: #endif
1.187 brouard 2010: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2011: }
1.191 brouard 2012: #ifdef DEBUGLINMIN
1.203 brouard 2013: printf("\n");
1.191 brouard 2014: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2015: 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 2016: for (j=1;j<=n;j++) {
1.202 brouard 2017: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2018: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2019: if(j % ncovmodel == 0){
1.191 brouard 2020: printf("\n");
1.202 brouard 2021: fprintf(ficlog,"\n");
2022: }
1.191 brouard 2023: }
1.203 brouard 2024: #else
1.191 brouard 2025: #endif
1.126 brouard 2026: free_vector(xicom,1,n);
2027: free_vector(pcom,1,n);
2028: }
2029:
2030:
2031: /*************** powell ************************/
1.162 brouard 2032: /*
2033: Minimization of a function func of n variables. Input consists of an initial starting point
2034: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2035: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2036: such that failure to decrease by more than this amount on one iteration signals doneness. On
2037: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2038: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2039: */
1.224 brouard 2040: #ifdef LINMINORIGINAL
2041: #else
2042: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2043: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2044: #endif
1.126 brouard 2045: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2046: double (*func)(double []))
2047: {
1.224 brouard 2048: #ifdef LINMINORIGINAL
2049: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2050: double (*func)(double []));
1.224 brouard 2051: #else
2052: void linmin(double p[], double xi[], int n, double *fret,
2053: double (*func)(double []),int *flat);
2054: #endif
1.126 brouard 2055: int i,ibig,j;
2056: double del,t,*pt,*ptt,*xit;
1.181 brouard 2057: double directest;
1.126 brouard 2058: double fp,fptt;
2059: double *xits;
2060: int niterf, itmp;
1.224 brouard 2061: #ifdef LINMINORIGINAL
2062: #else
2063:
2064: flatdir=ivector(1,n);
2065: for (j=1;j<=n;j++) flatdir[j]=0;
2066: #endif
1.126 brouard 2067:
2068: pt=vector(1,n);
2069: ptt=vector(1,n);
2070: xit=vector(1,n);
2071: xits=vector(1,n);
2072: *fret=(*func)(p);
2073: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2074: rcurr_time = time(NULL);
1.126 brouard 2075: for (*iter=1;;++(*iter)) {
1.187 brouard 2076: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2077: ibig=0;
2078: del=0.0;
1.157 brouard 2079: rlast_time=rcurr_time;
2080: /* (void) gettimeofday(&curr_time,&tzp); */
2081: rcurr_time = time(NULL);
2082: curr_time = *localtime(&rcurr_time);
2083: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2084: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2085: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2086: for (i=1;i<=n;i++) {
1.126 brouard 2087: printf(" %d %.12f",i, p[i]);
2088: fprintf(ficlog," %d %.12lf",i, p[i]);
2089: fprintf(ficrespow," %.12lf", p[i]);
2090: }
2091: printf("\n");
2092: fprintf(ficlog,"\n");
2093: fprintf(ficrespow,"\n");fflush(ficrespow);
2094: if(*iter <=3){
1.157 brouard 2095: tml = *localtime(&rcurr_time);
2096: strcpy(strcurr,asctime(&tml));
2097: rforecast_time=rcurr_time;
1.126 brouard 2098: itmp = strlen(strcurr);
2099: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.224 brouard 2100: strcurr[itmp-1]='\0';
1.162 brouard 2101: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2102: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2103: for(niterf=10;niterf<=30;niterf+=10){
1.224 brouard 2104: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2105: forecast_time = *localtime(&rforecast_time);
2106: strcpy(strfor,asctime(&forecast_time));
2107: itmp = strlen(strfor);
2108: if(strfor[itmp-1]=='\n')
2109: strfor[itmp-1]='\0';
2110: 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);
2111: 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 2112: }
2113: }
1.187 brouard 2114: for (i=1;i<=n;i++) { /* For each direction i */
2115: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2116: fptt=(*fret);
2117: #ifdef DEBUG
1.203 brouard 2118: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2119: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2120: #endif
1.203 brouard 2121: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2122: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2123: #ifdef LINMINORIGINAL
1.188 brouard 2124: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2125: #else
2126: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2127: flatdir[i]=flat; /* Function is vanishing in that direction i */
2128: #endif
2129: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2130: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2131: /* because that direction will be replaced unless the gain del is small */
2132: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2133: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2134: /* with the new direction. */
2135: del=fabs(fptt-(*fret));
2136: ibig=i;
1.126 brouard 2137: }
2138: #ifdef DEBUG
2139: printf("%d %.12e",i,(*fret));
2140: fprintf(ficlog,"%d %.12e",i,(*fret));
2141: for (j=1;j<=n;j++) {
1.224 brouard 2142: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2143: printf(" x(%d)=%.12e",j,xit[j]);
2144: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2145: }
2146: for(j=1;j<=n;j++) {
1.225 brouard 2147: printf(" p(%d)=%.12e",j,p[j]);
2148: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2149: }
2150: printf("\n");
2151: fprintf(ficlog,"\n");
2152: #endif
1.187 brouard 2153: } /* end loop on each direction i */
2154: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2155: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2156: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2157: for(j=1;j<=n;j++) {
1.225 brouard 2158: if(flatdir[j] >0){
2159: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2160: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2161: }
2162: /* printf("\n"); */
2163: /* fprintf(ficlog,"\n"); */
2164: }
1.182 brouard 2165: if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /* Did we reach enough precision? */
1.188 brouard 2166: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2167: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2168: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2169: /* decreased of more than 3.84 */
2170: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2171: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2172: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2173:
1.188 brouard 2174: /* Starting the program with initial values given by a former maximization will simply change */
2175: /* the scales of the directions and the directions, because the are reset to canonical directions */
2176: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2177: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2178: #ifdef DEBUG
2179: int k[2],l;
2180: k[0]=1;
2181: k[1]=-1;
2182: printf("Max: %.12e",(*func)(p));
2183: fprintf(ficlog,"Max: %.12e",(*func)(p));
2184: for (j=1;j<=n;j++) {
2185: printf(" %.12e",p[j]);
2186: fprintf(ficlog," %.12e",p[j]);
2187: }
2188: printf("\n");
2189: fprintf(ficlog,"\n");
2190: for(l=0;l<=1;l++) {
2191: for (j=1;j<=n;j++) {
2192: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2193: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2194: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2195: }
2196: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2197: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2198: }
2199: #endif
2200:
1.224 brouard 2201: #ifdef LINMINORIGINAL
2202: #else
2203: free_ivector(flatdir,1,n);
2204: #endif
1.126 brouard 2205: free_vector(xit,1,n);
2206: free_vector(xits,1,n);
2207: free_vector(ptt,1,n);
2208: free_vector(pt,1,n);
2209: return;
1.192 brouard 2210: } /* enough precision */
1.126 brouard 2211: if (*iter == ITMAX) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2212: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2213: ptt[j]=2.0*p[j]-pt[j];
2214: xit[j]=p[j]-pt[j];
2215: pt[j]=p[j];
2216: }
1.181 brouard 2217: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2218: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2219: if (*iter <=4) {
1.225 brouard 2220: #else
2221: #endif
1.224 brouard 2222: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2223: #else
1.161 brouard 2224: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2225: #endif
1.162 brouard 2226: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2227: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2228: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2229: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2230: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2231: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2232: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2233: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2234: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2235: /* Even if f3 <f1, directest can be negative and t >0 */
2236: /* mu² and del² are equal when f3=f1 */
2237: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2238: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2239: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2240: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2241: #ifdef NRCORIGINAL
2242: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2243: #else
2244: 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 2245: t= t- del*SQR(fp-fptt);
1.183 brouard 2246: #endif
1.202 brouard 2247: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2248: #ifdef DEBUG
1.181 brouard 2249: 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);
2250: 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 2251: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2252: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2253: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2254: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2255: 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);
2256: 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);
2257: #endif
1.183 brouard 2258: #ifdef POWELLORIGINAL
2259: if (t < 0.0) { /* Then we use it for new direction */
2260: #else
1.182 brouard 2261: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2262: 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 2263: 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 2264: 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 2265: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2266: }
1.181 brouard 2267: if (directest < 0.0) { /* Then we use it for new direction */
2268: #endif
1.191 brouard 2269: #ifdef DEBUGLINMIN
1.234 brouard 2270: printf("Before linmin in direction P%d-P0\n",n);
2271: for (j=1;j<=n;j++) {
2272: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2273: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2274: if(j % ncovmodel == 0){
2275: printf("\n");
2276: fprintf(ficlog,"\n");
2277: }
2278: }
1.224 brouard 2279: #endif
2280: #ifdef LINMINORIGINAL
1.234 brouard 2281: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2282: #else
1.234 brouard 2283: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2284: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2285: #endif
1.234 brouard 2286:
1.191 brouard 2287: #ifdef DEBUGLINMIN
1.234 brouard 2288: for (j=1;j<=n;j++) {
2289: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2290: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2291: if(j % ncovmodel == 0){
2292: printf("\n");
2293: fprintf(ficlog,"\n");
2294: }
2295: }
1.224 brouard 2296: #endif
1.234 brouard 2297: for (j=1;j<=n;j++) {
2298: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2299: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2300: }
1.224 brouard 2301: #ifdef LINMINORIGINAL
2302: #else
1.234 brouard 2303: for (j=1, flatd=0;j<=n;j++) {
2304: if(flatdir[j]>0)
2305: flatd++;
2306: }
2307: if(flatd >0){
2308: printf("%d flat directions\n",flatd);
2309: fprintf(ficlog,"%d flat directions\n",flatd);
2310: for (j=1;j<=n;j++) {
2311: if(flatdir[j]>0){
2312: printf("%d ",j);
2313: fprintf(ficlog,"%d ",j);
2314: }
2315: }
2316: printf("\n");
2317: fprintf(ficlog,"\n");
2318: }
1.191 brouard 2319: #endif
1.234 brouard 2320: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2321: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2322:
1.126 brouard 2323: #ifdef DEBUG
1.234 brouard 2324: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2325: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2326: for(j=1;j<=n;j++){
2327: printf(" %lf",xit[j]);
2328: fprintf(ficlog," %lf",xit[j]);
2329: }
2330: printf("\n");
2331: fprintf(ficlog,"\n");
1.126 brouard 2332: #endif
1.192 brouard 2333: } /* end of t or directest negative */
1.224 brouard 2334: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2335: #else
1.234 brouard 2336: } /* end if (fptt < fp) */
1.192 brouard 2337: #endif
1.225 brouard 2338: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2339: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2340: #else
1.224 brouard 2341: #endif
1.234 brouard 2342: } /* loop iteration */
1.126 brouard 2343: }
1.234 brouard 2344:
1.126 brouard 2345: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2346:
1.235 brouard 2347: 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 2348: {
1.235 brouard 2349: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2350: (and selected quantitative values in nres)
2351: by left multiplying the unit
1.234 brouard 2352: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2353: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2354: /* Wx is row vector: population in state 1, population in state 2, population dead */
2355: /* or prevalence in state 1, prevalence in state 2, 0 */
2356: /* newm is the matrix after multiplications, its rows are identical at a factor */
2357: /* Initial matrix pimij */
2358: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2359: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2360: /* 0, 0 , 1} */
2361: /*
2362: * and after some iteration: */
2363: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2364: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2365: /* 0, 0 , 1} */
2366: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2367: /* {0.51571254859325999, 0.4842874514067399, */
2368: /* 0.51326036147820708, 0.48673963852179264} */
2369: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2370:
1.126 brouard 2371: int i, ii,j,k;
1.209 brouard 2372: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2373: /* double **matprod2(); */ /* test */
1.218 brouard 2374: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2375: double **newm;
1.209 brouard 2376: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2377: int ncvloop=0;
1.169 brouard 2378:
1.209 brouard 2379: min=vector(1,nlstate);
2380: max=vector(1,nlstate);
2381: meandiff=vector(1,nlstate);
2382:
1.218 brouard 2383: /* Starting with matrix unity */
1.126 brouard 2384: for (ii=1;ii<=nlstate+ndeath;ii++)
2385: for (j=1;j<=nlstate+ndeath;j++){
2386: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2387: }
1.169 brouard 2388:
2389: cov[1]=1.;
2390:
2391: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2392: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2393: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2394: ncvloop++;
1.126 brouard 2395: newm=savm;
2396: /* Covariates have to be included here again */
1.138 brouard 2397: cov[2]=agefin;
1.187 brouard 2398: if(nagesqr==1)
2399: cov[3]= agefin*agefin;;
1.234 brouard 2400: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2401: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2402: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2403: /* 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 2404: }
2405: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2406: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2407: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2408: /* 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 2409: }
1.234 brouard 2410: for (k=1; k<=cptcovage;k++){
2411: if(Dummy[Tvar[Tage[k]]]){
2412: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2413: } else{
1.235 brouard 2414: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2415: }
1.235 brouard 2416: /* 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 2417: }
2418: for (k=1; k<=cptcovprod;k++){ /* */
1.235 brouard 2419: /* 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]); */
1.200 brouard 2420: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
1.234 brouard 2421: }
1.138 brouard 2422: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2423: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2424: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2425: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2426: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2427: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2428: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2429:
1.126 brouard 2430: savm=oldm;
2431: oldm=newm;
1.209 brouard 2432:
2433: for(j=1; j<=nlstate; j++){
2434: max[j]=0.;
2435: min[j]=1.;
2436: }
2437: for(i=1;i<=nlstate;i++){
2438: sumnew=0;
2439: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2440: for(j=1; j<=nlstate; j++){
2441: prlim[i][j]= newm[i][j]/(1-sumnew);
2442: max[j]=FMAX(max[j],prlim[i][j]);
2443: min[j]=FMIN(min[j],prlim[i][j]);
2444: }
2445: }
2446:
1.126 brouard 2447: maxmax=0.;
1.209 brouard 2448: for(j=1; j<=nlstate; j++){
2449: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2450: maxmax=FMAX(maxmax,meandiff[j]);
2451: /* 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 2452: } /* j loop */
1.203 brouard 2453: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2454: /* 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 2455: if(maxmax < ftolpl){
1.209 brouard 2456: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2457: free_vector(min,1,nlstate);
2458: free_vector(max,1,nlstate);
2459: free_vector(meandiff,1,nlstate);
1.126 brouard 2460: return prlim;
2461: }
1.169 brouard 2462: } /* age loop */
1.208 brouard 2463: /* After some age loop it doesn't converge */
1.209 brouard 2464: 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 2465: 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 2466: /* 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); */
2467: free_vector(min,1,nlstate);
2468: free_vector(max,1,nlstate);
2469: free_vector(meandiff,1,nlstate);
1.208 brouard 2470:
1.169 brouard 2471: return prlim; /* should not reach here */
1.126 brouard 2472: }
2473:
1.217 brouard 2474:
2475: /**** Back Prevalence limit (stable or period prevalence) ****************/
2476:
1.218 brouard 2477: /* 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) */
2478: /* 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) */
2479: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij)
1.217 brouard 2480: {
1.218 brouard 2481: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2482: matrix by transitions matrix until convergence is reached with precision ftolpl */
2483: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2484: /* Wx is row vector: population in state 1, population in state 2, population dead */
2485: /* or prevalence in state 1, prevalence in state 2, 0 */
2486: /* newm is the matrix after multiplications, its rows are identical at a factor */
2487: /* Initial matrix pimij */
2488: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2489: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2490: /* 0, 0 , 1} */
2491: /*
2492: * and after some iteration: */
2493: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2494: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2495: /* 0, 0 , 1} */
2496: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2497: /* {0.51571254859325999, 0.4842874514067399, */
2498: /* 0.51326036147820708, 0.48673963852179264} */
2499: /* If we start from prlim again, prlim tends to a constant matrix */
2500:
2501: int i, ii,j,k;
2502: double *min, *max, *meandiff, maxmax,sumnew=0.;
2503: /* double **matprod2(); */ /* test */
2504: double **out, cov[NCOVMAX+1], **bmij();
2505: double **newm;
1.218 brouard 2506: double **dnewm, **doldm, **dsavm; /* for use */
2507: double **oldm, **savm; /* for use */
2508:
1.217 brouard 2509: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2510: int ncvloop=0;
2511:
2512: min=vector(1,nlstate);
2513: max=vector(1,nlstate);
2514: meandiff=vector(1,nlstate);
2515:
1.218 brouard 2516: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2517: oldm=oldms; savm=savms;
2518:
2519: /* Starting with matrix unity */
2520: for (ii=1;ii<=nlstate+ndeath;ii++)
2521: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2522: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2523: }
2524:
2525: cov[1]=1.;
2526:
2527: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2528: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2529: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2530: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2531: ncvloop++;
1.218 brouard 2532: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2533: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2534: /* Covariates have to be included here again */
2535: cov[2]=agefin;
2536: if(nagesqr==1)
2537: cov[3]= agefin*agefin;;
2538: for (k=1; k<=cptcovn;k++) {
2539: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
2540: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
2541: /* 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])]); */
2542: }
2543: for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2];
2544: for (k=1; k<=cptcovprod;k++) /* Useless */
2545: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
2546: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2547:
2548: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2549: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2550: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2551: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2552: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2553: /* ij should be linked to the correct index of cov */
2554: /* age and covariate values ij are in 'cov', but we need to pass
2555: * ij for the observed prevalence at age and status and covariate
2556: * number: prevacurrent[(int)agefin][ii][ij]
2557: */
2558: /* 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 *\/ */
2559: /* 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 *\/ */
2560: 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 2561: savm=oldm;
2562: oldm=newm;
2563: for(j=1; j<=nlstate; j++){
2564: max[j]=0.;
2565: min[j]=1.;
2566: }
2567: for(j=1; j<=nlstate; j++){
2568: for(i=1;i<=nlstate;i++){
1.234 brouard 2569: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2570: bprlim[i][j]= newm[i][j];
2571: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2572: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2573: }
2574: }
1.218 brouard 2575:
1.217 brouard 2576: maxmax=0.;
2577: for(i=1; i<=nlstate; i++){
2578: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2579: maxmax=FMAX(maxmax,meandiff[i]);
2580: /* 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); */
2581: } /* j loop */
2582: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2583: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2584: if(maxmax < ftolpl){
1.220 brouard 2585: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2586: free_vector(min,1,nlstate);
2587: free_vector(max,1,nlstate);
2588: free_vector(meandiff,1,nlstate);
2589: return bprlim;
2590: }
2591: } /* age loop */
2592: /* After some age loop it doesn't converge */
2593: 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'. \n\
2594: 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);
2595: /* 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); */
2596: free_vector(min,1,nlstate);
2597: free_vector(max,1,nlstate);
2598: free_vector(meandiff,1,nlstate);
2599:
2600: return bprlim; /* should not reach here */
2601: }
2602:
1.126 brouard 2603: /*************** transition probabilities ***************/
2604:
2605: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2606: {
1.138 brouard 2607: /* According to parameters values stored in x and the covariate's values stored in cov,
2608: computes the probability to be observed in state j being in state i by appying the
2609: model to the ncovmodel covariates (including constant and age).
2610: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2611: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2612: ncth covariate in the global vector x is given by the formula:
2613: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2614: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2615: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2616: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2617: Outputs ps[i][j] the probability to be observed in j being in j according to
2618: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2619: */
2620: double s1, lnpijopii;
1.126 brouard 2621: /*double t34;*/
1.164 brouard 2622: int i,j, nc, ii, jj;
1.126 brouard 2623:
1.223 brouard 2624: for(i=1; i<= nlstate; i++){
2625: for(j=1; j<i;j++){
2626: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2627: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2628: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2629: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2630: }
2631: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2632: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2633: }
2634: for(j=i+1; j<=nlstate+ndeath;j++){
2635: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2636: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2637: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2638: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2639: }
2640: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2641: }
2642: }
1.218 brouard 2643:
1.223 brouard 2644: for(i=1; i<= nlstate; i++){
2645: s1=0;
2646: for(j=1; j<i; j++){
2647: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2648: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2649: }
2650: for(j=i+1; j<=nlstate+ndeath; j++){
2651: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2652: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2653: }
2654: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2655: ps[i][i]=1./(s1+1.);
2656: /* Computing other pijs */
2657: for(j=1; j<i; j++)
2658: ps[i][j]= exp(ps[i][j])*ps[i][i];
2659: for(j=i+1; j<=nlstate+ndeath; j++)
2660: ps[i][j]= exp(ps[i][j])*ps[i][i];
2661: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2662: } /* end i */
1.218 brouard 2663:
1.223 brouard 2664: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2665: for(jj=1; jj<= nlstate+ndeath; jj++){
2666: ps[ii][jj]=0;
2667: ps[ii][ii]=1;
2668: }
2669: }
1.218 brouard 2670:
2671:
1.223 brouard 2672: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2673: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2674: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2675: /* } */
2676: /* printf("\n "); */
2677: /* } */
2678: /* printf("\n ");printf("%lf ",cov[2]);*/
2679: /*
2680: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2681: goto end;*/
1.223 brouard 2682: return ps;
1.126 brouard 2683: }
2684:
1.218 brouard 2685: /*************** backward transition probabilities ***************/
2686:
2687: /* 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 ) */
2688: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2689: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2690: {
1.222 brouard 2691: /* Computes the backward probability at age agefin and covariate ij
2692: * and returns in **ps as well as **bmij.
2693: */
1.218 brouard 2694: int i, ii, j,k;
1.222 brouard 2695:
2696: double **out, **pmij();
2697: double sumnew=0.;
1.218 brouard 2698: double agefin;
1.222 brouard 2699:
2700: double **dnewm, **dsavm, **doldm;
2701: double **bbmij;
2702:
1.218 brouard 2703: doldm=ddoldms; /* global pointers */
1.222 brouard 2704: dnewm=ddnewms;
2705: dsavm=ddsavms;
2706:
2707: agefin=cov[2];
2708: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2709: the observed prevalence (with this covariate ij) */
2710: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2711: /* We do have the matrix Px in savm and we need pij */
2712: for (j=1;j<=nlstate+ndeath;j++){
2713: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2714: for (ii=1;ii<=nlstate;ii++){
2715: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2716: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2717: for (ii=1;ii<=nlstate+ndeath;ii++){
2718: if(sumnew >= 1.e-10){
2719: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2720: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2721: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2722: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2723: /* }else */
2724: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2725: }else{
2726: 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);
2727: }
2728: } /*End ii */
2729: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2730: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2731: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2732: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2733: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2734: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2735: /* left Product of this matrix by diag matrix of prevalences (savm) */
2736: for (j=1;j<=nlstate+ndeath;j++){
2737: for (ii=1;ii<=nlstate+ndeath;ii++){
2738: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2739: }
2740: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2741: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2742: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2743: /* end bmij */
2744: return ps;
1.218 brouard 2745: }
1.217 brouard 2746: /*************** transition probabilities ***************/
2747:
1.218 brouard 2748: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2749: {
2750: /* According to parameters values stored in x and the covariate's values stored in cov,
2751: computes the probability to be observed in state j being in state i by appying the
2752: model to the ncovmodel covariates (including constant and age).
2753: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2754: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2755: ncth covariate in the global vector x is given by the formula:
2756: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2757: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2758: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2759: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2760: Outputs ps[i][j] the probability to be observed in j being in j according to
2761: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2762: */
2763: double s1, lnpijopii;
2764: /*double t34;*/
2765: int i,j, nc, ii, jj;
2766:
1.234 brouard 2767: for(i=1; i<= nlstate; i++){
2768: for(j=1; j<i;j++){
2769: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2770: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2771: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2772: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2773: }
2774: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2775: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2776: }
2777: for(j=i+1; j<=nlstate+ndeath;j++){
2778: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2779: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2780: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2781: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2782: }
2783: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2784: }
2785: }
2786:
2787: for(i=1; i<= nlstate; i++){
2788: s1=0;
2789: for(j=1; j<i; j++){
2790: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2791: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2792: }
2793: for(j=i+1; j<=nlstate+ndeath; j++){
2794: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2795: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2796: }
2797: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2798: ps[i][i]=1./(s1+1.);
2799: /* Computing other pijs */
2800: for(j=1; j<i; j++)
2801: ps[i][j]= exp(ps[i][j])*ps[i][i];
2802: for(j=i+1; j<=nlstate+ndeath; j++)
2803: ps[i][j]= exp(ps[i][j])*ps[i][i];
2804: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2805: } /* end i */
2806:
2807: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2808: for(jj=1; jj<= nlstate+ndeath; jj++){
2809: ps[ii][jj]=0;
2810: ps[ii][ii]=1;
2811: }
2812: }
2813: /* Added for backcast */ /* Transposed matrix too */
2814: for(jj=1; jj<= nlstate+ndeath; jj++){
2815: s1=0.;
2816: for(ii=1; ii<= nlstate+ndeath; ii++){
2817: s1+=ps[ii][jj];
2818: }
2819: for(ii=1; ii<= nlstate; ii++){
2820: ps[ii][jj]=ps[ii][jj]/s1;
2821: }
2822: }
2823: /* Transposition */
2824: for(jj=1; jj<= nlstate+ndeath; jj++){
2825: for(ii=jj; ii<= nlstate+ndeath; ii++){
2826: s1=ps[ii][jj];
2827: ps[ii][jj]=ps[jj][ii];
2828: ps[jj][ii]=s1;
2829: }
2830: }
2831: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2832: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2833: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2834: /* } */
2835: /* printf("\n "); */
2836: /* } */
2837: /* printf("\n ");printf("%lf ",cov[2]);*/
2838: /*
2839: for(i=1; i<= npar; i++) printf("%f ",x[i]);
2840: goto end;*/
2841: return ps;
1.217 brouard 2842: }
2843:
2844:
1.126 brouard 2845: /**************** Product of 2 matrices ******************/
2846:
1.145 brouard 2847: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 2848: {
2849: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
2850: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
2851: /* in, b, out are matrice of pointers which should have been initialized
2852: before: only the contents of out is modified. The function returns
2853: a pointer to pointers identical to out */
1.145 brouard 2854: int i, j, k;
1.126 brouard 2855: for(i=nrl; i<= nrh; i++)
1.145 brouard 2856: for(k=ncolol; k<=ncoloh; k++){
2857: out[i][k]=0.;
2858: for(j=ncl; j<=nch; j++)
2859: out[i][k] +=in[i][j]*b[j][k];
2860: }
1.126 brouard 2861: return out;
2862: }
2863:
2864:
2865: /************* Higher Matrix Product ***************/
2866:
1.235 brouard 2867: 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 2868: {
1.218 brouard 2869: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 2870: 'nhstepm*hstepm*stepm' months (i.e. until
2871: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
2872: nhstepm*hstepm matrices.
2873: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
2874: (typically every 2 years instead of every month which is too big
2875: for the memory).
2876: Model is determined by parameters x and covariates have to be
2877: included manually here.
2878:
2879: */
2880:
2881: int i, j, d, h, k;
1.131 brouard 2882: double **out, cov[NCOVMAX+1];
1.126 brouard 2883: double **newm;
1.187 brouard 2884: double agexact;
1.214 brouard 2885: double agebegin, ageend;
1.126 brouard 2886:
2887: /* Hstepm could be zero and should return the unit matrix */
2888: for (i=1;i<=nlstate+ndeath;i++)
2889: for (j=1;j<=nlstate+ndeath;j++){
2890: oldm[i][j]=(i==j ? 1.0 : 0.0);
2891: po[i][j][0]=(i==j ? 1.0 : 0.0);
2892: }
2893: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2894: for(h=1; h <=nhstepm; h++){
2895: for(d=1; d <=hstepm; d++){
2896: newm=savm;
2897: /* Covariates have to be included here again */
2898: cov[1]=1.;
1.214 brouard 2899: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 2900: cov[2]=agexact;
2901: if(nagesqr==1)
1.227 brouard 2902: cov[3]= agexact*agexact;
1.235 brouard 2903: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2904: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2905: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
2906: /* 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)); */
2907: }
2908: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2909: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2910: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2911: /* 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]); */
2912: }
2913: for (k=1; k<=cptcovage;k++){
2914: if(Dummy[Tvar[Tage[k]]]){
2915: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2916: } else{
2917: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2918: }
2919: /* 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]); */
2920: }
2921: for (k=1; k<=cptcovprod;k++){ /* */
2922: /* 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]); */
2923: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2924: }
2925: /* for (k=1; k<=cptcovn;k++) */
2926: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2927: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
2928: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
2929: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
2930: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 2931:
2932:
1.126 brouard 2933: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
2934: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 2935: /* right multiplication of oldm by the current matrix */
1.126 brouard 2936: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
2937: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 2938: /* if((int)age == 70){ */
2939: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
2940: /* for(i=1; i<=nlstate+ndeath; i++) { */
2941: /* printf("%d pmmij ",i); */
2942: /* for(j=1;j<=nlstate+ndeath;j++) { */
2943: /* printf("%f ",pmmij[i][j]); */
2944: /* } */
2945: /* printf(" oldm "); */
2946: /* for(j=1;j<=nlstate+ndeath;j++) { */
2947: /* printf("%f ",oldm[i][j]); */
2948: /* } */
2949: /* printf("\n"); */
2950: /* } */
2951: /* } */
1.126 brouard 2952: savm=oldm;
2953: oldm=newm;
2954: }
2955: for(i=1; i<=nlstate+ndeath; i++)
2956: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 2957: po[i][j][h]=newm[i][j];
2958: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 2959: }
1.128 brouard 2960: /*printf("h=%d ",h);*/
1.126 brouard 2961: } /* end h */
1.218 brouard 2962: /* printf("\n H=%d \n",h); */
1.126 brouard 2963: return po;
2964: }
2965:
1.217 brouard 2966: /************* Higher Back Matrix Product ***************/
1.218 brouard 2967: /* 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 2968: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 2969: {
1.218 brouard 2970: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 2971: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 2972: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
2973: nhstepm*hstepm matrices.
2974: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
2975: (typically every 2 years instead of every month which is too big
1.217 brouard 2976: for the memory).
1.218 brouard 2977: Model is determined by parameters x and covariates have to be
2978: included manually here.
1.217 brouard 2979:
1.222 brouard 2980: */
1.217 brouard 2981:
2982: int i, j, d, h, k;
2983: double **out, cov[NCOVMAX+1];
2984: double **newm;
2985: double agexact;
2986: double agebegin, ageend;
1.222 brouard 2987: double **oldm, **savm;
1.217 brouard 2988:
1.222 brouard 2989: oldm=oldms;savm=savms;
1.217 brouard 2990: /* Hstepm could be zero and should return the unit matrix */
2991: for (i=1;i<=nlstate+ndeath;i++)
2992: for (j=1;j<=nlstate+ndeath;j++){
2993: oldm[i][j]=(i==j ? 1.0 : 0.0);
2994: po[i][j][0]=(i==j ? 1.0 : 0.0);
2995: }
2996: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2997: for(h=1; h <=nhstepm; h++){
2998: for(d=1; d <=hstepm; d++){
2999: newm=savm;
3000: /* Covariates have to be included here again */
3001: cov[1]=1.;
3002: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3003: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3004: cov[2]=agexact;
3005: if(nagesqr==1)
1.222 brouard 3006: cov[3]= agexact*agexact;
1.218 brouard 3007: for (k=1; k<=cptcovn;k++)
1.222 brouard 3008: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3009: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3010: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3011: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3012: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3013: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3014: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3015: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3016: /* 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 3017:
3018:
1.217 brouard 3019: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3020: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3021: /* Careful transposed matrix */
1.222 brouard 3022: /* age is in cov[2] */
1.218 brouard 3023: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3024: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3025: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3026: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3027: /* if((int)age == 70){ */
3028: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3029: /* for(i=1; i<=nlstate+ndeath; i++) { */
3030: /* printf("%d pmmij ",i); */
3031: /* for(j=1;j<=nlstate+ndeath;j++) { */
3032: /* printf("%f ",pmmij[i][j]); */
3033: /* } */
3034: /* printf(" oldm "); */
3035: /* for(j=1;j<=nlstate+ndeath;j++) { */
3036: /* printf("%f ",oldm[i][j]); */
3037: /* } */
3038: /* printf("\n"); */
3039: /* } */
3040: /* } */
3041: savm=oldm;
3042: oldm=newm;
3043: }
3044: for(i=1; i<=nlstate+ndeath; i++)
3045: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3046: po[i][j][h]=newm[i][j];
3047: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3048: }
3049: /*printf("h=%d ",h);*/
3050: } /* end h */
1.222 brouard 3051: /* printf("\n H=%d \n",h); */
1.217 brouard 3052: return po;
3053: }
3054:
3055:
1.162 brouard 3056: #ifdef NLOPT
3057: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3058: double fret;
3059: double *xt;
3060: int j;
3061: myfunc_data *d2 = (myfunc_data *) pd;
3062: /* xt = (p1-1); */
3063: xt=vector(1,n);
3064: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3065:
3066: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3067: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3068: printf("Function = %.12lf ",fret);
3069: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3070: printf("\n");
3071: free_vector(xt,1,n);
3072: return fret;
3073: }
3074: #endif
1.126 brouard 3075:
3076: /*************** log-likelihood *************/
3077: double func( double *x)
3078: {
1.226 brouard 3079: int i, ii, j, k, mi, d, kk;
3080: int ioffset=0;
3081: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3082: double **out;
3083: double lli; /* Individual log likelihood */
3084: int s1, s2;
1.228 brouard 3085: 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 3086: double bbh, survp;
3087: long ipmx;
3088: double agexact;
3089: /*extern weight */
3090: /* We are differentiating ll according to initial status */
3091: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3092: /*for(i=1;i<imx;i++)
3093: printf(" %d\n",s[4][i]);
3094: */
1.162 brouard 3095:
1.226 brouard 3096: ++countcallfunc;
1.162 brouard 3097:
1.226 brouard 3098: cov[1]=1.;
1.126 brouard 3099:
1.226 brouard 3100: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3101: ioffset=0;
1.226 brouard 3102: if(mle==1){
3103: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3104: /* Computes the values of the ncovmodel covariates of the model
3105: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3106: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3107: to be observed in j being in i according to the model.
3108: */
3109: ioffset=2+nagesqr+cptcovage;
1.233 brouard 3110: /* Fixed */
1.234 brouard 3111: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3112: 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)*/
3113: }
1.226 brouard 3114: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3115: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3116: has been calculated etc */
3117: /* For an individual i, wav[i] gives the number of effective waves */
3118: /* We compute the contribution to Likelihood of each effective transition
3119: mw[mi][i] is real wave of the mi th effectve wave */
3120: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3121: s2=s[mw[mi+1][i]][i];
3122: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3123: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3124: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3125: */
3126: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3127: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
3128: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i];
3129: }
3130: for (ii=1;ii<=nlstate+ndeath;ii++)
3131: for (j=1;j<=nlstate+ndeath;j++){
3132: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3133: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3134: }
3135: for(d=0; d<dh[mi][i]; d++){
3136: newm=savm;
3137: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3138: cov[2]=agexact;
3139: if(nagesqr==1)
3140: cov[3]= agexact*agexact; /* Should be changed here */
3141: for (kk=1; kk<=cptcovage;kk++) {
3142: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
3143: }
3144: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3145: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3146: savm=oldm;
3147: oldm=newm;
3148: } /* end mult */
3149:
3150: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3151: /* But now since version 0.9 we anticipate for bias at large stepm.
3152: * If stepm is larger than one month (smallest stepm) and if the exact delay
3153: * (in months) between two waves is not a multiple of stepm, we rounded to
3154: * the nearest (and in case of equal distance, to the lowest) interval but now
3155: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3156: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3157: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3158: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3159: * -stepm/2 to stepm/2 .
3160: * For stepm=1 the results are the same as for previous versions of Imach.
3161: * For stepm > 1 the results are less biased than in previous versions.
3162: */
1.234 brouard 3163: s1=s[mw[mi][i]][i];
3164: s2=s[mw[mi+1][i]][i];
3165: bbh=(double)bh[mi][i]/(double)stepm;
3166: /* bias bh is positive if real duration
3167: * is higher than the multiple of stepm and negative otherwise.
3168: */
3169: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3170: if( s2 > nlstate){
3171: /* i.e. if s2 is a death state and if the date of death is known
3172: then the contribution to the likelihood is the probability to
3173: die between last step unit time and current step unit time,
3174: which is also equal to probability to die before dh
3175: minus probability to die before dh-stepm .
3176: In version up to 0.92 likelihood was computed
3177: as if date of death was unknown. Death was treated as any other
3178: health state: the date of the interview describes the actual state
3179: and not the date of a change in health state. The former idea was
3180: to consider that at each interview the state was recorded
3181: (healthy, disable or death) and IMaCh was corrected; but when we
3182: introduced the exact date of death then we should have modified
3183: the contribution of an exact death to the likelihood. This new
3184: contribution is smaller and very dependent of the step unit
3185: stepm. It is no more the probability to die between last interview
3186: and month of death but the probability to survive from last
3187: interview up to one month before death multiplied by the
3188: probability to die within a month. Thanks to Chris
3189: Jackson for correcting this bug. Former versions increased
3190: mortality artificially. The bad side is that we add another loop
3191: which slows down the processing. The difference can be up to 10%
3192: lower mortality.
3193: */
3194: /* If, at the beginning of the maximization mostly, the
3195: cumulative probability or probability to be dead is
3196: constant (ie = 1) over time d, the difference is equal to
3197: 0. out[s1][3] = savm[s1][3]: probability, being at state
3198: s1 at precedent wave, to be dead a month before current
3199: wave is equal to probability, being at state s1 at
3200: precedent wave, to be dead at mont of the current
3201: wave. Then the observed probability (that this person died)
3202: is null according to current estimated parameter. In fact,
3203: it should be very low but not zero otherwise the log go to
3204: infinity.
3205: */
1.183 brouard 3206: /* #ifdef INFINITYORIGINAL */
3207: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3208: /* #else */
3209: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3210: /* lli=log(mytinydouble); */
3211: /* else */
3212: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3213: /* #endif */
1.226 brouard 3214: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3215:
1.226 brouard 3216: } else if ( s2==-1 ) { /* alive */
3217: for (j=1,survp=0. ; j<=nlstate; j++)
3218: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3219: /*survp += out[s1][j]; */
3220: lli= log(survp);
3221: }
3222: else if (s2==-4) {
3223: for (j=3,survp=0. ; j<=nlstate; j++)
3224: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3225: lli= log(survp);
3226: }
3227: else if (s2==-5) {
3228: for (j=1,survp=0. ; j<=2; j++)
3229: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3230: lli= log(survp);
3231: }
3232: else{
3233: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3234: /* 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 */
3235: }
3236: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3237: /*if(lli ==000.0)*/
3238: /*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); */
3239: ipmx +=1;
3240: sw += weight[i];
3241: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3242: /* if (lli < log(mytinydouble)){ */
3243: /* 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); */
3244: /* 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]); */
3245: /* } */
3246: } /* end of wave */
3247: } /* end of individual */
3248: } else if(mle==2){
3249: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3250: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3251: for(mi=1; mi<= wav[i]-1; mi++){
3252: for (ii=1;ii<=nlstate+ndeath;ii++)
3253: for (j=1;j<=nlstate+ndeath;j++){
3254: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3255: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3256: }
3257: for(d=0; d<=dh[mi][i]; d++){
3258: newm=savm;
3259: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3260: cov[2]=agexact;
3261: if(nagesqr==1)
3262: cov[3]= agexact*agexact;
3263: for (kk=1; kk<=cptcovage;kk++) {
3264: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3265: }
3266: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3267: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3268: savm=oldm;
3269: oldm=newm;
3270: } /* end mult */
3271:
3272: s1=s[mw[mi][i]][i];
3273: s2=s[mw[mi+1][i]][i];
3274: bbh=(double)bh[mi][i]/(double)stepm;
3275: 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 */
3276: ipmx +=1;
3277: sw += weight[i];
3278: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3279: } /* end of wave */
3280: } /* end of individual */
3281: } else if(mle==3){ /* exponential inter-extrapolation */
3282: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3283: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3284: for(mi=1; mi<= wav[i]-1; mi++){
3285: for (ii=1;ii<=nlstate+ndeath;ii++)
3286: for (j=1;j<=nlstate+ndeath;j++){
3287: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3288: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3289: }
3290: for(d=0; d<dh[mi][i]; d++){
3291: newm=savm;
3292: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3293: cov[2]=agexact;
3294: if(nagesqr==1)
3295: cov[3]= agexact*agexact;
3296: for (kk=1; kk<=cptcovage;kk++) {
3297: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3298: }
3299: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3300: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3301: savm=oldm;
3302: oldm=newm;
3303: } /* end mult */
3304:
3305: s1=s[mw[mi][i]][i];
3306: s2=s[mw[mi+1][i]][i];
3307: bbh=(double)bh[mi][i]/(double)stepm;
3308: 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 */
3309: ipmx +=1;
3310: sw += weight[i];
3311: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3312: } /* end of wave */
3313: } /* end of individual */
3314: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3315: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3316: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3317: for(mi=1; mi<= wav[i]-1; mi++){
3318: for (ii=1;ii<=nlstate+ndeath;ii++)
3319: for (j=1;j<=nlstate+ndeath;j++){
3320: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3321: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3322: }
3323: for(d=0; d<dh[mi][i]; d++){
3324: newm=savm;
3325: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3326: cov[2]=agexact;
3327: if(nagesqr==1)
3328: cov[3]= agexact*agexact;
3329: for (kk=1; kk<=cptcovage;kk++) {
3330: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3331: }
1.126 brouard 3332:
1.226 brouard 3333: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3334: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3335: savm=oldm;
3336: oldm=newm;
3337: } /* end mult */
3338:
3339: s1=s[mw[mi][i]][i];
3340: s2=s[mw[mi+1][i]][i];
3341: if( s2 > nlstate){
3342: lli=log(out[s1][s2] - savm[s1][s2]);
3343: } else if ( s2==-1 ) { /* alive */
3344: for (j=1,survp=0. ; j<=nlstate; j++)
3345: survp += out[s1][j];
3346: lli= log(survp);
3347: }else{
3348: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3349: }
3350: ipmx +=1;
3351: sw += weight[i];
3352: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3353: /* 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 3354: } /* end of wave */
3355: } /* end of individual */
3356: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3357: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3358: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3359: for(mi=1; mi<= wav[i]-1; mi++){
3360: for (ii=1;ii<=nlstate+ndeath;ii++)
3361: for (j=1;j<=nlstate+ndeath;j++){
3362: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3363: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3364: }
3365: for(d=0; d<dh[mi][i]; d++){
3366: newm=savm;
3367: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3368: cov[2]=agexact;
3369: if(nagesqr==1)
3370: cov[3]= agexact*agexact;
3371: for (kk=1; kk<=cptcovage;kk++) {
3372: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3373: }
1.126 brouard 3374:
1.226 brouard 3375: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3376: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3377: savm=oldm;
3378: oldm=newm;
3379: } /* end mult */
3380:
3381: s1=s[mw[mi][i]][i];
3382: s2=s[mw[mi+1][i]][i];
3383: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3384: ipmx +=1;
3385: sw += weight[i];
3386: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3387: /*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]);*/
3388: } /* end of wave */
3389: } /* end of individual */
3390: } /* End of if */
3391: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3392: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3393: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3394: return -l;
1.126 brouard 3395: }
3396:
3397: /*************** log-likelihood *************/
3398: double funcone( double *x)
3399: {
1.228 brouard 3400: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3401: int i, ii, j, k, mi, d, kk;
1.228 brouard 3402: int ioffset=0;
1.131 brouard 3403: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3404: double **out;
3405: double lli; /* Individual log likelihood */
3406: double llt;
3407: int s1, s2;
1.228 brouard 3408: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3409:
1.126 brouard 3410: double bbh, survp;
1.187 brouard 3411: double agexact;
1.214 brouard 3412: double agebegin, ageend;
1.126 brouard 3413: /*extern weight */
3414: /* We are differentiating ll according to initial status */
3415: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3416: /*for(i=1;i<imx;i++)
3417: printf(" %d\n",s[4][i]);
3418: */
3419: cov[1]=1.;
3420:
3421: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3422: ioffset=0;
3423: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.225 brouard 3424: ioffset=2+nagesqr+cptcovage;
1.232 brouard 3425: /* Fixed */
1.224 brouard 3426: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3427: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3428: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3429: 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)*/
3430: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3431: /* cov[2+6]=covar[Tvar[6]][i]; */
3432: /* cov[2+6]=covar[2][i]; V2 */
3433: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3434: /* cov[2+7]=covar[Tvar[7]][i]; */
3435: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3436: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3437: /* cov[2+9]=covar[Tvar[9]][i]; */
3438: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3439: }
1.232 brouard 3440: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3441: /* 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?)*\/ */
3442: /* } */
1.231 brouard 3443: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3444: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3445: /* } */
1.225 brouard 3446:
1.233 brouard 3447:
3448: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3449: /* Wave varying (but not age varying) */
3450: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.233 brouard 3451: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i];
1.232 brouard 3452: }
3453: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.231 brouard 3454: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3455: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
1.232 brouard 3456: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3457: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
1.231 brouard 3458: /* 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 3459: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
3460: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3461: /* /\* 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]); *\/ */
3462: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
3463: /* } */
1.126 brouard 3464: for (ii=1;ii<=nlstate+ndeath;ii++)
1.231 brouard 3465: for (j=1;j<=nlstate+ndeath;j++){
3466: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3467: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3468: }
1.214 brouard 3469:
3470: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3471: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3472: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.231 brouard 3473: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3474: and mw[mi+1][i]. dh depends on stepm.*/
3475: newm=savm;
3476: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3477: cov[2]=agexact;
3478: if(nagesqr==1)
3479: cov[3]= agexact*agexact;
3480: for (kk=1; kk<=cptcovage;kk++) {
3481: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3482: }
3483: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3484: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3485: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3486: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3487: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3488: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3489: savm=oldm;
3490: oldm=newm;
1.126 brouard 3491: } /* end mult */
3492:
3493: s1=s[mw[mi][i]][i];
3494: s2=s[mw[mi+1][i]][i];
1.217 brouard 3495: /* if(s2==-1){ */
3496: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3497: /* /\* exit(1); *\/ */
3498: /* } */
1.126 brouard 3499: bbh=(double)bh[mi][i]/(double)stepm;
3500: /* bias is positive if real duration
3501: * is higher than the multiple of stepm and negative otherwise.
3502: */
3503: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.232 brouard 3504: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3505: } else if ( s2==-1 ) { /* alive */
1.232 brouard 3506: for (j=1,survp=0. ; j<=nlstate; j++)
3507: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3508: lli= log(survp);
1.126 brouard 3509: }else if (mle==1){
1.232 brouard 3510: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3511: } else if(mle==2){
1.232 brouard 3512: 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 3513: } else if(mle==3){ /* exponential inter-extrapolation */
1.232 brouard 3514: 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 3515: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.232 brouard 3516: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3517: } else{ /* mle=0 back to 1 */
1.232 brouard 3518: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3519: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3520: } /* End of if */
3521: ipmx +=1;
3522: sw += weight[i];
3523: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3524: /*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 3525: if(globpr){
1.232 brouard 3526: fprintf(ficresilk,"%9ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3527: %11.6f %11.6f %11.6f ", \
1.232 brouard 3528: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3529: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3530: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3531: llt +=ll[k]*gipmx/gsw;
3532: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3533: }
3534: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3535: }
1.232 brouard 3536: } /* end of wave */
3537: } /* end of individual */
3538: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3539: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3540: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3541: if(globpr==0){ /* First time we count the contributions and weights */
3542: gipmx=ipmx;
3543: gsw=sw;
3544: }
3545: return -l;
1.126 brouard 3546: }
3547:
3548:
3549: /*************** function likelione ***********/
3550: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3551: {
3552: /* This routine should help understanding what is done with
3553: the selection of individuals/waves and
3554: to check the exact contribution to the likelihood.
3555: Plotting could be done.
3556: */
3557: int k;
3558:
3559: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3560: strcpy(fileresilk,"ILK_");
1.202 brouard 3561: strcat(fileresilk,fileresu);
1.126 brouard 3562: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3563: printf("Problem with resultfile: %s\n", fileresilk);
3564: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3565: }
1.214 brouard 3566: 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");
3567: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3568: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3569: for(k=1; k<=nlstate; k++)
3570: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3571: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3572: }
3573:
3574: *fretone=(*funcone)(p);
3575: if(*globpri !=0){
3576: fclose(ficresilk);
1.205 brouard 3577: if (mle ==0)
3578: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3579: else if(mle >=1)
3580: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3581: 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 3582:
1.208 brouard 3583:
3584: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3585: 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 3586: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3587: }
1.207 brouard 3588: 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 3589: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3590: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3591: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3592: fflush(fichtm);
1.205 brouard 3593: }
1.126 brouard 3594: return;
3595: }
3596:
3597:
3598: /*********** Maximum Likelihood Estimation ***************/
3599:
3600: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3601: {
1.165 brouard 3602: int i,j, iter=0;
1.126 brouard 3603: double **xi;
3604: double fret;
3605: double fretone; /* Only one call to likelihood */
3606: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3607:
3608: #ifdef NLOPT
3609: int creturn;
3610: nlopt_opt opt;
3611: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3612: double *lb;
3613: double minf; /* the minimum objective value, upon return */
3614: double * p1; /* Shifted parameters from 0 instead of 1 */
3615: myfunc_data dinst, *d = &dinst;
3616: #endif
3617:
3618:
1.126 brouard 3619: xi=matrix(1,npar,1,npar);
3620: for (i=1;i<=npar;i++)
3621: for (j=1;j<=npar;j++)
3622: xi[i][j]=(i==j ? 1.0 : 0.0);
3623: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3624: strcpy(filerespow,"POW_");
1.126 brouard 3625: strcat(filerespow,fileres);
3626: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3627: printf("Problem with resultfile: %s\n", filerespow);
3628: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3629: }
3630: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3631: for (i=1;i<=nlstate;i++)
3632: for(j=1;j<=nlstate+ndeath;j++)
3633: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3634: fprintf(ficrespow,"\n");
1.162 brouard 3635: #ifdef POWELL
1.126 brouard 3636: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3637: #endif
1.126 brouard 3638:
1.162 brouard 3639: #ifdef NLOPT
3640: #ifdef NEWUOA
3641: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3642: #else
3643: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3644: #endif
3645: lb=vector(0,npar-1);
3646: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3647: nlopt_set_lower_bounds(opt, lb);
3648: nlopt_set_initial_step1(opt, 0.1);
3649:
3650: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3651: d->function = func;
3652: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3653: nlopt_set_min_objective(opt, myfunc, d);
3654: nlopt_set_xtol_rel(opt, ftol);
3655: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3656: printf("nlopt failed! %d\n",creturn);
3657: }
3658: else {
3659: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3660: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3661: iter=1; /* not equal */
3662: }
3663: nlopt_destroy(opt);
3664: #endif
1.126 brouard 3665: free_matrix(xi,1,npar,1,npar);
3666: fclose(ficrespow);
1.203 brouard 3667: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3668: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3669: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3670:
3671: }
3672:
3673: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3674: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3675: {
3676: double **a,**y,*x,pd;
1.203 brouard 3677: /* double **hess; */
1.164 brouard 3678: int i, j;
1.126 brouard 3679: int *indx;
3680:
3681: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3682: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3683: void lubksb(double **a, int npar, int *indx, double b[]) ;
3684: void ludcmp(double **a, int npar, int *indx, double *d) ;
3685: double gompertz(double p[]);
1.203 brouard 3686: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3687:
3688: printf("\nCalculation of the hessian matrix. Wait...\n");
3689: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3690: for (i=1;i<=npar;i++){
1.203 brouard 3691: printf("%d-",i);fflush(stdout);
3692: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3693:
3694: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3695:
3696: /* printf(" %f ",p[i]);
3697: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3698: }
3699:
3700: for (i=1;i<=npar;i++) {
3701: for (j=1;j<=npar;j++) {
3702: if (j>i) {
1.203 brouard 3703: printf(".%d-%d",i,j);fflush(stdout);
3704: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3705: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3706:
3707: hess[j][i]=hess[i][j];
3708: /*printf(" %lf ",hess[i][j]);*/
3709: }
3710: }
3711: }
3712: printf("\n");
3713: fprintf(ficlog,"\n");
3714:
3715: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3716: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3717:
3718: a=matrix(1,npar,1,npar);
3719: y=matrix(1,npar,1,npar);
3720: x=vector(1,npar);
3721: indx=ivector(1,npar);
3722: for (i=1;i<=npar;i++)
3723: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3724: ludcmp(a,npar,indx,&pd);
3725:
3726: for (j=1;j<=npar;j++) {
3727: for (i=1;i<=npar;i++) x[i]=0;
3728: x[j]=1;
3729: lubksb(a,npar,indx,x);
3730: for (i=1;i<=npar;i++){
3731: matcov[i][j]=x[i];
3732: }
3733: }
3734:
3735: printf("\n#Hessian matrix#\n");
3736: fprintf(ficlog,"\n#Hessian matrix#\n");
3737: for (i=1;i<=npar;i++) {
3738: for (j=1;j<=npar;j++) {
1.203 brouard 3739: printf("%.6e ",hess[i][j]);
3740: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3741: }
3742: printf("\n");
3743: fprintf(ficlog,"\n");
3744: }
3745:
1.203 brouard 3746: /* printf("\n#Covariance matrix#\n"); */
3747: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3748: /* for (i=1;i<=npar;i++) { */
3749: /* for (j=1;j<=npar;j++) { */
3750: /* printf("%.6e ",matcov[i][j]); */
3751: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3752: /* } */
3753: /* printf("\n"); */
3754: /* fprintf(ficlog,"\n"); */
3755: /* } */
3756:
1.126 brouard 3757: /* Recompute Inverse */
1.203 brouard 3758: /* for (i=1;i<=npar;i++) */
3759: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3760: /* ludcmp(a,npar,indx,&pd); */
3761:
3762: /* printf("\n#Hessian matrix recomputed#\n"); */
3763:
3764: /* for (j=1;j<=npar;j++) { */
3765: /* for (i=1;i<=npar;i++) x[i]=0; */
3766: /* x[j]=1; */
3767: /* lubksb(a,npar,indx,x); */
3768: /* for (i=1;i<=npar;i++){ */
3769: /* y[i][j]=x[i]; */
3770: /* printf("%.3e ",y[i][j]); */
3771: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3772: /* } */
3773: /* printf("\n"); */
3774: /* fprintf(ficlog,"\n"); */
3775: /* } */
3776:
3777: /* Verifying the inverse matrix */
3778: #ifdef DEBUGHESS
3779: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3780:
1.203 brouard 3781: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3782: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3783:
3784: for (j=1;j<=npar;j++) {
3785: for (i=1;i<=npar;i++){
1.203 brouard 3786: printf("%.2f ",y[i][j]);
3787: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3788: }
3789: printf("\n");
3790: fprintf(ficlog,"\n");
3791: }
1.203 brouard 3792: #endif
1.126 brouard 3793:
3794: free_matrix(a,1,npar,1,npar);
3795: free_matrix(y,1,npar,1,npar);
3796: free_vector(x,1,npar);
3797: free_ivector(indx,1,npar);
1.203 brouard 3798: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3799:
3800:
3801: }
3802:
3803: /*************** hessian matrix ****************/
3804: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3805: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3806: int i;
3807: int l=1, lmax=20;
1.203 brouard 3808: double k1,k2, res, fx;
1.132 brouard 3809: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3810: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3811: int k=0,kmax=10;
3812: double l1;
3813:
3814: fx=func(x);
3815: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3816: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3817: l1=pow(10,l);
3818: delts=delt;
3819: for(k=1 ; k <kmax; k=k+1){
3820: delt = delta*(l1*k);
3821: p2[theta]=x[theta] +delt;
1.145 brouard 3822: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 3823: p2[theta]=x[theta]-delt;
3824: k2=func(p2)-fx;
3825: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 3826: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 3827:
1.203 brouard 3828: #ifdef DEBUGHESSII
1.126 brouard 3829: 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);
3830: 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);
3831: #endif
3832: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
3833: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
3834: k=kmax;
3835: }
3836: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 3837: k=kmax; l=lmax*10;
1.126 brouard 3838: }
3839: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
3840: delts=delt;
3841: }
1.203 brouard 3842: } /* End loop k */
1.126 brouard 3843: }
3844: delti[theta]=delts;
3845: return res;
3846:
3847: }
3848:
1.203 brouard 3849: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 3850: {
3851: int i;
1.164 brouard 3852: int l=1, lmax=20;
1.126 brouard 3853: double k1,k2,k3,k4,res,fx;
1.132 brouard 3854: double p2[MAXPARM+1];
1.203 brouard 3855: int k, kmax=1;
3856: double v1, v2, cv12, lc1, lc2;
1.208 brouard 3857:
3858: int firstime=0;
1.203 brouard 3859:
1.126 brouard 3860: fx=func(x);
1.203 brouard 3861: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 3862: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 3863: p2[thetai]=x[thetai]+delti[thetai]*k;
3864: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 3865: k1=func(p2)-fx;
3866:
1.203 brouard 3867: p2[thetai]=x[thetai]+delti[thetai]*k;
3868: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 3869: k2=func(p2)-fx;
3870:
1.203 brouard 3871: p2[thetai]=x[thetai]-delti[thetai]*k;
3872: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 3873: k3=func(p2)-fx;
3874:
1.203 brouard 3875: p2[thetai]=x[thetai]-delti[thetai]*k;
3876: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 3877: k4=func(p2)-fx;
1.203 brouard 3878: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
3879: if(k1*k2*k3*k4 <0.){
1.208 brouard 3880: firstime=1;
1.203 brouard 3881: kmax=kmax+10;
1.208 brouard 3882: }
3883: if(kmax >=10 || firstime ==1){
1.218 brouard 3884: printf("Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you may increase ftol=%.2e\n",thetai,thetaj, ftol);
3885: fprintf(ficlog,"Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you may increase ftol=%.2e\n",thetai,thetaj, ftol);
1.203 brouard 3886: 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);
3887: 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);
3888: }
3889: #ifdef DEBUGHESSIJ
3890: v1=hess[thetai][thetai];
3891: v2=hess[thetaj][thetaj];
3892: cv12=res;
3893: /* Computing eigen value of Hessian matrix */
3894: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
3895: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
3896: if ((lc2 <0) || (lc1 <0) ){
3897: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
3898: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
3899: 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);
3900: 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);
3901: }
1.126 brouard 3902: #endif
3903: }
3904: return res;
3905: }
3906:
1.203 brouard 3907: /* Not done yet: Was supposed to fix if not exactly at the maximum */
3908: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
3909: /* { */
3910: /* int i; */
3911: /* int l=1, lmax=20; */
3912: /* double k1,k2,k3,k4,res,fx; */
3913: /* double p2[MAXPARM+1]; */
3914: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
3915: /* int k=0,kmax=10; */
3916: /* double l1; */
3917:
3918: /* fx=func(x); */
3919: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
3920: /* l1=pow(10,l); */
3921: /* delts=delt; */
3922: /* for(k=1 ; k <kmax; k=k+1){ */
3923: /* delt = delti*(l1*k); */
3924: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
3925: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
3926: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
3927: /* k1=func(p2)-fx; */
3928:
3929: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
3930: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
3931: /* k2=func(p2)-fx; */
3932:
3933: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
3934: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
3935: /* k3=func(p2)-fx; */
3936:
3937: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
3938: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
3939: /* k4=func(p2)-fx; */
3940: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
3941: /* #ifdef DEBUGHESSIJ */
3942: /* 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); */
3943: /* 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); */
3944: /* #endif */
3945: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
3946: /* k=kmax; */
3947: /* } */
3948: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
3949: /* k=kmax; l=lmax*10; */
3950: /* } */
3951: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
3952: /* delts=delt; */
3953: /* } */
3954: /* } /\* End loop k *\/ */
3955: /* } */
3956: /* delti[theta]=delts; */
3957: /* return res; */
3958: /* } */
3959:
3960:
1.126 brouard 3961: /************** Inverse of matrix **************/
3962: void ludcmp(double **a, int n, int *indx, double *d)
3963: {
3964: int i,imax,j,k;
3965: double big,dum,sum,temp;
3966: double *vv;
3967:
3968: vv=vector(1,n);
3969: *d=1.0;
3970: for (i=1;i<=n;i++) {
3971: big=0.0;
3972: for (j=1;j<=n;j++)
3973: if ((temp=fabs(a[i][j])) > big) big=temp;
3974: if (big == 0.0) nrerror("Singular matrix in routine ludcmp");
3975: vv[i]=1.0/big;
3976: }
3977: for (j=1;j<=n;j++) {
3978: for (i=1;i<j;i++) {
3979: sum=a[i][j];
3980: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
3981: a[i][j]=sum;
3982: }
3983: big=0.0;
3984: for (i=j;i<=n;i++) {
3985: sum=a[i][j];
3986: for (k=1;k<j;k++)
3987: sum -= a[i][k]*a[k][j];
3988: a[i][j]=sum;
3989: if ( (dum=vv[i]*fabs(sum)) >= big) {
3990: big=dum;
3991: imax=i;
3992: }
3993: }
3994: if (j != imax) {
3995: for (k=1;k<=n;k++) {
3996: dum=a[imax][k];
3997: a[imax][k]=a[j][k];
3998: a[j][k]=dum;
3999: }
4000: *d = -(*d);
4001: vv[imax]=vv[j];
4002: }
4003: indx[j]=imax;
4004: if (a[j][j] == 0.0) a[j][j]=TINY;
4005: if (j != n) {
4006: dum=1.0/(a[j][j]);
4007: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4008: }
4009: }
4010: free_vector(vv,1,n); /* Doesn't work */
4011: ;
4012: }
4013:
4014: void lubksb(double **a, int n, int *indx, double b[])
4015: {
4016: int i,ii=0,ip,j;
4017: double sum;
4018:
4019: for (i=1;i<=n;i++) {
4020: ip=indx[i];
4021: sum=b[ip];
4022: b[ip]=b[i];
4023: if (ii)
4024: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4025: else if (sum) ii=i;
4026: b[i]=sum;
4027: }
4028: for (i=n;i>=1;i--) {
4029: sum=b[i];
4030: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4031: b[i]=sum/a[i][i];
4032: }
4033: }
4034:
4035: void pstamp(FILE *fichier)
4036: {
1.196 brouard 4037: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4038: }
4039:
4040: /************ Frequencies ********************/
1.226 brouard 4041: void freqsummary(char fileres[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
4042: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4043: int firstpass, int lastpass, int stepm, int weightopt, char model[])
4044: { /* Some frequencies */
4045:
1.227 brouard 4046: int i, m, jk, j1, bool, z1,j, k, iv;
1.226 brouard 4047: int iind=0, iage=0;
4048: int mi; /* Effective wave */
4049: int first;
4050: double ***freq; /* Frequencies */
4051: double *meanq;
4052: double **meanqt;
4053: double *pp, **prop, *posprop, *pospropt;
4054: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4055: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4056: double agebegin, ageend;
4057:
4058: pp=vector(1,nlstate);
4059: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4060: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4061: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4062: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4063: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4064: meanqt=matrix(1,lastpass,1,nqtveff);
4065: strcpy(fileresp,"P_");
4066: strcat(fileresp,fileresu);
4067: /*strcat(fileresphtm,fileresu);*/
4068: if((ficresp=fopen(fileresp,"w"))==NULL) {
4069: printf("Problem with prevalence resultfile: %s\n", fileresp);
4070: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4071: exit(0);
4072: }
1.214 brouard 4073:
1.226 brouard 4074: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4075: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4076: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4077: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4078: fflush(ficlog);
4079: exit(70);
4080: }
4081: else{
4082: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.214 brouard 4083: <hr size=\"2\" color=\"#EC5E5E\"> \n\
4084: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4085: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4086: }
4087: fprintf(ficresphtm,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies and prevalence by age at begin of transition</h4>\n",fileresphtm, fileresphtm);
1.214 brouard 4088:
1.226 brouard 4089: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4090: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4091: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4092: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4093: fflush(ficlog);
4094: exit(70);
4095: }
4096: else{
4097: fprintf(ficresphtmfr,"<html><head>\n<title>IMaCh PHTM_Frequency table %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.214 brouard 4098: <hr size=\"2\" color=\"#EC5E5E\"> \n\
4099: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4100: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4101: }
4102: fprintf(ficresphtmfr,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies of all effective transitions by age at begin of transition </h4>Unknown status is -1<br/>\n",fileresphtmfr, fileresphtmfr);
1.214 brouard 4103:
1.226 brouard 4104: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4105: j1=0;
1.126 brouard 4106:
1.227 brouard 4107: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4108: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4109: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.220 brouard 4110:
1.226 brouard 4111: first=1;
1.220 brouard 4112:
1.226 brouard 4113: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4114: reference=low_education V1=0,V2=0
4115: med_educ V1=1 V2=0,
4116: high_educ V1=0 V2=1
4117: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4118: */
1.126 brouard 4119:
1.227 brouard 4120: 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 4121: posproptt=0.;
4122: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4123: scanf("%d", i);*/
4124: for (i=-5; i<=nlstate+ndeath; i++)
4125: for (jk=-5; jk<=nlstate+ndeath; jk++)
1.231 brouard 4126: for(m=iagemin; m <= iagemax+3; m++)
4127: freq[i][jk][m]=0;
4128:
1.226 brouard 4129: for (i=1; i<=nlstate; i++) {
4130: for(m=iagemin; m <= iagemax+3; m++)
1.231 brouard 4131: prop[i][m]=0;
1.226 brouard 4132: posprop[i]=0;
4133: pospropt[i]=0;
4134: }
1.227 brouard 4135: /* for (z1=1; z1<= nqfveff; z1++) { */
4136: /* meanq[z1]+=0.; */
4137: /* for(m=1;m<=lastpass;m++){ */
4138: /* meanqt[m][z1]=0.; */
4139: /* } */
4140: /* } */
1.231 brouard 4141:
1.226 brouard 4142: dateintsum=0;
4143: k2cpt=0;
1.227 brouard 4144: /* For that combination of covariate j1, we count and print the frequencies in one pass */
1.226 brouard 4145: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4146: bool=1;
1.227 brouard 4147: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
1.234 brouard 4148: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
1.227 brouard 4149: /* for (z1=1; z1<= nqfveff; z1++) { */
4150: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4151: /* } */
1.234 brouard 4152: for (z1=1; z1<=cptcoveff; z1++) {
4153: /* if(Tvaraff[z1] ==-20){ */
4154: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4155: /* }else if(Tvaraff[z1] ==-10){ */
4156: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4157: /* }else */
4158: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){
4159: /* Tests if this individual iind responded to j1 (V4=1 V3=0) */
4160: bool=0;
4161: /* 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",
4162: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4163: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4164: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4165: } /* Onlyf fixed */
4166: } /* end z1 */
4167: } /* cptcovn > 0 */
1.227 brouard 4168: } /* end any */
4169: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
1.234 brouard 4170: /* for(m=firstpass; m<=lastpass; m++){ */
4171: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4172: m=mw[mi][iind];
4173: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4174: for (z1=1; z1<=cptcoveff; z1++) {
4175: if( Fixed[Tmodelind[z1]]==1){
4176: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4177: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4178: bool=0;
4179: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4180: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4181: bool=0;
4182: }
4183: }
4184: }
4185: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4186: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4187: if(bool==1){
4188: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4189: and mw[mi+1][iind]. dh depends on stepm. */
4190: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4191: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4192: if(m >=firstpass && m <=lastpass){
4193: k2=anint[m][iind]+(mint[m][iind]/12.);
4194: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4195: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4196: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4197: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4198: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4199: if (m<lastpass) {
4200: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4201: /* 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]); */
4202: if(s[m][iind]==-1)
4203: 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.));
4204: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4205: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4206: 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 */
4207: }
4208: } /* end if between passes */
4209: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99)) {
4210: dateintsum=dateintsum+k2;
4211: k2cpt++;
4212: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
4213: }
4214: } /* end bool 2 */
4215: } /* end m */
1.226 brouard 4216: } /* end bool */
4217: } /* end iind = 1 to imx */
4218: /* prop[s][age] is feeded for any initial and valid live state as well as
4219: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
1.231 brouard 4220:
4221:
1.226 brouard 4222: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4223: pstamp(ficresp);
1.227 brouard 4224: /* if (ncoveff>0) { */
4225: if (cptcoveff>0) {
1.226 brouard 4226: fprintf(ficresp, "\n#********** Variable ");
4227: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4228: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
1.227 brouard 4229: for (z1=1; z1<=cptcoveff; z1++){
1.234 brouard 4230: fprintf(ficresp, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4231: fprintf(ficresphtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4232: fprintf(ficresphtmfr, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.226 brouard 4233: }
4234: fprintf(ficresp, "**********\n#");
4235: fprintf(ficresphtm, "**********</h3>\n");
4236: fprintf(ficresphtmfr, "**********</h3>\n");
4237: fprintf(ficlog, "\n#********** Variable ");
1.227 brouard 4238: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficlog, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.226 brouard 4239: fprintf(ficlog, "**********\n");
4240: }
4241: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4242: for(i=1; i<=nlstate;i++) {
4243: fprintf(ficresp, " Age Prev(%d) N(%d) N",i,i);
4244: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4245: }
4246: fprintf(ficresp, "\n");
4247: fprintf(ficresphtm, "\n");
1.231 brouard 4248:
1.226 brouard 4249: /* Header of frequency table by age */
4250: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4251: fprintf(ficresphtmfr,"<th>Age</th> ");
4252: for(jk=-1; jk <=nlstate+ndeath; jk++){
4253: for(m=-1; m <=nlstate+ndeath; m++){
1.234 brouard 4254: if(jk!=0 && m!=0)
4255: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.226 brouard 4256: }
4257: }
4258: fprintf(ficresphtmfr, "\n");
1.231 brouard 4259:
1.226 brouard 4260: /* For each age */
4261: for(iage=iagemin; iage <= iagemax+3; iage++){
4262: fprintf(ficresphtm,"<tr>");
4263: if(iage==iagemax+1){
1.231 brouard 4264: fprintf(ficlog,"1");
4265: fprintf(ficresphtmfr,"<tr><th>0</th> ");
1.226 brouard 4266: }else if(iage==iagemax+2){
1.231 brouard 4267: fprintf(ficlog,"0");
4268: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
1.226 brouard 4269: }else if(iage==iagemax+3){
1.231 brouard 4270: fprintf(ficlog,"Total");
4271: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
1.226 brouard 4272: }else{
1.231 brouard 4273: if(first==1){
4274: first=0;
4275: printf("See log file for details...\n");
4276: }
4277: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4278: fprintf(ficlog,"Age %d", iage);
1.226 brouard 4279: }
4280: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4281: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4282: pp[jk] += freq[jk][m][iage];
1.226 brouard 4283: }
4284: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4285: for(m=-1, pos=0; m <=0 ; m++)
4286: pos += freq[jk][m][iage];
4287: if(pp[jk]>=1.e-10){
4288: if(first==1){
4289: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4290: }
4291: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4292: }else{
4293: if(first==1)
4294: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4295: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4296: }
1.226 brouard 4297: }
1.231 brouard 4298:
1.226 brouard 4299: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4300: /* posprop[jk]=0; */
4301: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4302: pp[jk] += freq[jk][m][iage];
1.226 brouard 4303: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
1.231 brouard 4304:
1.226 brouard 4305: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
1.231 brouard 4306: pos += pp[jk]; /* pos is the total number of transitions until this age */
4307: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4308: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4309: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
4310: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
1.226 brouard 4311: }
4312: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4313: if(pos>=1.e-5){
4314: if(first==1)
4315: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4316: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4317: }else{
4318: if(first==1)
4319: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4320: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4321: }
4322: if( iage <= iagemax){
4323: if(pos>=1.e-5){
4324: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4325: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4326: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4327: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4328: }
4329: else{
4330: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4331: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4332: }
4333: }
4334: pospropt[jk] +=posprop[jk];
1.226 brouard 4335: } /* end loop jk */
4336: /* pospropt=0.; */
4337: for(jk=-1; jk <=nlstate+ndeath; jk++){
1.231 brouard 4338: for(m=-1; m <=nlstate+ndeath; m++){
4339: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4340: if(first==1){
4341: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4342: }
4343: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4344: }
4345: if(jk!=0 && m!=0)
4346: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
4347: }
1.226 brouard 4348: } /* end loop jk */
4349: posproptt=0.;
4350: for(jk=1; jk <=nlstate; jk++){
1.231 brouard 4351: posproptt += pospropt[jk];
1.226 brouard 4352: }
4353: fprintf(ficresphtmfr,"</tr>\n ");
4354: if(iage <= iagemax){
1.231 brouard 4355: fprintf(ficresp,"\n");
4356: fprintf(ficresphtm,"</tr>\n");
1.226 brouard 4357: }
4358: if(first==1)
1.231 brouard 4359: printf("Others in log...\n");
1.226 brouard 4360: fprintf(ficlog,"\n");
4361: } /* end loop age iage */
4362: fprintf(ficresphtm,"<tr><th>Tot</th>");
4363: for(jk=1; jk <=nlstate ; jk++){
4364: if(posproptt < 1.e-5){
1.231 brouard 4365: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
1.226 brouard 4366: }else{
1.231 brouard 4367: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.226 brouard 4368: }
4369: }
4370: fprintf(ficresphtm,"</tr>\n");
4371: fprintf(ficresphtm,"</table>\n");
4372: fprintf(ficresphtmfr,"</table>\n");
4373: if(posproptt < 1.e-5){
4374: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4375: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4376: fprintf(ficres,"\n This combination (%d) is not valid and no result will be produced\n\n",j1);
4377: invalidvarcomb[j1]=1;
4378: }else{
4379: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4380: invalidvarcomb[j1]=0;
4381: }
4382: fprintf(ficresphtmfr,"</table>\n");
4383: } /* end selected combination of covariate j1 */
4384: dateintmean=dateintsum/k2cpt;
1.231 brouard 4385:
1.226 brouard 4386: fclose(ficresp);
4387: fclose(ficresphtm);
4388: fclose(ficresphtmfr);
4389: free_vector(meanq,1,nqfveff);
4390: free_matrix(meanqt,1,lastpass,1,nqtveff);
4391: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4392: free_vector(pospropt,1,nlstate);
4393: free_vector(posprop,1,nlstate);
4394: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4395: free_vector(pp,1,nlstate);
4396: /* End of freqsummary */
4397: }
1.126 brouard 4398:
4399: /************ Prevalence ********************/
1.227 brouard 4400: 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)
4401: {
4402: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4403: in each health status at the date of interview (if between dateprev1 and dateprev2).
4404: We still use firstpass and lastpass as another selection.
4405: */
1.126 brouard 4406:
1.227 brouard 4407: int i, m, jk, j1, bool, z1,j, iv;
4408: int mi; /* Effective wave */
4409: int iage;
4410: double agebegin, ageend;
4411:
4412: double **prop;
4413: double posprop;
4414: double y2; /* in fractional years */
4415: int iagemin, iagemax;
4416: int first; /** to stop verbosity which is redirected to log file */
4417:
4418: iagemin= (int) agemin;
4419: iagemax= (int) agemax;
4420: /*pp=vector(1,nlstate);*/
4421: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4422: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4423: j1=0;
1.222 brouard 4424:
1.227 brouard 4425: /*j=cptcoveff;*/
4426: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4427:
1.227 brouard 4428: first=1;
4429: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4430: for (i=1; i<=nlstate; i++)
4431: for(iage=iagemin-AGEMARGE; iage <= iagemax+3+AGEMARGE; iage++)
4432: prop[i][iage]=0.0;
4433: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4434: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4435: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4436:
4437: for (i=1; i<=imx; i++) { /* Each individual */
4438: bool=1;
4439: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4440: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4441: m=mw[mi][i];
4442: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4443: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4444: for (z1=1; z1<=cptcoveff; z1++){
4445: if( Fixed[Tmodelind[z1]]==1){
4446: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4447: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4448: bool=0;
4449: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4450: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4451: bool=0;
4452: }
4453: }
4454: if(bool==1){ /* Otherwise we skip that wave/person */
4455: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4456: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4457: if(m >=firstpass && m <=lastpass){
4458: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4459: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4460: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4461: if(agev[m][i]==1) agev[m][i]=iagemax+2;
4462: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+3+AGEMARGE){
4463: 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);
4464: exit(1);
4465: }
4466: if (s[m][i]>0 && s[m][i]<=nlstate) {
4467: /*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]]);*/
4468: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4469: prop[s[m][i]][iagemax+3] += weight[i];
4470: } /* end valid statuses */
4471: } /* end selection of dates */
4472: } /* end selection of waves */
4473: } /* end bool */
4474: } /* end wave */
4475: } /* end individual */
4476: for(i=iagemin; i <= iagemax+3; i++){
4477: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4478: posprop += prop[jk][i];
4479: }
4480:
4481: for(jk=1; jk <=nlstate ; jk++){
4482: if( i <= iagemax){
4483: if(posprop>=1.e-5){
4484: probs[i][jk][j1]= prop[jk][i]/posprop;
4485: } else{
4486: if(first==1){
4487: first=0;
4488: 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]);
4489: }
4490: }
4491: }
4492: }/* end jk */
4493: }/* end i */
1.222 brouard 4494: /*} *//* end i1 */
1.227 brouard 4495: } /* end j1 */
1.222 brouard 4496:
1.227 brouard 4497: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4498: /*free_vector(pp,1,nlstate);*/
4499: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4500: } /* End of prevalence */
1.126 brouard 4501:
4502: /************* Waves Concatenation ***************/
4503:
4504: 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)
4505: {
4506: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4507: Death is a valid wave (if date is known).
4508: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4509: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4510: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4511: */
1.126 brouard 4512:
1.224 brouard 4513: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4514: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4515: double sum=0., jmean=0.;*/
1.224 brouard 4516: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4517: int j, k=0,jk, ju, jl;
4518: double sum=0.;
4519: first=0;
1.214 brouard 4520: firstwo=0;
1.217 brouard 4521: firsthree=0;
1.218 brouard 4522: firstfour=0;
1.164 brouard 4523: jmin=100000;
1.126 brouard 4524: jmax=-1;
4525: jmean=0.;
1.224 brouard 4526:
4527: /* Treating live states */
1.214 brouard 4528: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4529: mi=0; /* First valid wave */
1.227 brouard 4530: mli=0; /* Last valid wave */
1.126 brouard 4531: m=firstpass;
1.214 brouard 4532: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4533: 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 */
4534: mli=m-1;/* mw[++mi][i]=m-1; */
4535: }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 */
4536: mw[++mi][i]=m;
4537: mli=m;
1.224 brouard 4538: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4539: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4540: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4541: }
1.227 brouard 4542: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4543: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4544: break;
1.224 brouard 4545: #else
1.227 brouard 4546: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4547: if(firsthree == 0){
4548: 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);
4549: firsthree=1;
4550: }
4551: 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);
4552: mw[++mi][i]=m;
4553: mli=m;
4554: }
4555: if(s[m][i]==-2){ /* Vital status is really unknown */
4556: nbwarn++;
4557: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4558: 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);
4559: 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);
4560: }
4561: break;
4562: }
4563: break;
1.224 brouard 4564: #endif
1.227 brouard 4565: }/* End m >= lastpass */
1.126 brouard 4566: }/* end while */
1.224 brouard 4567:
1.227 brouard 4568: /* 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 4569: /* After last pass */
1.224 brouard 4570: /* Treating death states */
1.214 brouard 4571: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4572: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4573: /* } */
1.126 brouard 4574: mi++; /* Death is another wave */
4575: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4576: /* Only death is a correct wave */
1.126 brouard 4577: mw[mi][i]=m;
1.224 brouard 4578: }
4579: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.227 brouard 4580: 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 4581: /* m++; */
4582: /* mi++; */
4583: /* s[m][i]=nlstate+1; /\* We are setting the status to the last of non live state *\/ */
4584: /* mw[mi][i]=m; */
1.218 brouard 4585: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4586: 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 */
4587: nbwarn++;
4588: if(firstfiv==0){
4589: 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 );
4590: firstfiv=1;
4591: }else{
4592: 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 );
4593: }
4594: }else{ /* Death occured afer last wave potential bias */
4595: nberr++;
4596: if(firstwo==0){
4597: 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 );
4598: firstwo=1;
4599: }
4600: 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 );
4601: }
1.218 brouard 4602: }else{ /* end date of interview is known */
1.227 brouard 4603: /* death is known but not confirmed by death status at any wave */
4604: if(firstfour==0){
4605: 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 );
4606: firstfour=1;
4607: }
4608: 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 4609: }
1.224 brouard 4610: } /* end if date of death is known */
4611: #endif
4612: wav[i]=mi; /* mi should be the last effective wave (or mli) */
4613: /* wav[i]=mw[mi][i]; */
1.126 brouard 4614: if(mi==0){
4615: nbwarn++;
4616: if(first==0){
1.227 brouard 4617: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
4618: first=1;
1.126 brouard 4619: }
4620: if(first==1){
1.227 brouard 4621: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 4622: }
4623: } /* end mi==0 */
4624: } /* End individuals */
1.214 brouard 4625: /* wav and mw are no more changed */
1.223 brouard 4626:
1.214 brouard 4627:
1.126 brouard 4628: for(i=1; i<=imx; i++){
4629: for(mi=1; mi<wav[i];mi++){
4630: if (stepm <=0)
1.227 brouard 4631: dh[mi][i]=1;
1.126 brouard 4632: else{
1.227 brouard 4633: if (s[mw[mi+1][i]][i] > nlstate) { /* A death */
4634: if (agedc[i] < 2*AGESUP) {
4635: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
4636: if(j==0) j=1; /* Survives at least one month after exam */
4637: else if(j<0){
4638: nberr++;
4639: 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]);
4640: j=1; /* Temporary Dangerous patch */
4641: 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);
4642: 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]);
4643: 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);
4644: }
4645: k=k+1;
4646: if (j >= jmax){
4647: jmax=j;
4648: ijmax=i;
4649: }
4650: if (j <= jmin){
4651: jmin=j;
4652: ijmin=i;
4653: }
4654: sum=sum+j;
4655: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
4656: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
4657: }
4658: }
4659: else{
4660: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 4661: /* 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 4662:
1.227 brouard 4663: k=k+1;
4664: if (j >= jmax) {
4665: jmax=j;
4666: ijmax=i;
4667: }
4668: else if (j <= jmin){
4669: jmin=j;
4670: ijmin=i;
4671: }
4672: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
4673: /*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]);*/
4674: if(j<0){
4675: nberr++;
4676: 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]);
4677: 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]);
4678: }
4679: sum=sum+j;
4680: }
4681: jk= j/stepm;
4682: jl= j -jk*stepm;
4683: ju= j -(jk+1)*stepm;
4684: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
4685: if(jl==0){
4686: dh[mi][i]=jk;
4687: bh[mi][i]=0;
4688: }else{ /* We want a negative bias in order to only have interpolation ie
4689: * to avoid the price of an extra matrix product in likelihood */
4690: dh[mi][i]=jk+1;
4691: bh[mi][i]=ju;
4692: }
4693: }else{
4694: if(jl <= -ju){
4695: dh[mi][i]=jk;
4696: bh[mi][i]=jl; /* bias is positive if real duration
4697: * is higher than the multiple of stepm and negative otherwise.
4698: */
4699: }
4700: else{
4701: dh[mi][i]=jk+1;
4702: bh[mi][i]=ju;
4703: }
4704: if(dh[mi][i]==0){
4705: dh[mi][i]=1; /* At least one step */
4706: bh[mi][i]=ju; /* At least one step */
4707: /* 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);*/
4708: }
4709: } /* end if mle */
1.126 brouard 4710: }
4711: } /* end wave */
4712: }
4713: jmean=sum/k;
4714: 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 4715: 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 4716: }
1.126 brouard 4717:
4718: /*********** Tricode ****************************/
1.220 brouard 4719: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.126 brouard 4720: {
1.144 brouard 4721: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
4722: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
1.169 brouard 4723: * Boring subroutine which should only output nbcode[Tvar[j]][k]
1.224 brouard 4724: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
4725: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
1.144 brouard 4726: */
1.130 brouard 4727:
1.145 brouard 4728: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
1.136 brouard 4729: int modmaxcovj=0; /* Modality max of covariates j */
1.145 brouard 4730: int cptcode=0; /* Modality max of covariates j */
4731: int modmincovj=0; /* Modality min of covariates j */
4732:
4733:
1.220 brouard 4734: /* cptcoveff=0; */
1.224 brouard 4735: /* *cptcov=0; */
1.126 brouard 4736:
1.144 brouard 4737: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 4738:
1.224 brouard 4739: /* Loop on covariates without age and products and no quantitative variable */
4740: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
1.227 brouard 4741: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
4742: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
4743: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
4744: switch(Fixed[k]) {
4745: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
1.231 brouard 4746: 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*/
4747: ij=(int)(covar[Tvar[k]][i]);
4748: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
4749: * If product of Vn*Vm, still boolean *:
4750: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
4751: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
4752: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
4753: modality of the nth covariate of individual i. */
4754: if (ij > modmaxcovj)
4755: modmaxcovj=ij;
4756: else if (ij < modmincovj)
4757: modmincovj=ij;
4758: if ((ij < -1) && (ij > NCOVMAX)){
4759: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
4760: exit(1);
4761: }else
4762: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
4763: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
4764: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
4765: /* getting the maximum value of the modality of the covariate
4766: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
4767: female ies 1, then modmaxcovj=1.
4768: */
4769: } /* end for loop on individuals i */
4770: printf(" Minimal and maximal values of %d th covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4771: fprintf(ficlog," Minimal and maximal values of %d th covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4772: cptcode=modmaxcovj;
4773: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
4774: /*for (i=0; i<=cptcode; i++) {*/
4775: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
4776: printf("Frequencies of covariates %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4777: fprintf(ficlog, "Frequencies of covariates %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4778: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
4779: if( j != -1){
4780: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
4781: covariate for which somebody answered excluding
4782: undefined. Usually 2: 0 and 1. */
4783: }
4784: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
4785: covariate for which somebody answered including
4786: undefined. Usually 3: -1, 0 and 1. */
4787: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
4788: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
4789: } /* Ndum[-1] number of undefined modalities */
4790:
4791: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
4792: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
4793: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
4794: /* modmincovj=3; modmaxcovj = 7; */
4795: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
4796: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
4797: /* defining two dummy variables: variables V1_1 and V1_2.*/
4798: /* nbcode[Tvar[j]][ij]=k; */
4799: /* nbcode[Tvar[j]][1]=0; */
4800: /* nbcode[Tvar[j]][2]=1; */
4801: /* nbcode[Tvar[j]][3]=2; */
4802: /* To be continued (not working yet). */
4803: ij=0; /* ij is similar to i but can jump over null modalities */
4804: 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*/
4805: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
4806: break;
4807: }
4808: ij++;
4809: 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*/
4810: cptcode = ij; /* New max modality for covar j */
4811: } /* end of loop on modality i=-1 to 1 or more */
4812: break;
1.227 brouard 4813: case 1: /* Testing on varying covariate, could be simple and
4814: * should look at waves or product of fixed *
4815: * varying. No time to test -1, assuming 0 and 1 only */
1.231 brouard 4816: ij=0;
4817: for(i=0; i<=1;i++){
4818: nbcode[Tvar[k]][++ij]=i;
4819: }
4820: break;
1.227 brouard 4821: default:
1.231 brouard 4822: break;
1.227 brouard 4823: } /* end switch */
4824: } /* end dummy test */
1.225 brouard 4825:
1.192 brouard 4826: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
4827: /* /\*recode from 0 *\/ */
4828: /* k is a modality. If we have model=V1+V1*sex */
4829: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
4830: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
4831: /* } */
4832: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
4833: /* if (ij > ncodemax[j]) { */
4834: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4835: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4836: /* break; */
4837: /* } */
4838: /* } /\* end of loop on modality k *\/ */
1.137 brouard 4839: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
4840:
1.225 brouard 4841: for (k=-1; k< maxncov; k++) Ndum[k]=0;
1.227 brouard 4842: /* Look at fixed dummy (single or product) covariates to check empty modalities */
1.187 brouard 4843: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
1.225 brouard 4844: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
1.227 brouard 4845: 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 */
4846: 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 */
4847: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
1.225 brouard 4848: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
4849:
4850: ij=0;
1.227 brouard 4851: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
4852: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
1.225 brouard 4853: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
1.227 brouard 4854: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
4855: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
4856: /* If product not in single variable we don't print results */
1.225 brouard 4857: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
1.230 brouard 4858: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
4859: 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*/
4860: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
1.231 brouard 4861: 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 */
1.227 brouard 4862: if(Fixed[k]!=0)
4863: anyvaryingduminmodel=1;
1.231 brouard 4864: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
4865: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
4866: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
4867: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
4868: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
4869: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
1.227 brouard 4870: }
1.225 brouard 4871: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
4872: /* ij--; */
4873: /* cptcoveff=ij; /\*Number of total covariates*\/ */
4874: *cptcov=ij; /*Number of total real effective covariates: effective
1.231 brouard 4875: * because they can be excluded from the model and real
4876: * if in the model but excluded because missing values, but how to get k from ij?*/
1.227 brouard 4877: for(j=ij+1; j<= cptcovt; j++){
4878: Tvaraff[j]=0;
4879: Tmodelind[j]=0;
4880: }
1.228 brouard 4881: for(j=ntveff+1; j<= cptcovt; j++){
4882: TmodelInvind[j]=0;
4883: }
1.227 brouard 4884: /* To be sorted */
4885: ;
1.126 brouard 4886: }
4887:
1.145 brouard 4888:
1.126 brouard 4889: /*********** Health Expectancies ****************/
4890:
1.235 brouard 4891: 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 4892:
4893: {
4894: /* Health expectancies, no variances */
1.164 brouard 4895: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 4896: int nhstepma, nstepma; /* Decreasing with age */
4897: double age, agelim, hf;
4898: double ***p3mat;
4899: double eip;
4900:
4901: pstamp(ficreseij);
4902: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
4903: fprintf(ficreseij,"# Age");
4904: for(i=1; i<=nlstate;i++){
4905: for(j=1; j<=nlstate;j++){
4906: fprintf(ficreseij," e%1d%1d ",i,j);
4907: }
4908: fprintf(ficreseij," e%1d. ",i);
4909: }
4910: fprintf(ficreseij,"\n");
4911:
4912:
4913: if(estepm < stepm){
4914: printf ("Problem %d lower than %d\n",estepm, stepm);
4915: }
4916: else hstepm=estepm;
4917: /* We compute the life expectancy from trapezoids spaced every estepm months
4918: * This is mainly to measure the difference between two models: for example
4919: * if stepm=24 months pijx are given only every 2 years and by summing them
4920: * we are calculating an estimate of the Life Expectancy assuming a linear
4921: * progression in between and thus overestimating or underestimating according
4922: * to the curvature of the survival function. If, for the same date, we
4923: * estimate the model with stepm=1 month, we can keep estepm to 24 months
4924: * to compare the new estimate of Life expectancy with the same linear
4925: * hypothesis. A more precise result, taking into account a more precise
4926: * curvature will be obtained if estepm is as small as stepm. */
4927:
4928: /* For example we decided to compute the life expectancy with the smallest unit */
4929: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
4930: nhstepm is the number of hstepm from age to agelim
4931: nstepm is the number of stepm from age to agelin.
4932: Look at hpijx to understand the reason of that which relies in memory size
4933: and note for a fixed period like estepm months */
4934: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
4935: survival function given by stepm (the optimization length). Unfortunately it
4936: means that if the survival funtion is printed only each two years of age and if
4937: you sum them up and add 1 year (area under the trapezoids) you won't get the same
4938: results. So we changed our mind and took the option of the best precision.
4939: */
4940: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
4941:
4942: agelim=AGESUP;
4943: /* If stepm=6 months */
4944: /* Computed by stepm unit matrices, product of hstepm matrices, stored
4945: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
4946:
4947: /* nhstepm age range expressed in number of stepm */
4948: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
4949: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
4950: /* if (stepm >= YEARM) hstepm=1;*/
4951: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
4952: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
4953:
4954: for (age=bage; age<=fage; age ++){
4955: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
4956: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
4957: /* if (stepm >= YEARM) hstepm=1;*/
4958: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
4959:
4960: /* If stepm=6 months */
4961: /* Computed by stepm unit matrices, product of hstepma matrices, stored
4962: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
4963:
1.235 brouard 4964: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 4965:
4966: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
4967:
4968: printf("%d|",(int)age);fflush(stdout);
4969: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
4970:
4971: /* Computing expectancies */
4972: for(i=1; i<=nlstate;i++)
4973: for(j=1; j<=nlstate;j++)
4974: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
4975: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
4976:
4977: /* 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]);*/
4978:
4979: }
4980:
4981: fprintf(ficreseij,"%3.0f",age );
4982: for(i=1; i<=nlstate;i++){
4983: eip=0;
4984: for(j=1; j<=nlstate;j++){
4985: eip +=eij[i][j][(int)age];
4986: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
4987: }
4988: fprintf(ficreseij,"%9.4f", eip );
4989: }
4990: fprintf(ficreseij,"\n");
4991:
4992: }
4993: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
4994: printf("\n");
4995: fprintf(ficlog,"\n");
4996:
4997: }
4998:
1.235 brouard 4999: 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 5000:
5001: {
5002: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5003: to initial status i, ei. .
1.126 brouard 5004: */
5005: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5006: int nhstepma, nstepma; /* Decreasing with age */
5007: double age, agelim, hf;
5008: double ***p3matp, ***p3matm, ***varhe;
5009: double **dnewm,**doldm;
5010: double *xp, *xm;
5011: double **gp, **gm;
5012: double ***gradg, ***trgradg;
5013: int theta;
5014:
5015: double eip, vip;
5016:
5017: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5018: xp=vector(1,npar);
5019: xm=vector(1,npar);
5020: dnewm=matrix(1,nlstate*nlstate,1,npar);
5021: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5022:
5023: pstamp(ficresstdeij);
5024: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5025: fprintf(ficresstdeij,"# Age");
5026: for(i=1; i<=nlstate;i++){
5027: for(j=1; j<=nlstate;j++)
5028: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5029: fprintf(ficresstdeij," e%1d. ",i);
5030: }
5031: fprintf(ficresstdeij,"\n");
5032:
5033: pstamp(ficrescveij);
5034: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5035: fprintf(ficrescveij,"# Age");
5036: for(i=1; i<=nlstate;i++)
5037: for(j=1; j<=nlstate;j++){
5038: cptj= (j-1)*nlstate+i;
5039: for(i2=1; i2<=nlstate;i2++)
5040: for(j2=1; j2<=nlstate;j2++){
5041: cptj2= (j2-1)*nlstate+i2;
5042: if(cptj2 <= cptj)
5043: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5044: }
5045: }
5046: fprintf(ficrescveij,"\n");
5047:
5048: if(estepm < stepm){
5049: printf ("Problem %d lower than %d\n",estepm, stepm);
5050: }
5051: else hstepm=estepm;
5052: /* We compute the life expectancy from trapezoids spaced every estepm months
5053: * This is mainly to measure the difference between two models: for example
5054: * if stepm=24 months pijx are given only every 2 years and by summing them
5055: * we are calculating an estimate of the Life Expectancy assuming a linear
5056: * progression in between and thus overestimating or underestimating according
5057: * to the curvature of the survival function. If, for the same date, we
5058: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5059: * to compare the new estimate of Life expectancy with the same linear
5060: * hypothesis. A more precise result, taking into account a more precise
5061: * curvature will be obtained if estepm is as small as stepm. */
5062:
5063: /* For example we decided to compute the life expectancy with the smallest unit */
5064: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5065: nhstepm is the number of hstepm from age to agelim
5066: nstepm is the number of stepm from age to agelin.
5067: Look at hpijx to understand the reason of that which relies in memory size
5068: and note for a fixed period like estepm months */
5069: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5070: survival function given by stepm (the optimization length). Unfortunately it
5071: means that if the survival funtion is printed only each two years of age and if
5072: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5073: results. So we changed our mind and took the option of the best precision.
5074: */
5075: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5076:
5077: /* If stepm=6 months */
5078: /* nhstepm age range expressed in number of stepm */
5079: agelim=AGESUP;
5080: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5081: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5082: /* if (stepm >= YEARM) hstepm=1;*/
5083: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5084:
5085: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5086: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5087: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5088: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5089: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5090: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5091:
5092: for (age=bage; age<=fage; age ++){
5093: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5094: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5095: /* if (stepm >= YEARM) hstepm=1;*/
5096: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5097:
1.126 brouard 5098: /* If stepm=6 months */
5099: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5100: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5101:
5102: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5103:
1.126 brouard 5104: /* Computing Variances of health expectancies */
5105: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5106: decrease memory allocation */
5107: for(theta=1; theta <=npar; theta++){
5108: for(i=1; i<=npar; i++){
1.222 brouard 5109: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5110: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5111: }
1.235 brouard 5112: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5113: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5114:
1.126 brouard 5115: for(j=1; j<= nlstate; j++){
1.222 brouard 5116: for(i=1; i<=nlstate; i++){
5117: for(h=0; h<=nhstepm-1; h++){
5118: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5119: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5120: }
5121: }
1.126 brouard 5122: }
1.218 brouard 5123:
1.126 brouard 5124: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5125: for(h=0; h<=nhstepm-1; h++){
5126: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5127: }
1.126 brouard 5128: }/* End theta */
5129:
5130:
5131: for(h=0; h<=nhstepm-1; h++)
5132: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5133: for(theta=1; theta <=npar; theta++)
5134: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5135:
1.218 brouard 5136:
1.222 brouard 5137: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5138: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5139: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5140:
1.222 brouard 5141: printf("%d|",(int)age);fflush(stdout);
5142: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5143: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5144: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5145: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5146: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5147: for(ij=1;ij<=nlstate*nlstate;ij++)
5148: for(ji=1;ji<=nlstate*nlstate;ji++)
5149: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5150: }
5151: }
1.218 brouard 5152:
1.126 brouard 5153: /* Computing expectancies */
1.235 brouard 5154: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5155: for(i=1; i<=nlstate;i++)
5156: for(j=1; j<=nlstate;j++)
1.222 brouard 5157: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5158: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5159:
1.222 brouard 5160: /* 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 5161:
1.222 brouard 5162: }
1.218 brouard 5163:
1.126 brouard 5164: fprintf(ficresstdeij,"%3.0f",age );
5165: for(i=1; i<=nlstate;i++){
5166: eip=0.;
5167: vip=0.;
5168: for(j=1; j<=nlstate;j++){
1.222 brouard 5169: eip += eij[i][j][(int)age];
5170: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5171: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5172: 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 5173: }
5174: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5175: }
5176: fprintf(ficresstdeij,"\n");
1.218 brouard 5177:
1.126 brouard 5178: fprintf(ficrescveij,"%3.0f",age );
5179: for(i=1; i<=nlstate;i++)
5180: for(j=1; j<=nlstate;j++){
1.222 brouard 5181: cptj= (j-1)*nlstate+i;
5182: for(i2=1; i2<=nlstate;i2++)
5183: for(j2=1; j2<=nlstate;j2++){
5184: cptj2= (j2-1)*nlstate+i2;
5185: if(cptj2 <= cptj)
5186: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5187: }
1.126 brouard 5188: }
5189: fprintf(ficrescveij,"\n");
1.218 brouard 5190:
1.126 brouard 5191: }
5192: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5193: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5194: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5195: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5196: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5197: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5198: printf("\n");
5199: fprintf(ficlog,"\n");
1.218 brouard 5200:
1.126 brouard 5201: free_vector(xm,1,npar);
5202: free_vector(xp,1,npar);
5203: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5204: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5205: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5206: }
1.218 brouard 5207:
1.126 brouard 5208: /************ Variance ******************/
1.235 brouard 5209: 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 5210: {
5211: /* Variance of health expectancies */
5212: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5213: /* double **newm;*/
5214: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5215:
5216: /* int movingaverage(); */
5217: double **dnewm,**doldm;
5218: double **dnewmp,**doldmp;
5219: int i, j, nhstepm, hstepm, h, nstepm ;
5220: int k;
5221: double *xp;
5222: double **gp, **gm; /* for var eij */
5223: double ***gradg, ***trgradg; /*for var eij */
5224: double **gradgp, **trgradgp; /* for var p point j */
5225: double *gpp, *gmp; /* for var p point j */
5226: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5227: double ***p3mat;
5228: double age,agelim, hf;
5229: /* double ***mobaverage; */
5230: int theta;
5231: char digit[4];
5232: char digitp[25];
5233:
5234: char fileresprobmorprev[FILENAMELENGTH];
5235:
5236: if(popbased==1){
5237: if(mobilav!=0)
5238: strcpy(digitp,"-POPULBASED-MOBILAV_");
5239: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5240: }
5241: else
5242: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5243:
1.218 brouard 5244: /* if (mobilav!=0) { */
5245: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5246: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5247: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5248: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5249: /* } */
5250: /* } */
5251:
5252: strcpy(fileresprobmorprev,"PRMORPREV-");
5253: sprintf(digit,"%-d",ij);
5254: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5255: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5256: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5257: strcat(fileresprobmorprev,fileresu);
5258: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5259: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5260: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5261: }
5262: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5263: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5264: pstamp(ficresprobmorprev);
5265: 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);
5266: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5267: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5268: fprintf(ficresprobmorprev," p.%-d SE",j);
5269: for(i=1; i<=nlstate;i++)
5270: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5271: }
5272: fprintf(ficresprobmorprev,"\n");
5273:
5274: fprintf(ficgp,"\n# Routine varevsij");
5275: fprintf(ficgp,"\nunset title \n");
5276: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5277: 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");
5278: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5279: /* } */
5280: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5281: pstamp(ficresvij);
5282: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5283: if(popbased==1)
5284: 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);
5285: else
5286: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5287: fprintf(ficresvij,"# Age");
5288: for(i=1; i<=nlstate;i++)
5289: for(j=1; j<=nlstate;j++)
5290: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5291: fprintf(ficresvij,"\n");
5292:
5293: xp=vector(1,npar);
5294: dnewm=matrix(1,nlstate,1,npar);
5295: doldm=matrix(1,nlstate,1,nlstate);
5296: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5297: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5298:
5299: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5300: gpp=vector(nlstate+1,nlstate+ndeath);
5301: gmp=vector(nlstate+1,nlstate+ndeath);
5302: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5303:
1.218 brouard 5304: if(estepm < stepm){
5305: printf ("Problem %d lower than %d\n",estepm, stepm);
5306: }
5307: else hstepm=estepm;
5308: /* For example we decided to compute the life expectancy with the smallest unit */
5309: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5310: nhstepm is the number of hstepm from age to agelim
5311: nstepm is the number of stepm from age to agelim.
5312: Look at function hpijx to understand why because of memory size limitations,
5313: we decided (b) to get a life expectancy respecting the most precise curvature of the
5314: survival function given by stepm (the optimization length). Unfortunately it
5315: means that if the survival funtion is printed every two years of age and if
5316: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5317: results. So we changed our mind and took the option of the best precision.
5318: */
5319: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5320: agelim = AGESUP;
5321: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5322: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5323: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5324: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5325: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5326: gp=matrix(0,nhstepm,1,nlstate);
5327: gm=matrix(0,nhstepm,1,nlstate);
5328:
5329:
5330: for(theta=1; theta <=npar; theta++){
5331: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5332: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5333: }
5334:
1.235 brouard 5335: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nresult);
1.218 brouard 5336:
5337: if (popbased==1) {
5338: if(mobilav ==0){
5339: for(i=1; i<=nlstate;i++)
5340: prlim[i][i]=probs[(int)age][i][ij];
5341: }else{ /* mobilav */
5342: for(i=1; i<=nlstate;i++)
5343: prlim[i][i]=mobaverage[(int)age][i][ij];
5344: }
5345: }
5346:
1.235 brouard 5347: 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 5348: for(j=1; j<= nlstate; j++){
5349: for(h=0; h<=nhstepm; h++){
5350: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5351: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5352: }
5353: }
5354: /* Next for computing probability of death (h=1 means
5355: computed over hstepm matrices product = hstepm*stepm months)
5356: as a weighted average of prlim.
5357: */
5358: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5359: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5360: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5361: }
5362: /* end probability of death */
5363:
5364: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5365: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5366:
1.235 brouard 5367: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nresult);
1.218 brouard 5368:
5369: if (popbased==1) {
5370: if(mobilav ==0){
5371: for(i=1; i<=nlstate;i++)
5372: prlim[i][i]=probs[(int)age][i][ij];
5373: }else{ /* mobilav */
5374: for(i=1; i<=nlstate;i++)
5375: prlim[i][i]=mobaverage[(int)age][i][ij];
5376: }
5377: }
5378:
1.235 brouard 5379: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5380:
5381: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5382: for(h=0; h<=nhstepm; h++){
5383: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5384: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5385: }
5386: }
5387: /* This for computing probability of death (h=1 means
5388: computed over hstepm matrices product = hstepm*stepm months)
5389: as a weighted average of prlim.
5390: */
5391: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5392: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5393: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5394: }
5395: /* end probability of death */
5396:
5397: for(j=1; j<= nlstate; j++) /* vareij */
5398: for(h=0; h<=nhstepm; h++){
5399: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5400: }
5401:
5402: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5403: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5404: }
5405:
5406: } /* End theta */
5407:
5408: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5409:
5410: for(h=0; h<=nhstepm; h++) /* veij */
5411: for(j=1; j<=nlstate;j++)
5412: for(theta=1; theta <=npar; theta++)
5413: trgradg[h][j][theta]=gradg[h][theta][j];
5414:
5415: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5416: for(theta=1; theta <=npar; theta++)
5417: trgradgp[j][theta]=gradgp[theta][j];
5418:
5419:
5420: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5421: for(i=1;i<=nlstate;i++)
5422: for(j=1;j<=nlstate;j++)
5423: vareij[i][j][(int)age] =0.;
5424:
5425: for(h=0;h<=nhstepm;h++){
5426: for(k=0;k<=nhstepm;k++){
5427: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5428: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5429: for(i=1;i<=nlstate;i++)
5430: for(j=1;j<=nlstate;j++)
5431: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5432: }
5433: }
5434:
5435: /* pptj */
5436: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5437: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5438: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5439: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5440: varppt[j][i]=doldmp[j][i];
5441: /* end ppptj */
5442: /* x centered again */
5443:
1.235 brouard 5444: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nresult);
1.218 brouard 5445:
5446: if (popbased==1) {
5447: if(mobilav ==0){
5448: for(i=1; i<=nlstate;i++)
5449: prlim[i][i]=probs[(int)age][i][ij];
5450: }else{ /* mobilav */
5451: for(i=1; i<=nlstate;i++)
5452: prlim[i][i]=mobaverage[(int)age][i][ij];
5453: }
5454: }
5455:
5456: /* This for computing probability of death (h=1 means
5457: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5458: as a weighted average of prlim.
5459: */
1.235 brouard 5460: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5461: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5462: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5463: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5464: }
5465: /* end probability of death */
5466:
5467: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5468: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5469: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5470: for(i=1; i<=nlstate;i++){
5471: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5472: }
5473: }
5474: fprintf(ficresprobmorprev,"\n");
5475:
5476: fprintf(ficresvij,"%.0f ",age );
5477: for(i=1; i<=nlstate;i++)
5478: for(j=1; j<=nlstate;j++){
5479: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5480: }
5481: fprintf(ficresvij,"\n");
5482: free_matrix(gp,0,nhstepm,1,nlstate);
5483: free_matrix(gm,0,nhstepm,1,nlstate);
5484: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5485: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5486: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5487: } /* End age */
5488: free_vector(gpp,nlstate+1,nlstate+ndeath);
5489: free_vector(gmp,nlstate+1,nlstate+ndeath);
5490: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5491: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5492: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5493: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5494: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5495: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5496: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5497: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5498: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5499: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5500: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5501: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5502: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5503: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5504: 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);
5505: /* 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 5506: */
1.218 brouard 5507: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5508: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5509:
1.218 brouard 5510: free_vector(xp,1,npar);
5511: free_matrix(doldm,1,nlstate,1,nlstate);
5512: free_matrix(dnewm,1,nlstate,1,npar);
5513: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5514: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5515: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5516: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5517: fclose(ficresprobmorprev);
5518: fflush(ficgp);
5519: fflush(fichtm);
5520: } /* end varevsij */
1.126 brouard 5521:
5522: /************ Variance of prevlim ******************/
1.235 brouard 5523: 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 5524: {
1.205 brouard 5525: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5526: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5527:
1.126 brouard 5528: double **dnewm,**doldm;
5529: int i, j, nhstepm, hstepm;
5530: double *xp;
5531: double *gp, *gm;
5532: double **gradg, **trgradg;
1.208 brouard 5533: double **mgm, **mgp;
1.126 brouard 5534: double age,agelim;
5535: int theta;
5536:
5537: pstamp(ficresvpl);
5538: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
5539: fprintf(ficresvpl,"# Age");
5540: for(i=1; i<=nlstate;i++)
5541: fprintf(ficresvpl," %1d-%1d",i,i);
5542: fprintf(ficresvpl,"\n");
5543:
5544: xp=vector(1,npar);
5545: dnewm=matrix(1,nlstate,1,npar);
5546: doldm=matrix(1,nlstate,1,nlstate);
5547:
5548: hstepm=1*YEARM; /* Every year of age */
5549: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5550: agelim = AGESUP;
5551: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5552: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5553: if (stepm >= YEARM) hstepm=1;
5554: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5555: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5556: mgp=matrix(1,npar,1,nlstate);
5557: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5558: gp=vector(1,nlstate);
5559: gm=vector(1,nlstate);
5560:
5561: for(theta=1; theta <=npar; theta++){
5562: for(i=1; i<=npar; i++){ /* Computes gradient */
5563: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5564: }
1.209 brouard 5565: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5566: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5567: else
1.235 brouard 5568: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5569: for(i=1;i<=nlstate;i++){
1.126 brouard 5570: gp[i] = prlim[i][i];
1.208 brouard 5571: mgp[theta][i] = prlim[i][i];
5572: }
1.126 brouard 5573: for(i=1; i<=npar; i++) /* Computes gradient */
5574: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5575: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5576: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5577: else
1.235 brouard 5578: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5579: for(i=1;i<=nlstate;i++){
1.126 brouard 5580: gm[i] = prlim[i][i];
1.208 brouard 5581: mgm[theta][i] = prlim[i][i];
5582: }
1.126 brouard 5583: for(i=1;i<=nlstate;i++)
5584: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5585: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5586: } /* End theta */
5587:
5588: trgradg =matrix(1,nlstate,1,npar);
5589:
5590: for(j=1; j<=nlstate;j++)
5591: for(theta=1; theta <=npar; theta++)
5592: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 5593: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5594: /* printf("\nmgm mgp %d ",(int)age); */
5595: /* for(j=1; j<=nlstate;j++){ */
5596: /* printf(" %d ",j); */
5597: /* for(theta=1; theta <=npar; theta++) */
5598: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
5599: /* printf("\n "); */
5600: /* } */
5601: /* } */
5602: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5603: /* printf("\n gradg %d ",(int)age); */
5604: /* for(j=1; j<=nlstate;j++){ */
5605: /* printf("%d ",j); */
5606: /* for(theta=1; theta <=npar; theta++) */
5607: /* printf("%d %lf ",theta,gradg[theta][j]); */
5608: /* printf("\n "); */
5609: /* } */
5610: /* } */
1.126 brouard 5611:
5612: for(i=1;i<=nlstate;i++)
5613: varpl[i][(int)age] =0.;
1.209 brouard 5614: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 5615: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5616: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
5617: }else{
1.126 brouard 5618: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5619: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 5620: }
1.126 brouard 5621: for(i=1;i<=nlstate;i++)
5622: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
5623:
5624: fprintf(ficresvpl,"%.0f ",age );
5625: for(i=1; i<=nlstate;i++)
5626: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
5627: fprintf(ficresvpl,"\n");
5628: free_vector(gp,1,nlstate);
5629: free_vector(gm,1,nlstate);
1.208 brouard 5630: free_matrix(mgm,1,npar,1,nlstate);
5631: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 5632: free_matrix(gradg,1,npar,1,nlstate);
5633: free_matrix(trgradg,1,nlstate,1,npar);
5634: } /* End age */
5635:
5636: free_vector(xp,1,npar);
5637: free_matrix(doldm,1,nlstate,1,npar);
5638: free_matrix(dnewm,1,nlstate,1,nlstate);
5639:
5640: }
5641:
5642: /************ Variance of one-step probabilities ******************/
5643: 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 5644: {
5645: int i, j=0, k1, l1, tj;
5646: int k2, l2, j1, z1;
5647: int k=0, l;
5648: int first=1, first1, first2;
5649: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
5650: double **dnewm,**doldm;
5651: double *xp;
5652: double *gp, *gm;
5653: double **gradg, **trgradg;
5654: double **mu;
5655: double age, cov[NCOVMAX+1];
5656: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
5657: int theta;
5658: char fileresprob[FILENAMELENGTH];
5659: char fileresprobcov[FILENAMELENGTH];
5660: char fileresprobcor[FILENAMELENGTH];
5661: double ***varpij;
5662:
5663: strcpy(fileresprob,"PROB_");
5664: strcat(fileresprob,fileres);
5665: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
5666: printf("Problem with resultfile: %s\n", fileresprob);
5667: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
5668: }
5669: strcpy(fileresprobcov,"PROBCOV_");
5670: strcat(fileresprobcov,fileresu);
5671: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
5672: printf("Problem with resultfile: %s\n", fileresprobcov);
5673: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
5674: }
5675: strcpy(fileresprobcor,"PROBCOR_");
5676: strcat(fileresprobcor,fileresu);
5677: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
5678: printf("Problem with resultfile: %s\n", fileresprobcor);
5679: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
5680: }
5681: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5682: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5683: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5684: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5685: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5686: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5687: pstamp(ficresprob);
5688: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
5689: fprintf(ficresprob,"# Age");
5690: pstamp(ficresprobcov);
5691: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
5692: fprintf(ficresprobcov,"# Age");
5693: pstamp(ficresprobcor);
5694: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
5695: fprintf(ficresprobcor,"# Age");
1.126 brouard 5696:
5697:
1.222 brouard 5698: for(i=1; i<=nlstate;i++)
5699: for(j=1; j<=(nlstate+ndeath);j++){
5700: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
5701: fprintf(ficresprobcov," p%1d-%1d ",i,j);
5702: fprintf(ficresprobcor," p%1d-%1d ",i,j);
5703: }
5704: /* fprintf(ficresprob,"\n");
5705: fprintf(ficresprobcov,"\n");
5706: fprintf(ficresprobcor,"\n");
5707: */
5708: xp=vector(1,npar);
5709: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5710: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5711: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
5712: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
5713: first=1;
5714: fprintf(ficgp,"\n# Routine varprob");
5715: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
5716: fprintf(fichtm,"\n");
5717:
5718: 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);
5719: 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);
5720: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 5721: and drawn. It helps understanding how is the covariance between two incidences.\
5722: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 5723: 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 5724: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
5725: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
5726: standard deviations wide on each axis. <br>\
5727: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
5728: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
5729: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
5730:
1.222 brouard 5731: cov[1]=1;
5732: /* tj=cptcoveff; */
1.225 brouard 5733: tj = (int) pow(2,cptcoveff);
1.222 brouard 5734: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
5735: j1=0;
1.224 brouard 5736: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 5737: if (cptcovn>0) {
5738: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 5739: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5740: fprintf(ficresprob, "**********\n#\n");
5741: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 5742: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5743: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 5744:
1.222 brouard 5745: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 5746: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5747: fprintf(ficgp, "**********\n#\n");
1.220 brouard 5748:
5749:
1.222 brouard 5750: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 5751: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5752: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 5753:
1.222 brouard 5754: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 5755: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5756: fprintf(ficresprobcor, "**********\n#");
5757: if(invalidvarcomb[j1]){
5758: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
5759: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
5760: continue;
5761: }
5762: }
5763: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
5764: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5765: gp=vector(1,(nlstate)*(nlstate+ndeath));
5766: gm=vector(1,(nlstate)*(nlstate+ndeath));
5767: for (age=bage; age<=fage; age ++){
5768: cov[2]=age;
5769: if(nagesqr==1)
5770: cov[3]= age*age;
5771: for (k=1; k<=cptcovn;k++) {
5772: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
5773: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
5774: * 1 1 1 1 1
5775: * 2 2 1 1 1
5776: * 3 1 2 1 1
5777: */
5778: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
5779: }
5780: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
5781: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
5782: for (k=1; k<=cptcovprod;k++)
5783: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 5784:
5785:
1.222 brouard 5786: for(theta=1; theta <=npar; theta++){
5787: for(i=1; i<=npar; i++)
5788: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 5789:
1.222 brouard 5790: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 5791:
1.222 brouard 5792: k=0;
5793: for(i=1; i<= (nlstate); i++){
5794: for(j=1; j<=(nlstate+ndeath);j++){
5795: k=k+1;
5796: gp[k]=pmmij[i][j];
5797: }
5798: }
1.220 brouard 5799:
1.222 brouard 5800: for(i=1; i<=npar; i++)
5801: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 5802:
1.222 brouard 5803: pmij(pmmij,cov,ncovmodel,xp,nlstate);
5804: k=0;
5805: for(i=1; i<=(nlstate); i++){
5806: for(j=1; j<=(nlstate+ndeath);j++){
5807: k=k+1;
5808: gm[k]=pmmij[i][j];
5809: }
5810: }
1.220 brouard 5811:
1.222 brouard 5812: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
5813: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
5814: }
1.126 brouard 5815:
1.222 brouard 5816: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
5817: for(theta=1; theta <=npar; theta++)
5818: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 5819:
1.222 brouard 5820: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
5821: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 5822:
1.222 brouard 5823: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 5824:
1.222 brouard 5825: k=0;
5826: for(i=1; i<=(nlstate); i++){
5827: for(j=1; j<=(nlstate+ndeath);j++){
5828: k=k+1;
5829: mu[k][(int) age]=pmmij[i][j];
5830: }
5831: }
5832: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
5833: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
5834: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 5835:
1.222 brouard 5836: /*printf("\n%d ",(int)age);
5837: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
5838: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5839: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5840: }*/
1.220 brouard 5841:
1.222 brouard 5842: fprintf(ficresprob,"\n%d ",(int)age);
5843: fprintf(ficresprobcov,"\n%d ",(int)age);
5844: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 5845:
1.222 brouard 5846: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
5847: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
5848: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
5849: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
5850: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
5851: }
5852: i=0;
5853: for (k=1; k<=(nlstate);k++){
5854: for (l=1; l<=(nlstate+ndeath);l++){
5855: i++;
5856: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
5857: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
5858: for (j=1; j<=i;j++){
5859: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
5860: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
5861: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
5862: }
5863: }
5864: }/* end of loop for state */
5865: } /* end of loop for age */
5866: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
5867: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
5868: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
5869: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
5870:
5871: /* Confidence intervalle of pij */
5872: /*
5873: fprintf(ficgp,"\nunset parametric;unset label");
5874: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
5875: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
5876: 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);
5877: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
5878: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
5879: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
5880: */
5881:
5882: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
5883: first1=1;first2=2;
5884: for (k2=1; k2<=(nlstate);k2++){
5885: for (l2=1; l2<=(nlstate+ndeath);l2++){
5886: if(l2==k2) continue;
5887: j=(k2-1)*(nlstate+ndeath)+l2;
5888: for (k1=1; k1<=(nlstate);k1++){
5889: for (l1=1; l1<=(nlstate+ndeath);l1++){
5890: if(l1==k1) continue;
5891: i=(k1-1)*(nlstate+ndeath)+l1;
5892: if(i<=j) continue;
5893: for (age=bage; age<=fage; age ++){
5894: if ((int)age %5==0){
5895: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
5896: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
5897: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
5898: mu1=mu[i][(int) age]/stepm*YEARM ;
5899: mu2=mu[j][(int) age]/stepm*YEARM;
5900: c12=cv12/sqrt(v1*v2);
5901: /* Computing eigen value of matrix of covariance */
5902: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
5903: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
5904: if ((lc2 <0) || (lc1 <0) ){
5905: if(first2==1){
5906: first1=0;
5907: 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);
5908: }
5909: 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);
5910: /* lc1=fabs(lc1); */ /* If we want to have them positive */
5911: /* lc2=fabs(lc2); */
5912: }
1.220 brouard 5913:
1.222 brouard 5914: /* Eigen vectors */
5915: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
5916: /*v21=sqrt(1.-v11*v11); *//* error */
5917: v21=(lc1-v1)/cv12*v11;
5918: v12=-v21;
5919: v22=v11;
5920: tnalp=v21/v11;
5921: if(first1==1){
5922: first1=0;
5923: 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);
5924: }
5925: 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);
5926: /*printf(fignu*/
5927: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
5928: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
5929: if(first==1){
5930: first=0;
5931: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
5932: fprintf(ficgp,"\nset parametric;unset label");
5933: 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);
5934: fprintf(ficgp,"\nset ter svg size 640, 480");
5935: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 5936: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 5937: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 5938: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
5939: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5940: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5941: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
5942: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5943: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
5944: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
5945: 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", \
5946: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
5947: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
5948: }else{
5949: first=0;
5950: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
5951: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
5952: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
5953: 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", \
5954: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
5955: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
5956: }/* if first */
5957: } /* age mod 5 */
5958: } /* end loop age */
5959: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5960: first=1;
5961: } /*l12 */
5962: } /* k12 */
5963: } /*l1 */
5964: }/* k1 */
5965: } /* loop on combination of covariates j1 */
5966: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
5967: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
5968: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5969: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
5970: free_vector(xp,1,npar);
5971: fclose(ficresprob);
5972: fclose(ficresprobcov);
5973: fclose(ficresprobcor);
5974: fflush(ficgp);
5975: fflush(fichtmcov);
5976: }
1.126 brouard 5977:
5978:
5979: /******************* Printing html file ***********/
1.201 brouard 5980: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 5981: int lastpass, int stepm, int weightopt, char model[],\
5982: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.217 brouard 5983: int popforecast, int prevfcast, int backcast, int estepm , \
1.213 brouard 5984: double jprev1, double mprev1,double anprev1, double dateprev1, \
5985: double jprev2, double mprev2,double anprev2, double dateprev2){
1.126 brouard 5986: int jj1, k1, i1, cpt;
5987:
5988: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
5989: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
5990: </ul>");
1.214 brouard 5991: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
5992: 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",
5993: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
5994: 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 5995: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
5996: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 5997: fprintf(fichtm,"\
5998: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 5999: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6000: fprintf(fichtm,"\
1.217 brouard 6001: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6002: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6003: fprintf(fichtm,"\
1.126 brouard 6004: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6005: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6006: fprintf(fichtm,"\
1.217 brouard 6007: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6008: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6009: fprintf(fichtm,"\
1.211 brouard 6010: - (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 6011: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6012: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6013: if(prevfcast==1){
6014: fprintf(fichtm,"\
6015: - Prevalence projections by age and states: \
1.201 brouard 6016: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6017: }
1.126 brouard 6018:
1.222 brouard 6019: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 6020:
1.225 brouard 6021: m=pow(2,cptcoveff);
1.222 brouard 6022: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6023:
1.222 brouard 6024: jj1=0;
6025: for(k1=1; k1<=m;k1++){
1.220 brouard 6026:
1.222 brouard 6027: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6028: jj1++;
6029: if (cptcovn > 0) {
6030: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6031: for (cpt=1; cpt<=cptcoveff;cpt++){
1.222 brouard 6032: fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);
6033: printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout);
6034: }
1.230 brouard 6035: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6036: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6037: if(invalidvarcomb[k1]){
6038: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6039: printf("\nCombination (%d) ignored because no cases \n",k1);
6040: continue;
6041: }
6042: }
6043: /* aij, bij */
6044: 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.svg\">%s_%d-1.svg</a><br> \
1.211 brouard 6045: <img src=\"%s_%d-1.svg\">",model,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 6046: /* Pij */
6047: 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.svg\">%s_%d-2.svg</a><br> \
1.201 brouard 6048: <img src=\"%s_%d-2.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 6049: /* Quasi-incidences */
6050: 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 6051: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6052: incidence (rates) are the limit when h tends to zero of the ratio of the probability <sub>h</sub>P<sub>ij</sub> \
6053: divided by h: <sub>h</sub>P<sub>ij</sub>/h : <a href=\"%s_%d-3.svg\">%s_%d-3.svg</a><br> \
1.201 brouard 6054: <img src=\"%s_%d-3.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 6055: /* Survival functions (period) in state j */
6056: for(cpt=1; cpt<=nlstate;cpt++){
6057: 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.svg\">%s%d_%d.svg</a><br> \
1.201 brouard 6058: <img src=\"%s_%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"LIJ_"),cpt,jj1,subdirf2(optionfilefiname,"LIJ_"),cpt,jj1,subdirf2(optionfilefiname,"LIJ_"),cpt,jj1);
1.222 brouard 6059: }
6060: /* State specific survival functions (period) */
6061: for(cpt=1; cpt<=nlstate;cpt++){
6062: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6063: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.201 brouard 6064: <a href=\"%s%d_%d.svg\">%s%d_%d.svg</a><br> <img src=\"%s_%d-%d.svg\">", cpt, nlstate, cpt, subdirf2(optionfilefiname,"LIJT_"),cpt,jj1,subdirf2(optionfilefiname,"LIJT_"),cpt,jj1,subdirf2(optionfilefiname,"LIJT_"),cpt,jj1);
1.222 brouard 6065: }
6066: /* Period (stable) prevalence in each health state */
6067: for(cpt=1; cpt<=nlstate;cpt++){
6068: 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.svg\">%s_%d-%d.svg</a><br> \
1.201 brouard 6069: <img src=\"%s_%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"P_"),cpt,jj1,subdirf2(optionfilefiname,"P_"),cpt,jj1,subdirf2(optionfilefiname,"P_"),cpt,jj1);
1.222 brouard 6070: }
6071: if(backcast==1){
6072: /* Period (stable) back prevalence in each health state */
6073: for(cpt=1; cpt<=nlstate;cpt++){
6074: 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.svg\">%s_%d-%d.svg</a><br> \
1.217 brouard 6075: <img src=\"%s_%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"PB_"),cpt,jj1,subdirf2(optionfilefiname,"PB_"),cpt,jj1,subdirf2(optionfilefiname,"PB_"),cpt,jj1);
1.222 brouard 6076: }
1.217 brouard 6077: }
1.222 brouard 6078: if(prevfcast==1){
6079: /* Projection of prevalence up to period (stable) prevalence in each health state */
6080: for(cpt=1; cpt<=nlstate;cpt++){
6081: 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.svg\">%s%d_%d.svg</a><br> \
1.213 brouard 6082: <img src=\"%s_%d-%d.svg\">", dateprev1, dateprev2, cpt, cpt, nlstate, subdirf2(optionfilefiname,"PROJ_"),cpt,jj1,subdirf2(optionfilefiname,"PROJ_"),cpt,jj1,subdirf2(optionfilefiname,"PROJ_"),cpt,jj1);
1.222 brouard 6083: }
6084: }
1.220 brouard 6085:
1.222 brouard 6086: for(cpt=1; cpt<=nlstate;cpt++) {
6087: 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.svg\">%s_%d%d.svg</a> <br> \
1.201 brouard 6088: <img src=\"%s_%d%d.svg\">",cpt,nlstate,subdirf2(optionfilefiname,"EXP_"),cpt,jj1,subdirf2(optionfilefiname,"EXP_"),cpt,jj1,subdirf2(optionfilefiname,"EXP_"),cpt,jj1);
1.222 brouard 6089: }
6090: /* } /\* end i1 *\/ */
6091: }/* End k1 */
6092: fprintf(fichtm,"</ul>");
1.126 brouard 6093:
1.222 brouard 6094: fprintf(fichtm,"\
1.126 brouard 6095: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6096: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6097: - 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 6098: But because parameters are usually highly correlated (a higher incidence of disability \
6099: and a higher incidence of recovery can give very close observed transition) it might \
6100: be very useful to look not only at linear confidence intervals estimated from the \
6101: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6102: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6103: covariance matrix of the one-step probabilities. \
6104: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6105:
1.222 brouard 6106: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6107: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6108: fprintf(fichtm,"\
1.126 brouard 6109: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6110: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6111:
1.222 brouard 6112: fprintf(fichtm,"\
1.126 brouard 6113: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6114: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6115: fprintf(fichtm,"\
1.126 brouard 6116: - 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): \
6117: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6118: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6119: fprintf(fichtm,"\
1.126 brouard 6120: - (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): \
6121: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6122: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6123: fprintf(fichtm,"\
1.128 brouard 6124: - 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 6125: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6126: fprintf(fichtm,"\
1.128 brouard 6127: - 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 6128: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6129: fprintf(fichtm,"\
1.126 brouard 6130: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6131: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6132:
6133: /* if(popforecast==1) fprintf(fichtm,"\n */
6134: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6135: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6136: /* <br>",fileres,fileres,fileres,fileres); */
6137: /* else */
6138: /* 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 6139: fflush(fichtm);
6140: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6141:
1.225 brouard 6142: m=pow(2,cptcoveff);
1.222 brouard 6143: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6144:
1.222 brouard 6145: jj1=0;
6146: for(k1=1; k1<=m;k1++){
6147: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6148: jj1++;
1.126 brouard 6149: if (cptcovn > 0) {
6150: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6151: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.222 brouard 6152: fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);
1.126 brouard 6153: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6154:
1.222 brouard 6155: if(invalidvarcomb[k1]){
6156: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6157: continue;
6158: }
1.126 brouard 6159: }
6160: for(cpt=1; cpt<=nlstate;cpt++) {
1.218 brouard 6161: fprintf(fichtm,"\n<br>- Observed (cross-sectional) and period (incidence based) \
6162: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d.svg\"> %s_%d-%d.svg</a>\n <br>\
1.205 brouard 6163: <img src=\"%s_%d-%d.svg\">",cpt,subdirf2(optionfilefiname,"V_"),cpt,jj1,subdirf2(optionfilefiname,"V_"),cpt,jj1,subdirf2(optionfilefiname,"V_"),cpt,jj1);
1.126 brouard 6164: }
6165: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6166: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6167: true period expectancies (those weighted with period prevalences are also\
6168: drawn in addition to the population based expectancies computed using\
1.218 brouard 6169: observed and cahotic prevalences: <a href=\"%s_%d.svg\">%s_%d.svg</a>\n<br>\
1.205 brouard 6170: <img src=\"%s_%d.svg\">",subdirf2(optionfilefiname,"E_"),jj1,subdirf2(optionfilefiname,"E_"),jj1,subdirf2(optionfilefiname,"E_"),jj1);
1.222 brouard 6171: /* } /\* end i1 *\/ */
6172: }/* End k1 */
6173: fprintf(fichtm,"</ul>");
6174: fflush(fichtm);
1.126 brouard 6175: }
6176:
6177: /******************* Gnuplot file **************/
1.223 brouard 6178: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6179:
6180: char dirfileres[132],optfileres[132];
1.223 brouard 6181: char gplotcondition[132];
1.235 brouard 6182: int cpt=0,k1=0,i=0,k=0,j=0,jk=0,k2=0,k3=0,k4=0,ij=0,l=0;
1.211 brouard 6183: int lv=0, vlv=0, kl=0;
1.130 brouard 6184: int ng=0;
1.201 brouard 6185: int vpopbased;
1.223 brouard 6186: int ioffset; /* variable offset for columns */
1.235 brouard 6187: int nres=0; /* Index of resultline */
1.219 brouard 6188:
1.126 brouard 6189: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6190: /* printf("Problem with file %s",optionfilegnuplot); */
6191: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6192: /* } */
6193:
6194: /*#ifdef windows */
6195: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6196: /*#endif */
1.225 brouard 6197: m=pow(2,cptcoveff);
1.126 brouard 6198:
1.202 brouard 6199: /* Contribution to likelihood */
6200: /* Plot the probability implied in the likelihood */
1.223 brouard 6201: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6202: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6203: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6204: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6205: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6206: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6207: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6208: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6209: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6210: 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));
6211: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6212: 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));
6213: for (i=1; i<= nlstate ; i ++) {
6214: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6215: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6216: 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);
6217: for (j=2; j<= nlstate+ndeath ; j ++) {
6218: 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);
6219: }
6220: fprintf(ficgp,";\nset out; unset ylabel;\n");
6221: }
6222: /* 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 */
6223: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6224: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6225: fprintf(ficgp,"\nset out;unset log\n");
6226: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6227:
1.126 brouard 6228: strcpy(dirfileres,optionfilefiname);
6229: strcpy(optfileres,"vpl");
1.223 brouard 6230: /* 1eme*/
1.211 brouard 6231: for (cpt=1; cpt<= nlstate ; cpt ++) { /* For each live state */
1.236 ! brouard 6232: for (k1=1; k1<= m ; k1 ++) /* For each valid combination of covariate */
! 6233: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.235 brouard 6234: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
6235: if(TKresult[nres]!= k1)
6236: continue;
6237: /* We are interested in selected combination by the resultline */
6238: printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
6239: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
1.225 brouard 6240: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6241: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
1.223 brouard 6242: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6243: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6244: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6245: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6246: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.235 brouard 6247: printf(" V%d=%d ",Tvaraff[k],vlv);
1.223 brouard 6248: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6249: }
1.235 brouard 6250: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6251: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6252: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6253: }
6254: printf("\n#\n");
1.211 brouard 6255: fprintf(ficgp,"\n#\n");
1.223 brouard 6256: if(invalidvarcomb[k1]){
6257: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6258: continue;
6259: }
1.235 brouard 6260:
1.223 brouard 6261: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1);
6262: fprintf(ficgp,"\n#set out \"V_%s_%d-%d.svg\" \n",optionfilefiname,cpt,k1);
6263: fprintf(ficgp,"set xlabel \"Age\" \n\
1.235 brouard 6264: set ylabel \"Probability\" \n \
6265: set ter svg size 640, 480\n \
1.201 brouard 6266: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:2 \"%%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),k1-1,k1-1);
1.235 brouard 6267:
1.223 brouard 6268: for (i=1; i<= nlstate ; i ++) {
6269: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6270: else fprintf(ficgp," %%*lf (%%*lf)");
6271: }
6272: fprintf(ficgp,"\" t\"Period (stable) prevalence\" w l lt 0,\"%s\" every :::%d::%d u 1:($2+1.96*$3) \"%%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1);
6273: for (i=1; i<= nlstate ; i ++) {
6274: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6275: else fprintf(ficgp," %%*lf (%%*lf)");
6276: }
6277: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2-1.96*$3) \"%%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1);
6278: for (i=1; i<= nlstate ; i ++) {
6279: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6280: else fprintf(ficgp," %%*lf (%%*lf)");
6281: }
6282: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6283: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6284: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
6285: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1 */
1.225 brouard 6286: if(cptcoveff ==0){
1.223 brouard 6287: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line ", 2+(cpt-1), cpt );
6288: }else{
6289: kl=0;
1.225 brouard 6290: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6291: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
1.223 brouard 6292: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6293: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6294: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6295: vlv= nbcode[Tvaraff[k]][lv];
6296: kl++;
6297: /* 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 *\/ */
6298: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6299: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6300: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
1.225 brouard 6301: if(k==cptcoveff){
1.227 brouard 6302: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
6303: 4+(cpt-1), cpt ); /* 4 or 6 ?*/
1.223 brouard 6304: }else{
6305: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6306: kl++;
6307: }
6308: } /* end covariate */
6309: } /* end if no covariate */
6310: } /* end if backcast */
6311: fprintf(ficgp,"\nset out \n");
1.201 brouard 6312: } /* k1 */
6313: } /* cpt */
1.235 brouard 6314:
6315:
1.126 brouard 6316: /*2 eme*/
1.236 ! brouard 6317: for (k1=1; k1<= m ; k1 ++)
! 6318: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
! 6319: if(TKresult[nres]!= k1)
! 6320: continue;
1.223 brouard 6321: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
1.225 brouard 6322: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6323: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6324: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6325: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6326: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6327: vlv= nbcode[Tvaraff[k]][lv];
6328: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.236 ! brouard 6329: }
! 6330: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
! 6331: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
! 6332: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.223 brouard 6333: }
6334: fprintf(ficgp,"\n#\n");
6335: if(invalidvarcomb[k1]){
6336: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6337: continue;
6338: }
1.219 brouard 6339:
1.223 brouard 6340: fprintf(ficgp,"\nset out \"%s_%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1);
6341: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6342: if(vpopbased==0)
6343: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6344: else
6345: fprintf(ficgp,"\nreplot ");
6346: for (i=1; i<= nlstate+1 ; i ++) {
6347: k=2*i;
6348: 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);
6349: for (j=1; j<= nlstate+1 ; j ++) {
6350: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6351: else fprintf(ficgp," %%*lf (%%*lf)");
6352: }
6353: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6354: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
6355: 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);
6356: for (j=1; j<= nlstate+1 ; j ++) {
6357: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6358: else fprintf(ficgp," %%*lf (%%*lf)");
6359: }
6360: fprintf(ficgp,"\" t\"\" w l lt 0,");
6361: 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);
6362: for (j=1; j<= nlstate+1 ; j ++) {
6363: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6364: else fprintf(ficgp," %%*lf (%%*lf)");
6365: }
6366: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6367: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6368: } /* state */
6369: } /* vpopbased */
6370: fprintf(ficgp,"\nset out;set out \"%s_%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1); /* Buggy gnuplot */
1.235 brouard 6371: } /* k1 end 2 eme*/
1.219 brouard 6372:
6373:
1.126 brouard 6374: /*3eme*/
1.236 ! brouard 6375: for (k1=1; k1<= m ; k1 ++)
! 6376: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
! 6377: if(TKresult[nres]!= k)
! 6378: continue;
1.220 brouard 6379:
1.126 brouard 6380: for (cpt=1; cpt<= nlstate ; cpt ++) {
1.236 ! brouard 6381: fprintf(ficgp,"\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
1.225 brouard 6382: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6383: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6384: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6385: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6386: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6387: vlv= nbcode[Tvaraff[k]][lv];
6388: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6389: }
1.236 ! brouard 6390: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
! 6391: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
! 6392: }
1.211 brouard 6393: fprintf(ficgp,"\n#\n");
1.223 brouard 6394: if(invalidvarcomb[k1]){
6395: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6396: continue;
6397: }
1.219 brouard 6398:
1.126 brouard 6399: /* k=2+nlstate*(2*cpt-2); */
6400: k=2+(nlstate+1)*(cpt-1);
1.201 brouard 6401: fprintf(ficgp,"\nset out \"%s_%d%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1);
1.199 brouard 6402: fprintf(ficgp,"set ter svg size 640, 480\n\
1.201 brouard 6403: 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);
1.126 brouard 6404: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
1.223 brouard 6405: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6406: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6407: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6408: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6409: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6410:
1.126 brouard 6411: */
6412: for (i=1; i< nlstate ; i ++) {
1.223 brouard 6413: 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);
6414: /* 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 6415:
1.126 brouard 6416: }
1.201 brouard 6417: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+nlstate,cpt);
1.126 brouard 6418: }
6419: }
6420:
1.223 brouard 6421: /* 4eme */
1.201 brouard 6422: /* Survival functions (period) from state i in state j by initial state i */
1.236 ! brouard 6423: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
! 6424: for(nres=1; nres <= nresult; nres++) /* For each resultline */
! 6425: if(TKresult[nres]!= k)
! 6426: continue;
1.220 brouard 6427:
1.201 brouard 6428: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.211 brouard 6429: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
1.225 brouard 6430: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6431: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6432: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6433: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6434: vlv= nbcode[Tvaraff[k]][lv];
6435: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6436: }
1.236 ! brouard 6437: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
! 6438: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
! 6439: }
1.211 brouard 6440: fprintf(ficgp,"\n#\n");
1.223 brouard 6441: if(invalidvarcomb[k1]){
6442: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6443: continue;
6444: }
1.220 brouard 6445:
1.201 brouard 6446: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1);
6447: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
1.220 brouard 6448: set ter svg size 640, 480\n \
6449: unset log y\n \
1.201 brouard 6450: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6451: k=3;
1.201 brouard 6452: for (i=1; i<= nlstate ; i ++){
1.223 brouard 6453: if(i==1){
6454: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6455: }else{
6456: fprintf(ficgp,", '' ");
6457: }
6458: l=(nlstate+ndeath)*(i-1)+1;
6459: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6460: for (j=2; j<= nlstate+ndeath ; j ++)
6461: fprintf(ficgp,"+$%d",k+l+j-1);
6462: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
1.201 brouard 6463: } /* nlstate */
6464: fprintf(ficgp,"\nset out\n");
6465: } /* end cpt state*/
6466: } /* end covariate */
1.220 brouard 6467:
6468: /* 5eme */
1.201 brouard 6469: /* Survival functions (period) from state i in state j by final state j */
1.236 ! brouard 6470: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
! 6471: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
! 6472: if(TKresult[nres]!= k1)
! 6473: continue;
1.201 brouard 6474: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
1.223 brouard 6475:
1.201 brouard 6476: 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);
1.225 brouard 6477: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6478: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6479: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6480: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6481: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6482: vlv= nbcode[Tvaraff[k]][lv];
6483: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6484: }
1.236 ! brouard 6485: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
! 6486: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
! 6487: }
1.211 brouard 6488: fprintf(ficgp,"\n#\n");
1.223 brouard 6489: if(invalidvarcomb[k1]){
1.227 brouard 6490: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6491: continue;
1.223 brouard 6492: }
1.227 brouard 6493:
1.201 brouard 6494: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1);
6495: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
1.227 brouard 6496: set ter svg size 640, 480\n \
6497: unset log y\n \
1.201 brouard 6498: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6499: k=3;
1.201 brouard 6500: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
1.227 brouard 6501: if(j==1)
6502: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6503: else
6504: fprintf(ficgp,", '' ");
6505: l=(nlstate+ndeath)*(cpt-1) +j;
6506: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6507: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6508: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6509: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
1.201 brouard 6510: } /* nlstate */
6511: fprintf(ficgp,", '' ");
6512: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6513: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
1.227 brouard 6514: l=(nlstate+ndeath)*(cpt-1) +j;
6515: if(j < nlstate)
6516: fprintf(ficgp,"$%d +",k+l);
6517: else
6518: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
1.201 brouard 6519: }
6520: fprintf(ficgp,"\nset out\n");
6521: } /* end cpt state*/
6522: } /* end covariate */
1.227 brouard 6523:
1.220 brouard 6524: /* 6eme */
1.202 brouard 6525: /* CV preval stable (period) for each covariate */
1.211 brouard 6526: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
1.153 brouard 6527: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6528:
1.211 brouard 6529: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6530: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6531: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6532: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6533: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6534: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6535: vlv= nbcode[Tvaraff[k]][lv];
6536: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6537: }
6538: fprintf(ficgp,"\n#\n");
1.223 brouard 6539: if(invalidvarcomb[k1]){
1.227 brouard 6540: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6541: continue;
1.223 brouard 6542: }
1.227 brouard 6543:
1.201 brouard 6544: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1);
1.126 brouard 6545: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.227 brouard 6546: set ter svg size 640, 480\n \
6547: unset log y\n \
1.153 brouard 6548: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6549: k=3; /* Offset */
1.153 brouard 6550: for (i=1; i<= nlstate ; i ++){
1.227 brouard 6551: if(i==1)
6552: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6553: else
6554: fprintf(ficgp,", '' ");
6555: l=(nlstate+ndeath)*(i-1)+1;
6556: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6557: for (j=2; j<= nlstate ; j ++)
6558: fprintf(ficgp,"+$%d",k+l+j-1);
6559: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6560: } /* nlstate */
1.201 brouard 6561: fprintf(ficgp,"\nset out\n");
1.153 brouard 6562: } /* end cpt state*/
6563: } /* end covariate */
1.227 brouard 6564:
6565:
1.220 brouard 6566: /* 7eme */
1.218 brouard 6567: if(backcast == 1){
1.217 brouard 6568: /* CV back preval stable (period) for each covariate */
1.218 brouard 6569: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
6570: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6571: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
6572: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6573: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6574: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6575: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 6576: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 6577: vlv= nbcode[Tvaraff[k]][lv];
6578: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6579: }
6580: fprintf(ficgp,"\n#\n");
6581: if(invalidvarcomb[k1]){
6582: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6583: continue;
6584: }
6585:
6586: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1);
6587: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
6588: set ter svg size 640, 480\n \
6589: unset log y\n \
1.218 brouard 6590: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6591: k=3; /* Offset */
6592: for (i=1; i<= nlstate ; i ++){
6593: if(i==1)
6594: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
6595: else
6596: fprintf(ficgp,", '' ");
6597: /* l=(nlstate+ndeath)*(i-1)+1; */
6598: l=(nlstate+ndeath)*(cpt-1)+1;
6599: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
6600: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
6601: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+(cpt-1)+i-1); /* a vérifier */
6602: /* for (j=2; j<= nlstate ; j ++) */
6603: /* fprintf(ficgp,"+$%d",k+l+j-1); */
6604: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
6605: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
6606: } /* nlstate */
6607: fprintf(ficgp,"\nset out\n");
1.218 brouard 6608: } /* end cpt state*/
6609: } /* end covariate */
6610: } /* End if backcast */
6611:
1.223 brouard 6612: /* 8eme */
1.218 brouard 6613: if(prevfcast==1){
6614: /* Projection from cross-sectional to stable (period) for each covariate */
6615:
6616: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
1.211 brouard 6617: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6618: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
6619: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
6620: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6621: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6622: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6623: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6624: vlv= nbcode[Tvaraff[k]][lv];
6625: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6626: }
6627: fprintf(ficgp,"\n#\n");
6628: if(invalidvarcomb[k1]){
6629: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6630: continue;
6631: }
6632:
6633: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
6634: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1);
6635: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
6636: set ter svg size 640, 480\n \
6637: unset log y\n \
1.219 brouard 6638: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6639: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
6640: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6641: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6642: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6643: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6644: if(i==1){
6645: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
6646: }else{
6647: fprintf(ficgp,",\\\n '' ");
6648: }
6649: if(cptcoveff ==0){ /* No covariate */
6650: ioffset=2; /* Age is in 2 */
6651: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6652: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6653: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6654: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6655: fprintf(ficgp," u %d:(", ioffset);
6656: if(i==nlstate+1)
6657: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
6658: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6659: else
6660: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
6661: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6662: }else{ /* more than 2 covariates */
6663: if(cptcoveff ==1){
6664: ioffset=4; /* Age is in 4 */
6665: }else{
6666: ioffset=6; /* Age is in 6 */
6667: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6668: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6669: }
6670: fprintf(ficgp," u %d:(",ioffset);
6671: kl=0;
6672: strcpy(gplotcondition,"(");
6673: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
6674: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
6675: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6676: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6677: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6678: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
6679: kl++;
6680: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
6681: kl++;
6682: if(k <cptcoveff && cptcoveff>1)
6683: sprintf(gplotcondition+strlen(gplotcondition)," && ");
6684: }
6685: strcpy(gplotcondition+strlen(gplotcondition),")");
6686: /* 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 *\/ */
6687: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6688: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6689: /* '' 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*/
6690: if(i==nlstate+1){
6691: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
6692: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6693: }else{
6694: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
6695: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6696: }
6697: } /* end if covariate */
6698: } /* nlstate */
6699: fprintf(ficgp,"\nset out\n");
1.223 brouard 6700: } /* end cpt state*/
6701: } /* end covariate */
6702: } /* End if prevfcast */
1.227 brouard 6703:
6704:
1.223 brouard 6705: /* proba elementaires */
6706: fprintf(ficgp,"\n##############\n#MLE estimated parameters\n#############\n");
1.126 brouard 6707: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 6708: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 6709: for(k=1; k <=(nlstate+ndeath); k++){
6710: if (k != i) {
1.227 brouard 6711: fprintf(ficgp,"# current state %d\n",k);
6712: for(j=1; j <=ncovmodel; j++){
6713: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
6714: jk++;
6715: }
6716: fprintf(ficgp,"\n");
1.126 brouard 6717: }
6718: }
1.223 brouard 6719: }
1.187 brouard 6720: fprintf(ficgp,"##############\n#\n");
1.227 brouard 6721:
1.145 brouard 6722: /*goto avoid;*/
1.200 brouard 6723: fprintf(ficgp,"\n##############\n#Graphics of probabilities or incidences\n#############\n");
1.187 brouard 6724: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
6725: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
6726: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
6727: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
6728: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6729: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6730: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6731: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6732: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
6733: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6734: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
6735: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
6736: fprintf(ficgp,"#\n");
1.223 brouard 6737: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
6738: fprintf(ficgp,"# ng=%d\n",ng);
1.225 brouard 6739: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);
1.223 brouard 6740: for(jk=1; jk <=m; jk++) {
6741: fprintf(ficgp,"# jk=%d\n",jk);
6742: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng);
6743: fprintf(ficgp,"\nset ter svg size 640, 480 ");
6744: if (ng==1){
6745: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
6746: fprintf(ficgp,"\nunset log y");
6747: }else if (ng==2){
6748: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
6749: fprintf(ficgp,"\nset log y");
6750: }else if (ng==3){
6751: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
6752: fprintf(ficgp,"\nset log y");
6753: }else
6754: fprintf(ficgp,"\nunset title ");
6755: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
6756: i=1;
6757: for(k2=1; k2<=nlstate; k2++) {
6758: k3=i;
6759: for(k=1; k<=(nlstate+ndeath); k++) {
6760: if (k != k2){
6761: switch( ng) {
6762: case 1:
6763: if(nagesqr==0)
6764: fprintf(ficgp," p%d+p%d*x",i,i+1);
6765: else /* nagesqr =1 */
6766: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6767: break;
6768: case 2: /* ng=2 */
6769: if(nagesqr==0)
6770: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
6771: else /* nagesqr =1 */
6772: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6773: break;
6774: case 3:
6775: if(nagesqr==0)
6776: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
6777: else /* nagesqr =1 */
6778: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
6779: break;
6780: }
6781: ij=1;/* To be checked else nbcode[0][0] wrong */
6782: for(j=3; j <=ncovmodel-nagesqr; j++) {
6783: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
6784: if(ij <=cptcovage) { /* Bug valgrind */
6785: if((j-2)==Tage[ij]) { /* Bug valgrind */
6786: fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
6787: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
6788: ij++;
6789: }
6790: }
6791: else
1.227 brouard 6792: fprintf(ficgp,"+p%d*%d",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,j-2)]); /* Valgrind bug nbcode */
1.223 brouard 6793: }
6794: }else{
6795: i=i-ncovmodel;
6796: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
6797: fprintf(ficgp," (1.");
6798: }
1.227 brouard 6799:
1.223 brouard 6800: if(ng != 1){
6801: fprintf(ficgp,")/(1");
1.227 brouard 6802:
1.223 brouard 6803: for(k1=1; k1 <=nlstate; k1++){
6804: if(nagesqr==0)
6805: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
6806: else /* nagesqr =1 */
6807: 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 6808:
1.223 brouard 6809: ij=1;
6810: for(j=3; j <=ncovmodel-nagesqr; j++){
6811: if(ij <=cptcovage) { /* Bug valgrind */
6812: if((j-2)==Tage[ij]) { /* Bug valgrind */
6813: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
6814: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
6815: ij++;
6816: }
6817: }
6818: else
1.225 brouard 6819: 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 6820: }
6821: fprintf(ficgp,")");
6822: }
6823: fprintf(ficgp,")");
6824: if(ng ==2)
6825: fprintf(ficgp," t \"p%d%d\" ", k2,k);
6826: else /* ng= 3 */
6827: fprintf(ficgp," t \"i%d%d\" ", k2,k);
6828: }else{ /* end ng <> 1 */
6829: if( k !=k2) /* logit p11 is hard to draw */
6830: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
6831: }
6832: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
6833: fprintf(ficgp,",");
6834: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
6835: fprintf(ficgp,",");
6836: i=i+ncovmodel;
6837: } /* end k */
6838: } /* end k2 */
6839: fprintf(ficgp,"\n set out\n");
6840: } /* end jk */
6841: } /* end ng */
6842: /* avoid: */
6843: fflush(ficgp);
1.126 brouard 6844: } /* end gnuplot */
6845:
6846:
6847: /*************** Moving average **************/
1.219 brouard 6848: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 6849: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 6850:
1.222 brouard 6851: int i, cpt, cptcod;
6852: int modcovmax =1;
6853: int mobilavrange, mob;
6854: int iage=0;
6855:
6856: double sum=0.;
6857: double age;
6858: double *sumnewp, *sumnewm;
6859: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
6860:
6861:
1.225 brouard 6862: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 6863: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
6864:
6865: sumnewp = vector(1,ncovcombmax);
6866: sumnewm = vector(1,ncovcombmax);
6867: agemingood = vector(1,ncovcombmax);
6868: agemaxgood = vector(1,ncovcombmax);
6869:
6870: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6871: sumnewm[cptcod]=0.;
6872: sumnewp[cptcod]=0.;
6873: agemingood[cptcod]=0;
6874: agemaxgood[cptcod]=0;
6875: }
6876: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
6877:
6878: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
6879: if(mobilav==1) mobilavrange=5; /* default */
6880: else mobilavrange=mobilav;
6881: for (age=bage; age<=fage; age++)
6882: for (i=1; i<=nlstate;i++)
6883: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
6884: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
6885: /* We keep the original values on the extreme ages bage, fage and for
6886: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
6887: we use a 5 terms etc. until the borders are no more concerned.
6888: */
6889: for (mob=3;mob <=mobilavrange;mob=mob+2){
6890: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
6891: for (i=1; i<=nlstate;i++){
6892: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6893: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
6894: for (cpt=1;cpt<=(mob-1)/2;cpt++){
6895: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
6896: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
6897: }
6898: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
6899: }
6900: }
6901: }/* end age */
6902: }/* end mob */
6903: }else
6904: return -1;
6905: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6906: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
6907: if(invalidvarcomb[cptcod]){
6908: printf("\nCombination (%d) ignored because no cases \n",cptcod);
6909: continue;
6910: }
1.219 brouard 6911:
1.222 brouard 6912: agemingood[cptcod]=fage-(mob-1)/2;
6913: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
6914: sumnewm[cptcod]=0.;
6915: for (i=1; i<=nlstate;i++){
6916: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6917: }
6918: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
6919: agemingood[cptcod]=age;
6920: }else{ /* bad */
6921: for (i=1; i<=nlstate;i++){
6922: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6923: } /* i */
6924: } /* end bad */
6925: }/* age */
6926: sum=0.;
6927: for (i=1; i<=nlstate;i++){
6928: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6929: }
6930: if(fabs(sum - 1.) > 1.e-3) { /* bad */
6931: 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);
6932: /* for (i=1; i<=nlstate;i++){ */
6933: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
6934: /* } /\* i *\/ */
6935: } /* end bad */
6936: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
6937: /* From youngest, finding the oldest wrong */
6938: agemaxgood[cptcod]=bage+(mob-1)/2;
6939: for (age=bage+(mob-1)/2; age<=fage; age++){
6940: sumnewm[cptcod]=0.;
6941: for (i=1; i<=nlstate;i++){
6942: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6943: }
6944: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
6945: agemaxgood[cptcod]=age;
6946: }else{ /* bad */
6947: for (i=1; i<=nlstate;i++){
6948: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6949: } /* i */
6950: } /* end bad */
6951: }/* age */
6952: sum=0.;
6953: for (i=1; i<=nlstate;i++){
6954: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6955: }
6956: if(fabs(sum - 1.) > 1.e-3) { /* bad */
6957: 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);
6958: /* for (i=1; i<=nlstate;i++){ */
6959: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
6960: /* } /\* i *\/ */
6961: } /* end bad */
6962:
6963: for (age=bage; age<=fage; age++){
1.235 brouard 6964: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 6965: sumnewp[cptcod]=0.;
6966: sumnewm[cptcod]=0.;
6967: for (i=1; i<=nlstate;i++){
6968: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
6969: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6970: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
6971: }
6972: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
6973: }
6974: /* printf("\n"); */
6975: /* } */
6976: /* brutal averaging */
6977: for (i=1; i<=nlstate;i++){
6978: for (age=1; age<=bage; age++){
6979: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6980: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
6981: }
6982: for (age=fage; age<=AGESUP; age++){
6983: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6984: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
6985: }
6986: } /* end i status */
6987: for (i=nlstate+1; i<=nlstate+ndeath;i++){
6988: for (age=1; age<=AGESUP; age++){
6989: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
6990: mobaverage[(int)age][i][cptcod]=0.;
6991: }
6992: }
6993: }/* end cptcod */
6994: free_vector(sumnewm,1, ncovcombmax);
6995: free_vector(sumnewp,1, ncovcombmax);
6996: free_vector(agemaxgood,1, ncovcombmax);
6997: free_vector(agemingood,1, ncovcombmax);
6998: return 0;
6999: }/* End movingaverage */
1.218 brouard 7000:
1.126 brouard 7001:
7002: /************** Forecasting ******************/
1.235 brouard 7003: 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 7004: /* proj1, year, month, day of starting projection
7005: agemin, agemax range of age
7006: dateprev1 dateprev2 range of dates during which prevalence is computed
7007: anproj2 year of en of projection (same day and month as proj1).
7008: */
1.235 brouard 7009: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7010: double agec; /* generic age */
7011: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7012: double *popeffectif,*popcount;
7013: double ***p3mat;
1.218 brouard 7014: /* double ***mobaverage; */
1.126 brouard 7015: char fileresf[FILENAMELENGTH];
7016:
7017: agelim=AGESUP;
1.211 brouard 7018: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7019: in each health status at the date of interview (if between dateprev1 and dateprev2).
7020: We still use firstpass and lastpass as another selection.
7021: */
1.214 brouard 7022: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7023: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7024:
1.201 brouard 7025: strcpy(fileresf,"F_");
7026: strcat(fileresf,fileresu);
1.126 brouard 7027: if((ficresf=fopen(fileresf,"w"))==NULL) {
7028: printf("Problem with forecast resultfile: %s\n", fileresf);
7029: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7030: }
1.235 brouard 7031: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7032: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7033:
1.225 brouard 7034: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7035:
7036:
7037: stepsize=(int) (stepm+YEARM-1)/YEARM;
7038: if (stepm<=12) stepsize=1;
7039: if(estepm < stepm){
7040: printf ("Problem %d lower than %d\n",estepm, stepm);
7041: }
7042: else hstepm=estepm;
7043:
7044: hstepm=hstepm/stepm;
7045: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7046: fractional in yp1 */
7047: anprojmean=yp;
7048: yp2=modf((yp1*12),&yp);
7049: mprojmean=yp;
7050: yp1=modf((yp2*30.5),&yp);
7051: jprojmean=yp;
7052: if(jprojmean==0) jprojmean=1;
7053: if(mprojmean==0) jprojmean=1;
7054:
1.227 brouard 7055: i1=pow(2,cptcoveff);
1.126 brouard 7056: if (cptcovn < 1){i1=1;}
7057:
7058: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7059:
7060: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7061:
1.126 brouard 7062: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7063: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7064: for(k=1; k<=i1;k++){
7065: if(TKresult[nres]!= k)
7066: continue;
1.227 brouard 7067: if(invalidvarcomb[k]){
7068: printf("\nCombination (%d) projection ignored because no cases \n",k);
7069: continue;
7070: }
7071: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7072: for(j=1;j<=cptcoveff;j++) {
7073: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7074: }
1.235 brouard 7075: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7076: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7077: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
7078: }
1.227 brouard 7079: fprintf(ficresf," yearproj age");
7080: for(j=1; j<=nlstate+ndeath;j++){
7081: for(i=1; i<=nlstate;i++)
7082: fprintf(ficresf," p%d%d",i,j);
7083: fprintf(ficresf," wp.%d",j);
7084: }
7085: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7086: fprintf(ficresf,"\n");
7087: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7088: for (agec=fage; agec>=(ageminpar-1); agec--){
7089: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7090: nhstepm = nhstepm/hstepm;
7091: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7092: oldm=oldms;savm=savms;
1.235 brouard 7093: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7094:
7095: for (h=0; h<=nhstepm; h++){
7096: if (h*hstepm/YEARM*stepm ==yearp) {
7097: fprintf(ficresf,"\n");
7098: for(j=1;j<=cptcoveff;j++)
7099: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7100: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7101: }
7102: for(j=1; j<=nlstate+ndeath;j++) {
7103: ppij=0.;
7104: for(i=1; i<=nlstate;i++) {
7105: if (mobilav==1)
7106: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7107: else {
7108: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7109: }
7110: if (h*hstepm/YEARM*stepm== yearp) {
7111: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7112: }
7113: } /* end i */
7114: if (h*hstepm/YEARM*stepm==yearp) {
7115: fprintf(ficresf," %.3f", ppij);
7116: }
7117: }/* end j */
7118: } /* end h */
7119: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7120: } /* end agec */
7121: } /* end yearp */
7122: } /* end k */
1.219 brouard 7123:
1.126 brouard 7124: fclose(ficresf);
1.215 brouard 7125: printf("End of Computing forecasting \n");
7126: fprintf(ficlog,"End of Computing forecasting\n");
7127:
1.126 brouard 7128: }
7129:
1.218 brouard 7130: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7131: /* 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 7132: /* /\* back1, year, month, day of starting backection */
7133: /* agemin, agemax range of age */
7134: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7135: /* anback2 year of en of backection (same day and month as back1). */
7136: /* *\/ */
7137: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7138: /* double agec; /\* generic age *\/ */
7139: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7140: /* double *popeffectif,*popcount; */
7141: /* double ***p3mat; */
7142: /* /\* double ***mobaverage; *\/ */
7143: /* char fileresfb[FILENAMELENGTH]; */
7144:
7145: /* agelim=AGESUP; */
7146: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7147: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7148: /* We still use firstpass and lastpass as another selection. */
7149: /* *\/ */
7150: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7151: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7152: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7153:
7154: /* strcpy(fileresfb,"FB_"); */
7155: /* strcat(fileresfb,fileresu); */
7156: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7157: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7158: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7159: /* } */
7160: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7161: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7162:
1.225 brouard 7163: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7164:
7165: /* /\* if (mobilav!=0) { *\/ */
7166: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7167: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7168: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7169: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7170: /* /\* } *\/ */
7171: /* /\* } *\/ */
7172:
7173: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7174: /* if (stepm<=12) stepsize=1; */
7175: /* if(estepm < stepm){ */
7176: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7177: /* } */
7178: /* else hstepm=estepm; */
7179:
7180: /* hstepm=hstepm/stepm; */
7181: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7182: /* fractional in yp1 *\/ */
7183: /* anprojmean=yp; */
7184: /* yp2=modf((yp1*12),&yp); */
7185: /* mprojmean=yp; */
7186: /* yp1=modf((yp2*30.5),&yp); */
7187: /* jprojmean=yp; */
7188: /* if(jprojmean==0) jprojmean=1; */
7189: /* if(mprojmean==0) jprojmean=1; */
7190:
1.225 brouard 7191: /* i1=cptcoveff; */
1.218 brouard 7192: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7193:
1.218 brouard 7194: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7195:
1.218 brouard 7196: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7197:
7198: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7199: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7200: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7201: /* k=k+1; */
7202: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7203: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7204: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7205: /* } */
7206: /* fprintf(ficresfb," yearbproj age"); */
7207: /* for(j=1; j<=nlstate+ndeath;j++){ */
7208: /* for(i=1; i<=nlstate;i++) */
7209: /* fprintf(ficresfb," p%d%d",i,j); */
7210: /* fprintf(ficresfb," p.%d",j); */
7211: /* } */
7212: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7213: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7214: /* fprintf(ficresfb,"\n"); */
7215: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7216: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7217: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7218: /* nhstepm = nhstepm/hstepm; */
7219: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7220: /* oldm=oldms;savm=savms; */
7221: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7222: /* for (h=0; h<=nhstepm; h++){ */
7223: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7224: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7225: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7226: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7227: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7228: /* } */
7229: /* for(j=1; j<=nlstate+ndeath;j++) { */
7230: /* ppij=0.; */
7231: /* for(i=1; i<=nlstate;i++) { */
7232: /* if (mobilav==1) */
7233: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7234: /* else { */
7235: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7236: /* } */
7237: /* if (h*hstepm/YEARM*stepm== yearp) { */
7238: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7239: /* } */
7240: /* } /\* end i *\/ */
7241: /* if (h*hstepm/YEARM*stepm==yearp) { */
7242: /* fprintf(ficresfb," %.3f", ppij); */
7243: /* } */
7244: /* }/\* end j *\/ */
7245: /* } /\* end h *\/ */
7246: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7247: /* } /\* end agec *\/ */
7248: /* } /\* end yearp *\/ */
7249: /* } /\* end cptcod *\/ */
7250: /* } /\* end cptcov *\/ */
7251:
7252: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7253:
7254: /* fclose(ficresfb); */
7255: /* printf("End of Computing Back forecasting \n"); */
7256: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7257:
1.218 brouard 7258: /* } */
1.217 brouard 7259:
1.126 brouard 7260: /************** Forecasting *****not tested NB*************/
1.227 brouard 7261: /* 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 7262:
1.227 brouard 7263: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7264: /* int *popage; */
7265: /* double calagedatem, agelim, kk1, kk2; */
7266: /* double *popeffectif,*popcount; */
7267: /* double ***p3mat,***tabpop,***tabpopprev; */
7268: /* /\* double ***mobaverage; *\/ */
7269: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7270:
1.227 brouard 7271: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7272: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7273: /* agelim=AGESUP; */
7274: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7275:
1.227 brouard 7276: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7277:
7278:
1.227 brouard 7279: /* strcpy(filerespop,"POP_"); */
7280: /* strcat(filerespop,fileresu); */
7281: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7282: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7283: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7284: /* } */
7285: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7286: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7287:
1.227 brouard 7288: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7289:
1.227 brouard 7290: /* /\* if (mobilav!=0) { *\/ */
7291: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7292: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7293: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7294: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7295: /* /\* } *\/ */
7296: /* /\* } *\/ */
1.126 brouard 7297:
1.227 brouard 7298: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7299: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7300:
1.227 brouard 7301: /* agelim=AGESUP; */
1.126 brouard 7302:
1.227 brouard 7303: /* hstepm=1; */
7304: /* hstepm=hstepm/stepm; */
1.218 brouard 7305:
1.227 brouard 7306: /* if (popforecast==1) { */
7307: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7308: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7309: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7310: /* } */
7311: /* popage=ivector(0,AGESUP); */
7312: /* popeffectif=vector(0,AGESUP); */
7313: /* popcount=vector(0,AGESUP); */
1.126 brouard 7314:
1.227 brouard 7315: /* i=1; */
7316: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7317:
1.227 brouard 7318: /* imx=i; */
7319: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7320: /* } */
1.218 brouard 7321:
1.227 brouard 7322: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7323: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7324: /* k=k+1; */
7325: /* fprintf(ficrespop,"\n#******"); */
7326: /* for(j=1;j<=cptcoveff;j++) { */
7327: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7328: /* } */
7329: /* fprintf(ficrespop,"******\n"); */
7330: /* fprintf(ficrespop,"# Age"); */
7331: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7332: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7333:
1.227 brouard 7334: /* for (cpt=0; cpt<=0;cpt++) { */
7335: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7336:
1.227 brouard 7337: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7338: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7339: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7340:
1.227 brouard 7341: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7342: /* oldm=oldms;savm=savms; */
7343: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7344:
1.227 brouard 7345: /* for (h=0; h<=nhstepm; h++){ */
7346: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7347: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7348: /* } */
7349: /* for(j=1; j<=nlstate+ndeath;j++) { */
7350: /* kk1=0.;kk2=0; */
7351: /* for(i=1; i<=nlstate;i++) { */
7352: /* if (mobilav==1) */
7353: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7354: /* else { */
7355: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7356: /* } */
7357: /* } */
7358: /* if (h==(int)(calagedatem+12*cpt)){ */
7359: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7360: /* /\*fprintf(ficrespop," %.3f", kk1); */
7361: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7362: /* } */
7363: /* } */
7364: /* for(i=1; i<=nlstate;i++){ */
7365: /* kk1=0.; */
7366: /* for(j=1; j<=nlstate;j++){ */
7367: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7368: /* } */
7369: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7370: /* } */
1.218 brouard 7371:
1.227 brouard 7372: /* if (h==(int)(calagedatem+12*cpt)) */
7373: /* for(j=1; j<=nlstate;j++) */
7374: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7375: /* } */
7376: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7377: /* } */
7378: /* } */
1.218 brouard 7379:
1.227 brouard 7380: /* /\******\/ */
1.218 brouard 7381:
1.227 brouard 7382: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7383: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7384: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7385: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7386: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7387:
1.227 brouard 7388: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7389: /* oldm=oldms;savm=savms; */
7390: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7391: /* for (h=0; h<=nhstepm; h++){ */
7392: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7393: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7394: /* } */
7395: /* for(j=1; j<=nlstate+ndeath;j++) { */
7396: /* kk1=0.;kk2=0; */
7397: /* for(i=1; i<=nlstate;i++) { */
7398: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7399: /* } */
7400: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7401: /* } */
7402: /* } */
7403: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7404: /* } */
7405: /* } */
7406: /* } */
7407: /* } */
1.218 brouard 7408:
1.227 brouard 7409: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7410:
1.227 brouard 7411: /* if (popforecast==1) { */
7412: /* free_ivector(popage,0,AGESUP); */
7413: /* free_vector(popeffectif,0,AGESUP); */
7414: /* free_vector(popcount,0,AGESUP); */
7415: /* } */
7416: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7417: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7418: /* fclose(ficrespop); */
7419: /* } /\* End of popforecast *\/ */
1.218 brouard 7420:
1.126 brouard 7421: int fileappend(FILE *fichier, char *optionfich)
7422: {
7423: if((fichier=fopen(optionfich,"a"))==NULL) {
7424: printf("Problem with file: %s\n", optionfich);
7425: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7426: return (0);
7427: }
7428: fflush(fichier);
7429: return (1);
7430: }
7431:
7432:
7433: /**************** function prwizard **********************/
7434: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7435: {
7436:
7437: /* Wizard to print covariance matrix template */
7438:
1.164 brouard 7439: char ca[32], cb[32];
7440: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7441: int numlinepar;
7442:
7443: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7444: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7445: for(i=1; i <=nlstate; i++){
7446: jj=0;
7447: for(j=1; j <=nlstate+ndeath; j++){
7448: if(j==i) continue;
7449: jj++;
7450: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7451: printf("%1d%1d",i,j);
7452: fprintf(ficparo,"%1d%1d",i,j);
7453: for(k=1; k<=ncovmodel;k++){
7454: /* printf(" %lf",param[i][j][k]); */
7455: /* fprintf(ficparo," %lf",param[i][j][k]); */
7456: printf(" 0.");
7457: fprintf(ficparo," 0.");
7458: }
7459: printf("\n");
7460: fprintf(ficparo,"\n");
7461: }
7462: }
7463: printf("# Scales (for hessian or gradient estimation)\n");
7464: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7465: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
7466: for(i=1; i <=nlstate; i++){
7467: jj=0;
7468: for(j=1; j <=nlstate+ndeath; j++){
7469: if(j==i) continue;
7470: jj++;
7471: fprintf(ficparo,"%1d%1d",i,j);
7472: printf("%1d%1d",i,j);
7473: fflush(stdout);
7474: for(k=1; k<=ncovmodel;k++){
7475: /* printf(" %le",delti3[i][j][k]); */
7476: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7477: printf(" 0.");
7478: fprintf(ficparo," 0.");
7479: }
7480: numlinepar++;
7481: printf("\n");
7482: fprintf(ficparo,"\n");
7483: }
7484: }
7485: printf("# Covariance matrix\n");
7486: /* # 121 Var(a12)\n\ */
7487: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7488: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7489: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7490: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7491: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7492: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7493: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7494: fflush(stdout);
7495: fprintf(ficparo,"# Covariance matrix\n");
7496: /* # 121 Var(a12)\n\ */
7497: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7498: /* # ...\n\ */
7499: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7500:
7501: for(itimes=1;itimes<=2;itimes++){
7502: jj=0;
7503: for(i=1; i <=nlstate; i++){
7504: for(j=1; j <=nlstate+ndeath; j++){
7505: if(j==i) continue;
7506: for(k=1; k<=ncovmodel;k++){
7507: jj++;
7508: ca[0]= k+'a'-1;ca[1]='\0';
7509: if(itimes==1){
7510: printf("#%1d%1d%d",i,j,k);
7511: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7512: }else{
7513: printf("%1d%1d%d",i,j,k);
7514: fprintf(ficparo,"%1d%1d%d",i,j,k);
7515: /* printf(" %.5le",matcov[i][j]); */
7516: }
7517: ll=0;
7518: for(li=1;li <=nlstate; li++){
7519: for(lj=1;lj <=nlstate+ndeath; lj++){
7520: if(lj==li) continue;
7521: for(lk=1;lk<=ncovmodel;lk++){
7522: ll++;
7523: if(ll<=jj){
7524: cb[0]= lk +'a'-1;cb[1]='\0';
7525: if(ll<jj){
7526: if(itimes==1){
7527: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7528: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7529: }else{
7530: printf(" 0.");
7531: fprintf(ficparo," 0.");
7532: }
7533: }else{
7534: if(itimes==1){
7535: printf(" Var(%s%1d%1d)",ca,i,j);
7536: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
7537: }else{
7538: printf(" 0.");
7539: fprintf(ficparo," 0.");
7540: }
7541: }
7542: }
7543: } /* end lk */
7544: } /* end lj */
7545: } /* end li */
7546: printf("\n");
7547: fprintf(ficparo,"\n");
7548: numlinepar++;
7549: } /* end k*/
7550: } /*end j */
7551: } /* end i */
7552: } /* end itimes */
7553:
7554: } /* end of prwizard */
7555: /******************* Gompertz Likelihood ******************************/
7556: double gompertz(double x[])
7557: {
7558: double A,B,L=0.0,sump=0.,num=0.;
7559: int i,n=0; /* n is the size of the sample */
7560:
1.220 brouard 7561: for (i=1;i<=imx ; i++) {
1.126 brouard 7562: sump=sump+weight[i];
7563: /* sump=sump+1;*/
7564: num=num+1;
7565: }
7566:
7567:
7568: /* for (i=0; i<=imx; i++)
7569: 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]);*/
7570:
7571: for (i=1;i<=imx ; i++)
7572: {
7573: if (cens[i] == 1 && wav[i]>1)
7574: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
7575:
7576: if (cens[i] == 0 && wav[i]>1)
7577: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
7578: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
7579:
7580: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7581: if (wav[i] > 1 ) { /* ??? */
7582: L=L+A*weight[i];
7583: /* 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]);*/
7584: }
7585: }
7586:
7587: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7588:
7589: return -2*L*num/sump;
7590: }
7591:
1.136 brouard 7592: #ifdef GSL
7593: /******************* Gompertz_f Likelihood ******************************/
7594: double gompertz_f(const gsl_vector *v, void *params)
7595: {
7596: double A,B,LL=0.0,sump=0.,num=0.;
7597: double *x= (double *) v->data;
7598: int i,n=0; /* n is the size of the sample */
7599:
7600: for (i=0;i<=imx-1 ; i++) {
7601: sump=sump+weight[i];
7602: /* sump=sump+1;*/
7603: num=num+1;
7604: }
7605:
7606:
7607: /* for (i=0; i<=imx; i++)
7608: 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]);*/
7609: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
7610: for (i=1;i<=imx ; i++)
7611: {
7612: if (cens[i] == 1 && wav[i]>1)
7613: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
7614:
7615: if (cens[i] == 0 && wav[i]>1)
7616: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
7617: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
7618:
7619: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7620: if (wav[i] > 1 ) { /* ??? */
7621: LL=LL+A*weight[i];
7622: /* 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]);*/
7623: }
7624: }
7625:
7626: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7627: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
7628:
7629: return -2*LL*num/sump;
7630: }
7631: #endif
7632:
1.126 brouard 7633: /******************* Printing html file ***********/
1.201 brouard 7634: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 7635: int lastpass, int stepm, int weightopt, char model[],\
7636: int imx, double p[],double **matcov,double agemortsup){
7637: int i,k;
7638:
7639: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
7640: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
7641: for (i=1;i<=2;i++)
7642: 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 7643: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 7644: fprintf(fichtm,"</ul>");
7645:
7646: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
7647:
7648: 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>");
7649:
7650: for (k=agegomp;k<(agemortsup-2);k++)
7651: 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]);
7652:
7653:
7654: fflush(fichtm);
7655: }
7656:
7657: /******************* Gnuplot file **************/
1.201 brouard 7658: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 7659:
7660: char dirfileres[132],optfileres[132];
1.164 brouard 7661:
1.126 brouard 7662: int ng;
7663:
7664:
7665: /*#ifdef windows */
7666: fprintf(ficgp,"cd \"%s\" \n",pathc);
7667: /*#endif */
7668:
7669:
7670: strcpy(dirfileres,optionfilefiname);
7671: strcpy(optfileres,"vpl");
1.199 brouard 7672: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 7673: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 7674: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 7675: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 7676: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
7677:
7678: }
7679:
1.136 brouard 7680: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
7681: {
1.126 brouard 7682:
1.136 brouard 7683: /*-------- data file ----------*/
7684: FILE *fic;
7685: char dummy[]=" ";
1.223 brouard 7686: int i=0, j=0, n=0, iv=0;
7687: int lstra;
1.136 brouard 7688: int linei, month, year,iout;
7689: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 7690: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 7691: char *stratrunc;
1.223 brouard 7692:
1.126 brouard 7693:
7694:
1.136 brouard 7695: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 7696: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
7697: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 7698: }
1.126 brouard 7699:
1.136 brouard 7700: i=1;
7701: linei=0;
7702: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
7703: linei=linei+1;
7704: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
7705: if(line[j] == '\t')
7706: line[j] = ' ';
7707: }
7708: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
7709: ;
7710: };
7711: line[j+1]=0; /* Trims blanks at end of line */
7712: if(line[0]=='#'){
7713: fprintf(ficlog,"Comment line\n%s\n",line);
7714: printf("Comment line\n%s\n",line);
7715: continue;
7716: }
7717: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 7718: strcpy(line, linetmp);
1.223 brouard 7719:
7720: /* Loops on waves */
7721: for (j=maxwav;j>=1;j--){
7722: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.232 brouard 7723: cutv(stra, strb, line, ' ');
7724: if(strb[0]=='.') { /* Missing value */
7725: lval=-1;
7726: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
7727: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
7728: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
7729: 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);
7730: 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);
7731: return 1;
7732: }
7733: }else{
7734: errno=0;
7735: /* what_kind_of_number(strb); */
7736: dval=strtod(strb,&endptr);
7737: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
7738: /* if(strb != endptr && *endptr == '\0') */
7739: /* dval=dlval; */
7740: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
7741: if( strb[0]=='\0' || (*endptr != '\0')){
7742: 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);
7743: 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);
7744: return 1;
7745: }
7746: cotqvar[j][iv][i]=dval;
7747: cotvar[j][ntv+iv][i]=dval;
7748: }
7749: strcpy(line,stra);
1.223 brouard 7750: }/* end loop ntqv */
1.225 brouard 7751:
1.223 brouard 7752: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.232 brouard 7753: cutv(stra, strb, line, ' ');
7754: if(strb[0]=='.') { /* Missing value */
7755: lval=-1;
7756: }else{
7757: errno=0;
7758: lval=strtol(strb,&endptr,10);
7759: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
7760: if( strb[0]=='\0' || (*endptr != '\0')){
7761: 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);
7762: 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);
7763: return 1;
7764: }
7765: }
7766: if(lval <-1 || lval >1){
7767: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 7768: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7769: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.232 brouard 7770: For example, for multinomial values like 1, 2 and 3,\n \
7771: build V1=0 V2=0 for the reference value (1),\n \
7772: V1=1 V2=0 for (2) \n \
1.223 brouard 7773: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.232 brouard 7774: output of IMaCh is often meaningless.\n \
1.223 brouard 7775: Exiting.\n",lval,linei, i,line,j);
1.232 brouard 7776: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 7777: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7778: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.232 brouard 7779: For example, for multinomial values like 1, 2 and 3,\n \
7780: build V1=0 V2=0 for the reference value (1),\n \
7781: V1=1 V2=0 for (2) \n \
1.223 brouard 7782: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.232 brouard 7783: output of IMaCh is often meaningless.\n \
1.223 brouard 7784: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.232 brouard 7785: return 1;
7786: }
7787: cotvar[j][iv][i]=(double)(lval);
7788: strcpy(line,stra);
1.223 brouard 7789: }/* end loop ntv */
1.225 brouard 7790:
1.223 brouard 7791: /* Statuses at wave */
1.137 brouard 7792: cutv(stra, strb, line, ' ');
1.223 brouard 7793: if(strb[0]=='.') { /* Missing value */
1.232 brouard 7794: lval=-1;
1.136 brouard 7795: }else{
1.232 brouard 7796: errno=0;
7797: lval=strtol(strb,&endptr,10);
7798: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
7799: if( strb[0]=='\0' || (*endptr != '\0')){
7800: 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);
7801: 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);
7802: return 1;
7803: }
1.136 brouard 7804: }
1.225 brouard 7805:
1.136 brouard 7806: s[j][i]=lval;
1.225 brouard 7807:
1.223 brouard 7808: /* Date of Interview */
1.136 brouard 7809: strcpy(line,stra);
7810: cutv(stra, strb,line,' ');
1.169 brouard 7811: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7812: }
1.169 brouard 7813: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 7814: month=99;
7815: year=9999;
1.136 brouard 7816: }else{
1.225 brouard 7817: 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);
7818: 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);
7819: return 1;
1.136 brouard 7820: }
7821: anint[j][i]= (double) year;
7822: mint[j][i]= (double)month;
7823: strcpy(line,stra);
1.223 brouard 7824: } /* End loop on waves */
1.225 brouard 7825:
1.223 brouard 7826: /* Date of death */
1.136 brouard 7827: cutv(stra, strb,line,' ');
1.169 brouard 7828: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7829: }
1.169 brouard 7830: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 7831: month=99;
7832: year=9999;
7833: }else{
1.141 brouard 7834: 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 7835: 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);
7836: return 1;
1.136 brouard 7837: }
7838: andc[i]=(double) year;
7839: moisdc[i]=(double) month;
7840: strcpy(line,stra);
7841:
1.223 brouard 7842: /* Date of birth */
1.136 brouard 7843: cutv(stra, strb,line,' ');
1.169 brouard 7844: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7845: }
1.169 brouard 7846: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 7847: month=99;
7848: year=9999;
7849: }else{
1.141 brouard 7850: 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);
7851: 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 7852: return 1;
1.136 brouard 7853: }
7854: if (year==9999) {
1.141 brouard 7855: 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);
7856: 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 7857: return 1;
7858:
1.136 brouard 7859: }
7860: annais[i]=(double)(year);
7861: moisnais[i]=(double)(month);
7862: strcpy(line,stra);
1.225 brouard 7863:
1.223 brouard 7864: /* Sample weight */
1.136 brouard 7865: cutv(stra, strb,line,' ');
7866: errno=0;
7867: dval=strtod(strb,&endptr);
7868: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 7869: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
7870: 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 7871: fflush(ficlog);
7872: return 1;
7873: }
7874: weight[i]=dval;
7875: strcpy(line,stra);
1.225 brouard 7876:
1.223 brouard 7877: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
7878: cutv(stra, strb, line, ' ');
7879: if(strb[0]=='.') { /* Missing value */
1.225 brouard 7880: lval=-1;
1.223 brouard 7881: }else{
1.225 brouard 7882: errno=0;
7883: /* what_kind_of_number(strb); */
7884: dval=strtod(strb,&endptr);
7885: /* if(strb != endptr && *endptr == '\0') */
7886: /* dval=dlval; */
7887: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
7888: if( strb[0]=='\0' || (*endptr != '\0')){
7889: 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);
7890: 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);
7891: return 1;
7892: }
7893: coqvar[iv][i]=dval;
1.226 brouard 7894: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 7895: }
7896: strcpy(line,stra);
7897: }/* end loop nqv */
1.136 brouard 7898:
1.223 brouard 7899: /* Covariate values */
1.136 brouard 7900: for (j=ncovcol;j>=1;j--){
7901: cutv(stra, strb,line,' ');
1.223 brouard 7902: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 7903: lval=-1;
1.136 brouard 7904: }else{
1.225 brouard 7905: errno=0;
7906: lval=strtol(strb,&endptr,10);
7907: if( strb[0]=='\0' || (*endptr != '\0')){
7908: 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);
7909: 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);
7910: return 1;
7911: }
1.136 brouard 7912: }
7913: if(lval <-1 || lval >1){
1.225 brouard 7914: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 7915: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7916: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 7917: For example, for multinomial values like 1, 2 and 3,\n \
7918: build V1=0 V2=0 for the reference value (1),\n \
7919: V1=1 V2=0 for (2) \n \
1.136 brouard 7920: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 7921: output of IMaCh is often meaningless.\n \
1.136 brouard 7922: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 7923: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 7924: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7925: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 7926: For example, for multinomial values like 1, 2 and 3,\n \
7927: build V1=0 V2=0 for the reference value (1),\n \
7928: V1=1 V2=0 for (2) \n \
1.136 brouard 7929: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 7930: output of IMaCh is often meaningless.\n \
1.136 brouard 7931: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 7932: return 1;
1.136 brouard 7933: }
7934: covar[j][i]=(double)(lval);
7935: strcpy(line,stra);
7936: }
7937: lstra=strlen(stra);
1.225 brouard 7938:
1.136 brouard 7939: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
7940: stratrunc = &(stra[lstra-9]);
7941: num[i]=atol(stratrunc);
7942: }
7943: else
7944: num[i]=atol(stra);
7945: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
7946: 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;}*/
7947:
7948: i=i+1;
7949: } /* End loop reading data */
1.225 brouard 7950:
1.136 brouard 7951: *imax=i-1; /* Number of individuals */
7952: fclose(fic);
1.225 brouard 7953:
1.136 brouard 7954: return (0);
1.164 brouard 7955: /* endread: */
1.225 brouard 7956: printf("Exiting readdata: ");
7957: fclose(fic);
7958: return (1);
1.223 brouard 7959: }
1.126 brouard 7960:
1.234 brouard 7961: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 7962: char *p1 = *stri, *p2 = *stri;
1.235 brouard 7963: while (*p2 == ' ')
1.234 brouard 7964: p2++;
7965: /* while ((*p1++ = *p2++) !=0) */
7966: /* ; */
7967: /* do */
7968: /* while (*p2 == ' ') */
7969: /* p2++; */
7970: /* while (*p1++ == *p2++); */
7971: *stri=p2;
1.145 brouard 7972: }
7973:
1.235 brouard 7974: int decoderesult ( char resultline[], int nres)
1.230 brouard 7975: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
7976: {
1.235 brouard 7977: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 7978: char resultsav[MAXLINE];
1.234 brouard 7979: int resultmodel[MAXLINE];
7980: int modelresult[MAXLINE];
1.230 brouard 7981: char stra[80], strb[80], strc[80], strd[80],stre[80];
7982:
1.234 brouard 7983: removefirstspace(&resultline);
1.233 brouard 7984: printf("decoderesult:%s\n",resultline);
1.230 brouard 7985:
7986: if (strstr(resultline,"v") !=0){
7987: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
7988: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
7989: return 1;
7990: }
7991: trimbb(resultsav, resultline);
7992: if (strlen(resultsav) >1){
7993: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
7994: }
1.234 brouard 7995: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
7996: 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);
7997: 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);
7998: }
7999: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8000: if(nbocc(resultsav,'=') >1){
8001: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8002: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8003: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8004: }else
8005: cutl(strc,strd,resultsav,'=');
1.230 brouard 8006: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8007:
1.230 brouard 8008: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8009: Tvarsel[k]=atoi(strc);
8010: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8011: /* cptcovsel++; */
8012: if (nbocc(stra,'=') >0)
8013: strcpy(resultsav,stra); /* and analyzes it */
8014: }
1.235 brouard 8015: /* Checking for missing or useless values in comparison of current model needs */
1.236 ! brouard 8016: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
! 8017: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8018: match=0;
1.236 ! brouard 8019: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
! 8020: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[2]=5 == Tvarsel[1]=4 */
! 8021: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8022: match=1;
8023: break;
8024: }
8025: }
8026: if(match == 0){
8027: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8028: }
8029: }
8030: }
1.235 brouard 8031: /* Checking for missing or useless values in comparison of current model needs */
8032: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8033: match=0;
1.235 brouard 8034: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8035: if(Typevar[k1]==0){ /* Single */
8036: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=5 == Tvarsel[1]=4 */
8037: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8038: ++match;
8039: }
8040: }
8041: }
8042: if(match == 0){
8043: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8044: }else if(match > 1){
8045: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8046: }
8047: }
1.235 brouard 8048:
1.234 brouard 8049: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8050: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8051: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8052: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8053: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8054: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8055: /* 1 0 0 0 */
8056: /* 2 1 0 0 */
8057: /* 3 0 1 0 */
8058: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8059: /* 5 0 0 1 */
8060: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8061: /* 7 0 1 1 */
8062: /* 8 1 1 1 */
8063: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8064: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
8065: k3= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8066: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8067: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
8068: Tresult[nres][k4+1]=Tvalsel[k3];
8069: Tvresult[nres][k4+1]=(int)Tvarsel[k3];
8070: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8071: k4++;;
8072: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8073: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8074: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8075: Tqresult[nres][k4q+1]=Tvalsel[k3q];
8076: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q];
8077: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8078: k4q++;;
8079: }
8080: }
1.234 brouard 8081:
1.235 brouard 8082: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8083: return (0);
8084: }
1.235 brouard 8085:
1.230 brouard 8086: int decodemodel( char model[], int lastobs)
8087: /**< This routine decodes the model and returns:
1.224 brouard 8088: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8089: * - nagesqr = 1 if age*age in the model, otherwise 0.
8090: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8091: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8092: * - cptcovage number of covariates with age*products =2
8093: * - cptcovs number of simple covariates
8094: * - 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
8095: * which is a new column after the 9 (ncovcol) variables.
8096: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8097: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8098: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8099: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8100: */
1.136 brouard 8101: {
1.145 brouard 8102: int i, j, k, ks;
1.227 brouard 8103: int j1, k1, k2, k3, k4;
1.136 brouard 8104: char modelsav[80];
1.145 brouard 8105: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8106: char *strpt;
1.136 brouard 8107:
1.145 brouard 8108: /*removespace(model);*/
1.136 brouard 8109: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8110: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8111: if (strstr(model,"AGE") !=0){
1.192 brouard 8112: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8113: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8114: return 1;
8115: }
1.141 brouard 8116: if (strstr(model,"v") !=0){
8117: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8118: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8119: return 1;
8120: }
1.187 brouard 8121: strcpy(modelsav,model);
8122: if ((strpt=strstr(model,"age*age")) !=0){
8123: printf(" strpt=%s, model=%s\n",strpt, model);
8124: if(strpt != model){
1.234 brouard 8125: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8126: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8127: corresponding column of parameters.\n",model);
1.234 brouard 8128: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8129: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8130: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8131: return 1;
1.225 brouard 8132: }
1.187 brouard 8133: nagesqr=1;
8134: if (strstr(model,"+age*age") !=0)
1.234 brouard 8135: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8136: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8137: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8138: else
1.234 brouard 8139: substrchaine(modelsav, model, "age*age");
1.187 brouard 8140: }else
8141: nagesqr=0;
8142: if (strlen(modelsav) >1){
8143: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8144: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8145: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8146: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8147: * cst, age and age*age
8148: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8149: /* including age products which are counted in cptcovage.
8150: * but the covariates which are products must be treated
8151: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8152: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8153: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8154:
8155:
1.187 brouard 8156: /* Design
8157: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8158: * < ncovcol=8 >
8159: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8160: * k= 1 2 3 4 5 6 7 8
8161: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8162: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8163: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8164: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8165: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8166: * Tage[++cptcovage]=k
8167: * if products, new covar are created after ncovcol with k1
8168: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8169: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8170: * 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
8171: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8172: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8173: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8174: * < ncovcol=8 >
8175: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8176: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8177: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8178: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8179: * p Tprod[1]@2={ 6, 5}
8180: *p Tvard[1][1]@4= {7, 8, 5, 6}
8181: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8182: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8183: *How to reorganize?
8184: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8185: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8186: * {2, 1, 4, 8, 5, 6, 3, 7}
8187: * Struct []
8188: */
1.225 brouard 8189:
1.187 brouard 8190: /* This loop fills the array Tvar from the string 'model'.*/
8191: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8192: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8193: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8194: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8195: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8196: /* k=1 Tvar[1]=2 (from V2) */
8197: /* k=5 Tvar[5] */
8198: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8199: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8200: /* } */
1.198 brouard 8201: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8202: /*
8203: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8204: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8205: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8206: }
1.187 brouard 8207: cptcovage=0;
8208: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8209: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8210: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8211: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8212: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8213: /*scanf("%d",i);*/
8214: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8215: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8216: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8217: /* covar is not filled and then is empty */
8218: cptcovprod--;
8219: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8220: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8221: Typevar[k]=1; /* 1 for age product */
8222: cptcovage++; /* Sums the number of covariates which include age as a product */
8223: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8224: /*printf("stre=%s ", stre);*/
8225: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8226: cptcovprod--;
8227: cutl(stre,strb,strc,'V');
8228: Tvar[k]=atoi(stre);
8229: Typevar[k]=1; /* 1 for age product */
8230: cptcovage++;
8231: Tage[cptcovage]=k;
8232: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8233: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8234: cptcovn++;
8235: cptcovprodnoage++;k1++;
8236: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8237: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8238: because this model-covariate is a construction we invent a new column
8239: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8240: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8241: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8242: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8243: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8244: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8245: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8246: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8247: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8248: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8249: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8250: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8251: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8252: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8253: for (i=1; i<=lastobs;i++){
8254: /* Computes the new covariate which is a product of
8255: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8256: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8257: }
8258: } /* End age is not in the model */
8259: } /* End if model includes a product */
8260: else { /* no more sum */
8261: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8262: /* scanf("%d",i);*/
8263: cutl(strd,strc,strb,'V');
8264: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8265: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8266: Tvar[k]=atoi(strd);
8267: Typevar[k]=0; /* 0 for simple covariates */
8268: }
8269: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8270: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8271: scanf("%d",i);*/
1.187 brouard 8272: } /* end of loop + on total covariates */
8273: } /* end if strlen(modelsave == 0) age*age might exist */
8274: } /* end if strlen(model == 0) */
1.136 brouard 8275:
8276: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8277: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8278:
1.136 brouard 8279: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8280: printf("cptcovprod=%d ", cptcovprod);
8281: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8282: scanf("%d ",i);*/
8283:
8284:
1.230 brouard 8285: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8286: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8287: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8288: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8289: k = 1 2 3 4 5 6 7 8 9
8290: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8291: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8292: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8293: Dummy[k] 1 0 0 0 3 1 1 2 3
8294: Tmodelind[combination of covar]=k;
1.225 brouard 8295: */
8296: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8297: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8298: /* 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 8299: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8300: printf("Model=%s\n\
8301: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8302: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8303: 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);
8304: fprintf(ficlog,"Model=%s\n\
8305: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8306: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8307: 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);
8308:
1.234 brouard 8309: 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 */
8310: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 8311: Fixed[k]= 0;
8312: Dummy[k]= 0;
1.225 brouard 8313: ncoveff++;
1.232 brouard 8314: ncovf++;
1.234 brouard 8315: nsd++;
8316: modell[k].maintype= FTYPE;
8317: TvarsD[nsd]=Tvar[k];
8318: TvarsDind[nsd]=k;
8319: TvarF[ncovf]=Tvar[k];
8320: TvarFind[ncovf]=k;
8321: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8322: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8323: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
8324: Fixed[k]= 0;
8325: Dummy[k]= 0;
8326: ncoveff++;
8327: ncovf++;
8328: modell[k].maintype= FTYPE;
8329: TvarF[ncovf]=Tvar[k];
8330: TvarFind[ncovf]=k;
1.230 brouard 8331: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8332: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8333: }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 8334: Fixed[k]= 0;
8335: Dummy[k]= 1;
1.230 brouard 8336: nqfveff++;
1.234 brouard 8337: modell[k].maintype= FTYPE;
8338: modell[k].subtype= FQ;
8339: nsq++;
8340: TvarsQ[nsq]=Tvar[k];
8341: TvarsQind[nsq]=k;
1.232 brouard 8342: ncovf++;
1.234 brouard 8343: TvarF[ncovf]=Tvar[k];
8344: TvarFind[ncovf]=k;
1.231 brouard 8345: 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 8346: 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 */
1.234 brouard 8347: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying variables */
1.227 brouard 8348: Fixed[k]= 1;
8349: Dummy[k]= 0;
1.225 brouard 8350: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 8351: modell[k].maintype= VTYPE;
8352: modell[k].subtype= VD;
8353: nsd++;
8354: TvarsD[nsd]=Tvar[k];
8355: TvarsDind[nsd]=k;
8356: ncovv++; /* Only simple time varying variables */
8357: TvarV[ncovv]=Tvar[k];
8358: TvarVind[ncovv]=k;
1.231 brouard 8359: 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 */
8360: 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 8361: 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);
8362: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8363: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 8364: Fixed[k]= 1;
8365: Dummy[k]= 1;
8366: nqtveff++;
8367: modell[k].maintype= VTYPE;
8368: modell[k].subtype= VQ;
8369: ncovv++; /* Only simple time varying variables */
8370: nsq++;
8371: TvarsQ[nsq]=Tvar[k];
8372: TvarsQind[nsq]=k;
8373: TvarV[ncovv]=Tvar[k];
8374: TvarVind[ncovv]=k;
1.231 brouard 8375: 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 */
8376: 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 8377: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8378: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8379: 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 8380: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8381: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 8382: ncova++;
8383: TvarA[ncova]=Tvar[k];
8384: TvarAind[ncova]=k;
1.231 brouard 8385: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.234 brouard 8386: Fixed[k]= 2;
8387: Dummy[k]= 2;
8388: modell[k].maintype= ATYPE;
8389: modell[k].subtype= APFD;
8390: /* ncoveff++; */
1.227 brouard 8391: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.234 brouard 8392: Fixed[k]= 2;
8393: Dummy[k]= 3;
8394: modell[k].maintype= ATYPE;
8395: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8396: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8397: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.234 brouard 8398: Fixed[k]= 3;
8399: Dummy[k]= 2;
8400: modell[k].maintype= ATYPE;
8401: modell[k].subtype= APVD; /* Product age * varying dummy */
8402: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8403: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.234 brouard 8404: Fixed[k]= 3;
8405: Dummy[k]= 3;
8406: modell[k].maintype= ATYPE;
8407: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8408: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8409: }
8410: }else if (Typevar[k] == 2) { /* product without age */
8411: k1=Tposprod[k];
8412: if(Tvard[k1][1] <=ncovcol){
1.234 brouard 8413: if(Tvard[k1][2] <=ncovcol){
8414: Fixed[k]= 1;
8415: Dummy[k]= 0;
8416: modell[k].maintype= FTYPE;
8417: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
8418: ncovf++; /* Fixed variables without age */
8419: TvarF[ncovf]=Tvar[k];
8420: TvarFind[ncovf]=k;
8421: }else if(Tvard[k1][2] <=ncovcol+nqv){
8422: Fixed[k]= 0; /* or 2 ?*/
8423: Dummy[k]= 1;
8424: modell[k].maintype= FTYPE;
8425: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
8426: ncovf++; /* Varying variables without age */
8427: TvarF[ncovf]=Tvar[k];
8428: TvarFind[ncovf]=k;
8429: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8430: Fixed[k]= 1;
8431: Dummy[k]= 0;
8432: modell[k].maintype= VTYPE;
8433: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
8434: ncovv++; /* Varying variables without age */
8435: TvarV[ncovv]=Tvar[k];
8436: TvarVind[ncovv]=k;
8437: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8438: Fixed[k]= 1;
8439: Dummy[k]= 1;
8440: modell[k].maintype= VTYPE;
8441: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
8442: ncovv++; /* Varying variables without age */
8443: TvarV[ncovv]=Tvar[k];
8444: TvarVind[ncovv]=k;
8445: }
1.227 brouard 8446: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.234 brouard 8447: if(Tvard[k1][2] <=ncovcol){
8448: Fixed[k]= 0; /* or 2 ?*/
8449: Dummy[k]= 1;
8450: modell[k].maintype= FTYPE;
8451: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
8452: ncovf++; /* Fixed variables without age */
8453: TvarF[ncovf]=Tvar[k];
8454: TvarFind[ncovf]=k;
8455: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8456: Fixed[k]= 1;
8457: Dummy[k]= 1;
8458: modell[k].maintype= VTYPE;
8459: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
8460: ncovv++; /* Varying variables without age */
8461: TvarV[ncovv]=Tvar[k];
8462: TvarVind[ncovv]=k;
8463: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8464: Fixed[k]= 1;
8465: Dummy[k]= 1;
8466: modell[k].maintype= VTYPE;
8467: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
8468: ncovv++; /* Varying variables without age */
8469: TvarV[ncovv]=Tvar[k];
8470: TvarVind[ncovv]=k;
8471: ncovv++; /* Varying variables without age */
8472: TvarV[ncovv]=Tvar[k];
8473: TvarVind[ncovv]=k;
8474: }
1.227 brouard 8475: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.234 brouard 8476: if(Tvard[k1][2] <=ncovcol){
8477: Fixed[k]= 1;
8478: Dummy[k]= 1;
8479: modell[k].maintype= VTYPE;
8480: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
8481: ncovv++; /* Varying variables without age */
8482: TvarV[ncovv]=Tvar[k];
8483: TvarVind[ncovv]=k;
8484: }else if(Tvard[k1][2] <=ncovcol+nqv){
8485: Fixed[k]= 1;
8486: Dummy[k]= 1;
8487: modell[k].maintype= VTYPE;
8488: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
8489: ncovv++; /* Varying variables without age */
8490: TvarV[ncovv]=Tvar[k];
8491: TvarVind[ncovv]=k;
8492: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8493: Fixed[k]= 1;
8494: Dummy[k]= 0;
8495: modell[k].maintype= VTYPE;
8496: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
8497: ncovv++; /* Varying variables without age */
8498: TvarV[ncovv]=Tvar[k];
8499: TvarVind[ncovv]=k;
8500: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8501: Fixed[k]= 1;
8502: Dummy[k]= 1;
8503: modell[k].maintype= VTYPE;
8504: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
8505: ncovv++; /* Varying variables without age */
8506: TvarV[ncovv]=Tvar[k];
8507: TvarVind[ncovv]=k;
8508: }
1.227 brouard 8509: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.234 brouard 8510: if(Tvard[k1][2] <=ncovcol){
8511: Fixed[k]= 1;
8512: Dummy[k]= 1;
8513: modell[k].maintype= VTYPE;
8514: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
8515: ncovv++; /* Varying variables without age */
8516: TvarV[ncovv]=Tvar[k];
8517: TvarVind[ncovv]=k;
8518: }else if(Tvard[k1][2] <=ncovcol+nqv){
8519: Fixed[k]= 1;
8520: Dummy[k]= 1;
8521: modell[k].maintype= VTYPE;
8522: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
8523: ncovv++; /* Varying variables without age */
8524: TvarV[ncovv]=Tvar[k];
8525: TvarVind[ncovv]=k;
8526: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8527: Fixed[k]= 1;
8528: Dummy[k]= 1;
8529: modell[k].maintype= VTYPE;
8530: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
8531: ncovv++; /* Varying variables without age */
8532: TvarV[ncovv]=Tvar[k];
8533: TvarVind[ncovv]=k;
8534: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8535: Fixed[k]= 1;
8536: Dummy[k]= 1;
8537: modell[k].maintype= VTYPE;
8538: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
8539: ncovv++; /* Varying variables without age */
8540: TvarV[ncovv]=Tvar[k];
8541: TvarVind[ncovv]=k;
8542: }
1.227 brouard 8543: }else{
1.234 brouard 8544: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8545: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
1.226 brouard 8546: } /* end k1 */
1.225 brouard 8547: }else{
1.226 brouard 8548: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
8549: 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 8550: }
1.227 brouard 8551: 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 8552: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 8553: 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]);
8554: }
8555: /* Searching for doublons in the model */
8556: for(k1=1; k1<= cptcovt;k1++){
8557: for(k2=1; k2 <k1;k2++){
8558: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 8559: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
8560: if(Tvar[k1]==Tvar[k2]){
8561: 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]]);
8562: 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);
8563: return(1);
8564: }
8565: }else if (Typevar[k1] ==2){
8566: k3=Tposprod[k1];
8567: k4=Tposprod[k2];
8568: 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])) ){
8569: 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]]);
8570: 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);
8571: return(1);
8572: }
8573: }
1.227 brouard 8574: }
8575: }
1.225 brouard 8576: }
8577: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
8578: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 8579: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
8580: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 8581: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 8582: /*endread:*/
1.225 brouard 8583: printf("Exiting decodemodel: ");
8584: return (1);
1.136 brouard 8585: }
8586:
1.169 brouard 8587: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.136 brouard 8588: {
8589: int i, m;
1.218 brouard 8590: int firstone=0;
8591:
1.136 brouard 8592: for (i=1; i<=imx; i++) {
8593: for(m=2; (m<= maxwav); m++) {
8594: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
8595: anint[m][i]=9999;
1.216 brouard 8596: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
8597: s[m][i]=-1;
1.136 brouard 8598: }
8599: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.169 brouard 8600: *nberr = *nberr + 1;
1.218 brouard 8601: if(firstone == 0){
8602: firstone=1;
8603: 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);
8604: }
8605: 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 8606: s[m][i]=-1;
8607: }
8608: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 8609: (*nberr)++;
1.136 brouard 8610: 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]);
8611: 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]);
8612: s[m][i]=-1; /* We prefer to skip it (and to skip it in version 0.8a1 too */
8613: }
8614: }
8615: }
8616:
8617: for (i=1; i<=imx; i++) {
8618: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
8619: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 8620: 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 8621: if (s[m][i] >= nlstate+1) {
1.169 brouard 8622: if(agedc[i]>0){
8623: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 8624: agev[m][i]=agedc[i];
1.214 brouard 8625: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 8626: }else {
1.136 brouard 8627: if ((int)andc[i]!=9999){
8628: nbwarn++;
8629: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
8630: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
8631: agev[m][i]=-1;
8632: }
8633: }
1.169 brouard 8634: } /* agedc > 0 */
1.214 brouard 8635: } /* end if */
1.136 brouard 8636: else if(s[m][i] !=9){ /* Standard case, age in fractional
8637: years but with the precision of a month */
8638: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
8639: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
8640: agev[m][i]=1;
8641: else if(agev[m][i] < *agemin){
8642: *agemin=agev[m][i];
8643: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
8644: }
8645: else if(agev[m][i] >*agemax){
8646: *agemax=agev[m][i];
1.156 brouard 8647: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 8648: }
8649: /*agev[m][i]=anint[m][i]-annais[i];*/
8650: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 8651: } /* en if 9*/
1.136 brouard 8652: else { /* =9 */
1.214 brouard 8653: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 8654: agev[m][i]=1;
8655: s[m][i]=-1;
8656: }
8657: }
1.214 brouard 8658: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 8659: agev[m][i]=1;
1.214 brouard 8660: else{
8661: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8662: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8663: agev[m][i]=0;
8664: }
8665: } /* End for lastpass */
8666: }
1.136 brouard 8667:
8668: for (i=1; i<=imx; i++) {
8669: for(m=firstpass; (m<=lastpass); m++){
8670: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 8671: (*nberr)++;
1.136 brouard 8672: 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);
8673: 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);
8674: return 1;
8675: }
8676: }
8677: }
8678:
8679: /*for (i=1; i<=imx; i++){
8680: for (m=firstpass; (m<lastpass); m++){
8681: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
8682: }
8683:
8684: }*/
8685:
8686:
1.139 brouard 8687: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
8688: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 8689:
8690: return (0);
1.164 brouard 8691: /* endread:*/
1.136 brouard 8692: printf("Exiting calandcheckages: ");
8693: return (1);
8694: }
8695:
1.172 brouard 8696: #if defined(_MSC_VER)
8697: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8698: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8699: //#include "stdafx.h"
8700: //#include <stdio.h>
8701: //#include <tchar.h>
8702: //#include <windows.h>
8703: //#include <iostream>
8704: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
8705:
8706: LPFN_ISWOW64PROCESS fnIsWow64Process;
8707:
8708: BOOL IsWow64()
8709: {
8710: BOOL bIsWow64 = FALSE;
8711:
8712: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
8713: // (HANDLE, PBOOL);
8714:
8715: //LPFN_ISWOW64PROCESS fnIsWow64Process;
8716:
8717: HMODULE module = GetModuleHandle(_T("kernel32"));
8718: const char funcName[] = "IsWow64Process";
8719: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
8720: GetProcAddress(module, funcName);
8721:
8722: if (NULL != fnIsWow64Process)
8723: {
8724: if (!fnIsWow64Process(GetCurrentProcess(),
8725: &bIsWow64))
8726: //throw std::exception("Unknown error");
8727: printf("Unknown error\n");
8728: }
8729: return bIsWow64 != FALSE;
8730: }
8731: #endif
1.177 brouard 8732:
1.191 brouard 8733: void syscompilerinfo(int logged)
1.167 brouard 8734: {
8735: /* #include "syscompilerinfo.h"*/
1.185 brouard 8736: /* command line Intel compiler 32bit windows, XP compatible:*/
8737: /* /GS /W3 /Gy
8738: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
8739: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
8740: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 8741: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
8742: */
8743: /* 64 bits */
1.185 brouard 8744: /*
8745: /GS /W3 /Gy
8746: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
8747: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
8748: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
8749: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
8750: /* Optimization are useless and O3 is slower than O2 */
8751: /*
8752: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
8753: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
8754: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
8755: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
8756: */
1.186 brouard 8757: /* Link is */ /* /OUT:"visual studio
1.185 brouard 8758: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
8759: /PDB:"visual studio
8760: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
8761: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
8762: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
8763: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
8764: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
8765: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
8766: uiAccess='false'"
8767: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
8768: /NOLOGO /TLBID:1
8769: */
1.177 brouard 8770: #if defined __INTEL_COMPILER
1.178 brouard 8771: #if defined(__GNUC__)
8772: struct utsname sysInfo; /* For Intel on Linux and OS/X */
8773: #endif
1.177 brouard 8774: #elif defined(__GNUC__)
1.179 brouard 8775: #ifndef __APPLE__
1.174 brouard 8776: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 8777: #endif
1.177 brouard 8778: struct utsname sysInfo;
1.178 brouard 8779: int cross = CROSS;
8780: if (cross){
8781: printf("Cross-");
1.191 brouard 8782: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 8783: }
1.174 brouard 8784: #endif
8785:
1.171 brouard 8786: #include <stdint.h>
1.178 brouard 8787:
1.191 brouard 8788: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 8789: #if defined(__clang__)
1.191 brouard 8790: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 8791: #endif
8792: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 8793: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 8794: #endif
8795: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 8796: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 8797: #endif
8798: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 8799: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 8800: #endif
8801: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 8802: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 8803: #endif
8804: #if defined(_MSC_VER)
1.191 brouard 8805: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 8806: #endif
8807: #if defined(__PGI)
1.191 brouard 8808: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 8809: #endif
8810: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 8811: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 8812: #endif
1.191 brouard 8813: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 8814:
1.167 brouard 8815: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
8816: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
8817: // Windows (x64 and x86)
1.191 brouard 8818: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 8819: #elif __unix__ // all unices, not all compilers
8820: // Unix
1.191 brouard 8821: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 8822: #elif __linux__
8823: // linux
1.191 brouard 8824: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 8825: #elif __APPLE__
1.174 brouard 8826: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 8827: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 8828: #endif
8829:
8830: /* __MINGW32__ */
8831: /* __CYGWIN__ */
8832: /* __MINGW64__ */
8833: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
8834: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
8835: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
8836: /* _WIN64 // Defined for applications for Win64. */
8837: /* _M_X64 // Defined for compilations that target x64 processors. */
8838: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 8839:
1.167 brouard 8840: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 8841: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 8842: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 8843: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 8844: #else
1.191 brouard 8845: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 8846: #endif
8847:
1.169 brouard 8848: #if defined(__GNUC__)
8849: # if defined(__GNUC_PATCHLEVEL__)
8850: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
8851: + __GNUC_MINOR__ * 100 \
8852: + __GNUC_PATCHLEVEL__)
8853: # else
8854: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
8855: + __GNUC_MINOR__ * 100)
8856: # endif
1.174 brouard 8857: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 8858: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 8859:
8860: if (uname(&sysInfo) != -1) {
8861: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 8862: 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 8863: }
8864: else
8865: perror("uname() error");
1.179 brouard 8866: //#ifndef __INTEL_COMPILER
8867: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 8868: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 8869: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 8870: #endif
1.169 brouard 8871: #endif
1.172 brouard 8872:
8873: // void main()
8874: // {
1.169 brouard 8875: #if defined(_MSC_VER)
1.174 brouard 8876: if (IsWow64()){
1.191 brouard 8877: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
8878: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 8879: }
8880: else{
1.191 brouard 8881: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
8882: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 8883: }
1.172 brouard 8884: // printf("\nPress Enter to continue...");
8885: // getchar();
8886: // }
8887:
1.169 brouard 8888: #endif
8889:
1.167 brouard 8890:
1.219 brouard 8891: }
1.136 brouard 8892:
1.219 brouard 8893: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 8894: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 8895: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 8896: /* double ftolpl = 1.e-10; */
1.180 brouard 8897: double age, agebase, agelim;
1.203 brouard 8898: double tot;
1.180 brouard 8899:
1.202 brouard 8900: strcpy(filerespl,"PL_");
8901: strcat(filerespl,fileresu);
8902: if((ficrespl=fopen(filerespl,"w"))==NULL) {
8903: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
8904: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
8905: }
1.227 brouard 8906: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
8907: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 8908: pstamp(ficrespl);
1.203 brouard 8909: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 8910: fprintf(ficrespl,"#Age ");
8911: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
8912: fprintf(ficrespl,"\n");
1.180 brouard 8913:
1.219 brouard 8914: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 8915:
1.219 brouard 8916: agebase=ageminpar;
8917: agelim=agemaxpar;
1.180 brouard 8918:
1.227 brouard 8919: /* i1=pow(2,ncoveff); */
1.234 brouard 8920: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 8921: if (cptcovn < 1){i1=1;}
1.180 brouard 8922:
1.235 brouard 8923: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.220 brouard 8924: for(k=1; k<=i1;k++){
1.235 brouard 8925: if(TKresult[nres]!= k)
8926: continue;
8927:
1.220 brouard 8928: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
1.180 brouard 8929: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
1.219 brouard 8930: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
1.220 brouard 8931: /* k=k+1; */
1.219 brouard 8932: /* to clean */
8933: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
8934: fprintf(ficrespl,"#******");
8935: printf("#******");
8936: fprintf(ficlog,"#******");
1.227 brouard 8937: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
8938: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
1.219 brouard 8939: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8940: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8941: }
1.235 brouard 8942: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
8943: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
8944: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
8945: }
1.219 brouard 8946: fprintf(ficrespl,"******\n");
8947: printf("******\n");
8948: fprintf(ficlog,"******\n");
1.227 brouard 8949: if(invalidvarcomb[k]){
8950: printf("\nCombination (%d) ignored because no case \n",k);
8951: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
8952: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
1.220 brouard 8953: continue;
1.227 brouard 8954: }
1.219 brouard 8955:
8956: fprintf(ficrespl,"#Age ");
1.227 brouard 8957: for(j=1;j<=cptcoveff;j++) {
1.219 brouard 8958: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8959: }
8960: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
8961: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 8962:
1.219 brouard 8963: for (age=agebase; age<=agelim; age++){
8964: /* for (age=agebase; age<=agebase; age++){ */
1.235 brouard 8965: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
1.219 brouard 8966: fprintf(ficrespl,"%.0f ",age );
1.227 brouard 8967: for(j=1;j<=cptcoveff;j++)
8968: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.219 brouard 8969: tot=0.;
8970: for(i=1; i<=nlstate;i++){
1.227 brouard 8971: tot += prlim[i][i];
8972: fprintf(ficrespl," %.5f", prlim[i][i]);
1.219 brouard 8973: }
8974: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
8975: } /* Age */
8976: /* was end of cptcod */
8977: } /* cptcov */
8978: return 0;
1.180 brouard 8979: }
8980:
1.218 brouard 8981: 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){
8982: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
8983:
8984: /* Computes the back prevalence limit for any combination of covariate values
8985: * at any age between ageminpar and agemaxpar
8986: */
1.235 brouard 8987: int i, j, k, i1, nres=0 ;
1.217 brouard 8988: /* double ftolpl = 1.e-10; */
8989: double age, agebase, agelim;
8990: double tot;
1.218 brouard 8991: /* double ***mobaverage; */
8992: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 8993:
8994: strcpy(fileresplb,"PLB_");
8995: strcat(fileresplb,fileresu);
8996: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
8997: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
8998: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
8999: }
9000: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9001: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9002: pstamp(ficresplb);
9003: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9004: fprintf(ficresplb,"#Age ");
9005: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9006: fprintf(ficresplb,"\n");
9007:
1.218 brouard 9008:
9009: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9010:
9011: agebase=ageminpar;
9012: agelim=agemaxpar;
9013:
9014:
1.227 brouard 9015: i1=pow(2,cptcoveff);
1.218 brouard 9016: if (cptcovn < 1){i1=1;}
1.227 brouard 9017:
1.235 brouard 9018: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9019: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
9020: if(TKresult[nres]!= k)
9021: continue;
1.218 brouard 9022: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9023: fprintf(ficresplb,"#******");
9024: printf("#******");
9025: fprintf(ficlog,"#******");
1.227 brouard 9026: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
1.218 brouard 9027: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9028: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9029: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9030: }
1.235 brouard 9031: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9032: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9033: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9034: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9035: }
1.218 brouard 9036: fprintf(ficresplb,"******\n");
9037: printf("******\n");
9038: fprintf(ficlog,"******\n");
1.227 brouard 9039: if(invalidvarcomb[k]){
9040: printf("\nCombination (%d) ignored because no cases \n",k);
9041: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9042: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9043: continue;
9044: }
1.218 brouard 9045:
9046: fprintf(ficresplb,"#Age ");
1.227 brouard 9047: for(j=1;j<=cptcoveff;j++) {
1.218 brouard 9048: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9049: }
9050: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9051: fprintf(ficresplb,"Total Years_to_converge\n");
9052:
9053:
9054: for (age=agebase; age<=agelim; age++){
9055: /* for (age=agebase; age<=agebase; age++){ */
9056: if(mobilavproj > 0){
9057: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9058: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.227 brouard 9059: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k);
1.218 brouard 9060: }else if (mobilavproj == 0){
1.227 brouard 9061: 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);
9062: 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);
9063: exit(1);
1.218 brouard 9064: }else{
1.227 brouard 9065: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
9066: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k);
1.218 brouard 9067: }
9068: fprintf(ficresplb,"%.0f ",age );
1.227 brouard 9069: for(j=1;j<=cptcoveff;j++)
9070: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.218 brouard 9071: tot=0.;
9072: for(i=1; i<=nlstate;i++){
1.227 brouard 9073: tot += bprlim[i][i];
9074: fprintf(ficresplb," %.5f", bprlim[i][i]);
1.218 brouard 9075: }
9076: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9077: } /* Age */
9078: /* was end of cptcod */
9079: } /* cptcov */
9080:
9081: /* hBijx(p, bage, fage); */
9082: /* fclose(ficrespijb); */
9083:
9084: return 0;
1.217 brouard 9085: }
1.218 brouard 9086:
1.180 brouard 9087: int hPijx(double *p, int bage, int fage){
9088: /*------------- h Pij x at various ages ------------*/
9089:
9090: int stepsize;
9091: int agelim;
9092: int hstepm;
9093: int nhstepm;
1.235 brouard 9094: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9095:
9096: double agedeb;
9097: double ***p3mat;
9098:
1.201 brouard 9099: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9100: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9101: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9102: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9103: }
9104: printf("Computing pij: result on file '%s' \n", filerespij);
9105: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9106:
9107: stepsize=(int) (stepm+YEARM-1)/YEARM;
9108: /*if (stepm<=24) stepsize=2;*/
9109:
9110: agelim=AGESUP;
9111: hstepm=stepsize*YEARM; /* Every year of age */
9112: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9113:
1.180 brouard 9114: /* hstepm=1; aff par mois*/
9115: pstamp(ficrespij);
9116: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9117: i1= pow(2,cptcoveff);
1.218 brouard 9118: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9119: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9120: /* k=k+1; */
1.235 brouard 9121: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9122: for(k=1; k<=i1;k++){
9123: if(TKresult[nres]!= k)
9124: continue;
1.183 brouard 9125: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9126: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9127: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9128: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9129: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9130: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9131: }
1.183 brouard 9132: fprintf(ficrespij,"******\n");
9133:
9134: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9135: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9136: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9137:
9138: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9139:
1.183 brouard 9140: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9141: oldm=oldms;savm=savms;
1.235 brouard 9142: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9143: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9144: for(i=1; i<=nlstate;i++)
9145: for(j=1; j<=nlstate+ndeath;j++)
9146: fprintf(ficrespij," %1d-%1d",i,j);
9147: fprintf(ficrespij,"\n");
9148: for (h=0; h<=nhstepm; h++){
9149: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9150: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9151: for(i=1; i<=nlstate;i++)
9152: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9153: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9154: fprintf(ficrespij,"\n");
9155: }
1.183 brouard 9156: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9157: fprintf(ficrespij,"\n");
9158: }
1.180 brouard 9159: /*}*/
9160: }
1.218 brouard 9161: return 0;
1.180 brouard 9162: }
1.218 brouard 9163:
9164: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9165: /*------------- h Bij x at various ages ------------*/
9166:
9167: int stepsize;
1.218 brouard 9168: /* int agelim; */
9169: int ageminl;
1.217 brouard 9170: int hstepm;
9171: int nhstepm;
9172: int h, i, i1, j, k;
1.218 brouard 9173:
1.217 brouard 9174: double agedeb;
9175: double ***p3mat;
1.218 brouard 9176:
9177: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9178: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9179: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9180: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9181: }
9182: printf("Computing pij back: result on file '%s' \n", filerespijb);
9183: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9184:
9185: stepsize=(int) (stepm+YEARM-1)/YEARM;
9186: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9187:
1.218 brouard 9188: /* agelim=AGESUP; */
9189: ageminl=30;
9190: hstepm=stepsize*YEARM; /* Every year of age */
9191: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9192:
9193: /* hstepm=1; aff par mois*/
9194: pstamp(ficrespijb);
9195: fprintf(ficrespijb,"#****** h Pij x Back Probability to be in state i at age x-h being in j at x ");
1.227 brouard 9196: i1= pow(2,cptcoveff);
1.218 brouard 9197: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9198: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9199: /* k=k+1; */
1.227 brouard 9200: for (k=1; k <= (int) pow(2,cptcoveff); k++){
1.218 brouard 9201: fprintf(ficrespijb,"\n#****** ");
1.227 brouard 9202: for(j=1;j<=cptcoveff;j++)
1.218 brouard 9203: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9204: fprintf(ficrespijb,"******\n");
1.222 brouard 9205: if(invalidvarcomb[k]){
9206: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9207: continue;
9208: }
1.218 brouard 9209:
9210: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9211: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9212: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9213: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9214: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9215:
9216: /* nhstepm=nhstepm*YEARM; aff par mois*/
9217:
9218: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9219: /* oldm=oldms;savm=savms; */
9220: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9221: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9222: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
9223: fprintf(ficrespijb,"# Cov Agex agex-h hpijx with i,j=");
9224: for(i=1; i<=nlstate;i++)
9225: for(j=1; j<=nlstate+ndeath;j++)
9226: fprintf(ficrespijb," %1d-%1d",i,j);
9227: fprintf(ficrespijb,"\n");
9228: for (h=0; h<=nhstepm; h++){
9229: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9230: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9231: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
1.217 brouard 9232: for(i=1; i<=nlstate;i++)
9233: for(j=1; j<=nlstate+ndeath;j++)
1.218 brouard 9234: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
1.217 brouard 9235: fprintf(ficrespijb,"\n");
9236: }
1.218 brouard 9237: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9238: fprintf(ficrespijb,"\n");
1.217 brouard 9239: }
1.218 brouard 9240: /*}*/
9241: }
9242: return 0;
9243: } /* hBijx */
1.217 brouard 9244:
1.180 brouard 9245:
1.136 brouard 9246: /***********************************************/
9247: /**************** Main Program *****************/
9248: /***********************************************/
9249:
9250: int main(int argc, char *argv[])
9251: {
9252: #ifdef GSL
9253: const gsl_multimin_fminimizer_type *T;
9254: size_t iteri = 0, it;
9255: int rval = GSL_CONTINUE;
9256: int status = GSL_SUCCESS;
9257: double ssval;
9258: #endif
9259: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9260: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9261: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9262: int jj, ll, li, lj, lk;
1.136 brouard 9263: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9264: int num_filled;
1.136 brouard 9265: int itimes;
9266: int NDIM=2;
9267: int vpopbased=0;
1.235 brouard 9268: int nres=0;
1.136 brouard 9269:
1.164 brouard 9270: char ca[32], cb[32];
1.136 brouard 9271: /* FILE *fichtm; *//* Html File */
9272: /* FILE *ficgp;*/ /*Gnuplot File */
9273: struct stat info;
1.191 brouard 9274: double agedeb=0.;
1.194 brouard 9275:
9276: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9277: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9278:
1.165 brouard 9279: double fret;
1.191 brouard 9280: double dum=0.; /* Dummy variable */
1.136 brouard 9281: double ***p3mat;
1.218 brouard 9282: /* double ***mobaverage; */
1.164 brouard 9283:
9284: char line[MAXLINE];
1.197 brouard 9285: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9286:
1.234 brouard 9287: char modeltemp[MAXLINE];
1.230 brouard 9288: char resultline[MAXLINE];
9289:
1.136 brouard 9290: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9291: char *tok, *val; /* pathtot */
1.136 brouard 9292: int firstobs=1, lastobs=10;
1.195 brouard 9293: int c, h , cpt, c2;
1.191 brouard 9294: int jl=0;
9295: int i1, j1, jk, stepsize=0;
1.194 brouard 9296: int count=0;
9297:
1.164 brouard 9298: int *tab;
1.136 brouard 9299: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9300: int backcast=0;
1.136 brouard 9301: int mobilav=0,popforecast=0;
1.191 brouard 9302: int hstepm=0, nhstepm=0;
1.136 brouard 9303: int agemortsup;
9304: float sumlpop=0.;
9305: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9306: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9307:
1.191 brouard 9308: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9309: double ftolpl=FTOL;
9310: double **prlim;
1.217 brouard 9311: double **bprlim;
1.136 brouard 9312: double ***param; /* Matrix of parameters */
9313: double *p;
9314: double **matcov; /* Matrix of covariance */
1.203 brouard 9315: double **hess; /* Hessian matrix */
1.136 brouard 9316: double ***delti3; /* Scale */
9317: double *delti; /* Scale */
9318: double ***eij, ***vareij;
9319: double **varpl; /* Variances of prevalence limits by age */
9320: double *epj, vepp;
1.164 brouard 9321:
1.136 brouard 9322: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9323: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9324:
1.136 brouard 9325: double **ximort;
1.145 brouard 9326: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9327: int *dcwave;
9328:
1.164 brouard 9329: char z[1]="c";
1.136 brouard 9330:
9331: /*char *strt;*/
9332: char strtend[80];
1.126 brouard 9333:
1.164 brouard 9334:
1.126 brouard 9335: /* setlocale (LC_ALL, ""); */
9336: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9337: /* textdomain (PACKAGE); */
9338: /* setlocale (LC_CTYPE, ""); */
9339: /* setlocale (LC_MESSAGES, ""); */
9340:
9341: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9342: rstart_time = time(NULL);
9343: /* (void) gettimeofday(&start_time,&tzp);*/
9344: start_time = *localtime(&rstart_time);
1.126 brouard 9345: curr_time=start_time;
1.157 brouard 9346: /*tml = *localtime(&start_time.tm_sec);*/
9347: /* strcpy(strstart,asctime(&tml)); */
9348: strcpy(strstart,asctime(&start_time));
1.126 brouard 9349:
9350: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9351: /* tp.tm_sec = tp.tm_sec +86400; */
9352: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9353: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9354: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9355: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9356: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9357: /* strt=asctime(&tmg); */
9358: /* printf("Time(after) =%s",strstart); */
9359: /* (void) time (&time_value);
9360: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9361: * tm = *localtime(&time_value);
9362: * strstart=asctime(&tm);
9363: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9364: */
9365:
9366: nberr=0; /* Number of errors and warnings */
9367: nbwarn=0;
1.184 brouard 9368: #ifdef WIN32
9369: _getcwd(pathcd, size);
9370: #else
1.126 brouard 9371: getcwd(pathcd, size);
1.184 brouard 9372: #endif
1.191 brouard 9373: syscompilerinfo(0);
1.196 brouard 9374: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9375: if(argc <=1){
9376: printf("\nEnter the parameter file name: ");
1.205 brouard 9377: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9378: printf("ERROR Empty parameter file name\n");
9379: goto end;
9380: }
1.126 brouard 9381: i=strlen(pathr);
9382: if(pathr[i-1]=='\n')
9383: pathr[i-1]='\0';
1.156 brouard 9384: i=strlen(pathr);
1.205 brouard 9385: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9386: pathr[i-1]='\0';
1.205 brouard 9387: }
9388: i=strlen(pathr);
9389: if( i==0 ){
9390: printf("ERROR Empty parameter file name\n");
9391: goto end;
9392: }
9393: for (tok = pathr; tok != NULL; ){
1.126 brouard 9394: printf("Pathr |%s|\n",pathr);
9395: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9396: printf("val= |%s| pathr=%s\n",val,pathr);
9397: strcpy (pathtot, val);
9398: if(pathr[0] == '\0') break; /* Dirty */
9399: }
9400: }
9401: else{
9402: strcpy(pathtot,argv[1]);
9403: }
9404: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9405: /*cygwin_split_path(pathtot,path,optionfile);
9406: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9407: /* cutv(path,optionfile,pathtot,'\\');*/
9408:
9409: /* Split argv[0], imach program to get pathimach */
9410: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9411: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9412: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9413: /* strcpy(pathimach,argv[0]); */
9414: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9415: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9416: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9417: #ifdef WIN32
9418: _chdir(path); /* Can be a relative path */
9419: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9420: #else
1.126 brouard 9421: chdir(path); /* Can be a relative path */
1.184 brouard 9422: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9423: #endif
9424: printf("Current directory %s!\n",pathcd);
1.126 brouard 9425: strcpy(command,"mkdir ");
9426: strcat(command,optionfilefiname);
9427: if((outcmd=system(command)) != 0){
1.169 brouard 9428: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9429: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9430: /* fclose(ficlog); */
9431: /* exit(1); */
9432: }
9433: /* if((imk=mkdir(optionfilefiname))<0){ */
9434: /* perror("mkdir"); */
9435: /* } */
9436:
9437: /*-------- arguments in the command line --------*/
9438:
1.186 brouard 9439: /* Main Log file */
1.126 brouard 9440: strcat(filelog, optionfilefiname);
9441: strcat(filelog,".log"); /* */
9442: if((ficlog=fopen(filelog,"w"))==NULL) {
9443: printf("Problem with logfile %s\n",filelog);
9444: goto end;
9445: }
9446: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9447: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9448: fprintf(ficlog,"\nEnter the parameter file name: \n");
9449: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9450: path=%s \n\
9451: optionfile=%s\n\
9452: optionfilext=%s\n\
1.156 brouard 9453: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9454:
1.197 brouard 9455: syscompilerinfo(1);
1.167 brouard 9456:
1.126 brouard 9457: printf("Local time (at start):%s",strstart);
9458: fprintf(ficlog,"Local time (at start): %s",strstart);
9459: fflush(ficlog);
9460: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9461: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9462:
9463: /* */
9464: strcpy(fileres,"r");
9465: strcat(fileres, optionfilefiname);
1.201 brouard 9466: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9467: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9468: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 9469:
1.186 brouard 9470: /* Main ---------arguments file --------*/
1.126 brouard 9471:
9472: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 9473: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
9474: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 9475: fflush(ficlog);
1.149 brouard 9476: /* goto end; */
9477: exit(70);
1.126 brouard 9478: }
9479:
9480:
9481:
9482: strcpy(filereso,"o");
1.201 brouard 9483: strcat(filereso,fileresu);
1.126 brouard 9484: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
9485: printf("Problem with Output resultfile: %s\n", filereso);
9486: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
9487: fflush(ficlog);
9488: goto end;
9489: }
9490:
9491: /* Reads comments: lines beginning with '#' */
9492: numlinepar=0;
1.197 brouard 9493:
9494: /* First parameter line */
9495: while(fgets(line, MAXLINE, ficpar)) {
9496: /* If line starts with a # it is a comment */
9497: if (line[0] == '#') {
9498: numlinepar++;
9499: fputs(line,stdout);
9500: fputs(line,ficparo);
9501: fputs(line,ficlog);
9502: continue;
9503: }else
9504: break;
9505: }
9506: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
9507: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
9508: if (num_filled != 5) {
9509: printf("Should be 5 parameters\n");
9510: }
1.126 brouard 9511: numlinepar++;
1.197 brouard 9512: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
9513: }
9514: /* Second parameter line */
9515: while(fgets(line, MAXLINE, ficpar)) {
9516: /* If line starts with a # it is a comment */
9517: if (line[0] == '#') {
9518: numlinepar++;
9519: fputs(line,stdout);
9520: fputs(line,ficparo);
9521: fputs(line,ficlog);
9522: continue;
9523: }else
9524: break;
9525: }
1.223 brouard 9526: 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", \
9527: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
9528: if (num_filled != 11) {
9529: 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 9530: printf("but line=%s\n",line);
1.197 brouard 9531: }
1.223 brouard 9532: 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 9533: }
1.203 brouard 9534: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 9535: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 9536: /* Third parameter line */
9537: while(fgets(line, MAXLINE, ficpar)) {
9538: /* If line starts with a # it is a comment */
9539: if (line[0] == '#') {
9540: numlinepar++;
9541: fputs(line,stdout);
9542: fputs(line,ficparo);
9543: fputs(line,ficlog);
9544: continue;
9545: }else
9546: break;
9547: }
1.201 brouard 9548: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
9549: if (num_filled == 0)
9550: model[0]='\0';
9551: else if (num_filled != 1){
1.197 brouard 9552: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9553: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9554: model[0]='\0';
9555: goto end;
9556: }
9557: else{
9558: if (model[0]=='+'){
9559: for(i=1; i<=strlen(model);i++)
9560: modeltemp[i-1]=model[i];
1.201 brouard 9561: strcpy(model,modeltemp);
1.197 brouard 9562: }
9563: }
1.199 brouard 9564: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 9565: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 9566: }
9567: /* 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); */
9568: /* numlinepar=numlinepar+3; /\* In general *\/ */
9569: /* 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 9570: 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);
9571: 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 9572: fflush(ficlog);
1.190 brouard 9573: /* if(model[0]=='#'|| model[0]== '\0'){ */
9574: if(model[0]=='#'){
1.187 brouard 9575: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
9576: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
9577: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
9578: if(mle != -1){
9579: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
9580: exit(1);
9581: }
9582: }
1.126 brouard 9583: while((c=getc(ficpar))=='#' && c!= EOF){
9584: ungetc(c,ficpar);
9585: fgets(line, MAXLINE, ficpar);
9586: numlinepar++;
1.195 brouard 9587: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
9588: z[0]=line[1];
9589: }
9590: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 9591: fputs(line, stdout);
9592: //puts(line);
1.126 brouard 9593: fputs(line,ficparo);
9594: fputs(line,ficlog);
9595: }
9596: ungetc(c,ficpar);
9597:
9598:
1.145 brouard 9599: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 9600: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 9601: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 9602: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 9603: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
9604: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
9605: v1+v2*age+v2*v3 makes cptcovn = 3
9606: */
9607: if (strlen(model)>1)
1.187 brouard 9608: 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 9609: else
1.187 brouard 9610: ncovmodel=2; /* Constant and age */
1.133 brouard 9611: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
9612: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 9613: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
9614: 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);
9615: 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);
9616: fflush(stdout);
9617: fclose (ficlog);
9618: goto end;
9619: }
1.126 brouard 9620: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9621: delti=delti3[1][1];
9622: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
9623: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
9624: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 9625: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
9626: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9627: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
9628: fclose (ficparo);
9629: fclose (ficlog);
9630: goto end;
9631: exit(0);
1.220 brouard 9632: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 9633: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 9634: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
9635: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9636: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9637: matcov=matrix(1,npar,1,npar);
1.203 brouard 9638: hess=matrix(1,npar,1,npar);
1.220 brouard 9639: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 9640: /* Read guessed parameters */
1.126 brouard 9641: /* Reads comments: lines beginning with '#' */
9642: while((c=getc(ficpar))=='#' && c!= EOF){
9643: ungetc(c,ficpar);
9644: fgets(line, MAXLINE, ficpar);
9645: numlinepar++;
1.141 brouard 9646: fputs(line,stdout);
1.126 brouard 9647: fputs(line,ficparo);
9648: fputs(line,ficlog);
9649: }
9650: ungetc(c,ficpar);
9651:
9652: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9653: for(i=1; i <=nlstate; i++){
1.234 brouard 9654: j=0;
1.126 brouard 9655: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 9656: if(jj==i) continue;
9657: j++;
9658: fscanf(ficpar,"%1d%1d",&i1,&j1);
9659: if ((i1 != i) || (j1 != jj)){
9660: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 9661: It might be a problem of design; if ncovcol and the model are correct\n \
9662: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 9663: exit(1);
9664: }
9665: fprintf(ficparo,"%1d%1d",i1,j1);
9666: if(mle==1)
9667: printf("%1d%1d",i,jj);
9668: fprintf(ficlog,"%1d%1d",i,jj);
9669: for(k=1; k<=ncovmodel;k++){
9670: fscanf(ficpar," %lf",¶m[i][j][k]);
9671: if(mle==1){
9672: printf(" %lf",param[i][j][k]);
9673: fprintf(ficlog," %lf",param[i][j][k]);
9674: }
9675: else
9676: fprintf(ficlog," %lf",param[i][j][k]);
9677: fprintf(ficparo," %lf",param[i][j][k]);
9678: }
9679: fscanf(ficpar,"\n");
9680: numlinepar++;
9681: if(mle==1)
9682: printf("\n");
9683: fprintf(ficlog,"\n");
9684: fprintf(ficparo,"\n");
1.126 brouard 9685: }
9686: }
9687: fflush(ficlog);
1.234 brouard 9688:
1.145 brouard 9689: /* Reads scales values */
1.126 brouard 9690: p=param[1][1];
9691:
9692: /* Reads comments: lines beginning with '#' */
9693: while((c=getc(ficpar))=='#' && c!= EOF){
9694: ungetc(c,ficpar);
9695: fgets(line, MAXLINE, ficpar);
9696: numlinepar++;
1.141 brouard 9697: fputs(line,stdout);
1.126 brouard 9698: fputs(line,ficparo);
9699: fputs(line,ficlog);
9700: }
9701: ungetc(c,ficpar);
9702:
9703: for(i=1; i <=nlstate; i++){
9704: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 9705: fscanf(ficpar,"%1d%1d",&i1,&j1);
9706: if ( (i1-i) * (j1-j) != 0){
9707: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
9708: exit(1);
9709: }
9710: printf("%1d%1d",i,j);
9711: fprintf(ficparo,"%1d%1d",i1,j1);
9712: fprintf(ficlog,"%1d%1d",i1,j1);
9713: for(k=1; k<=ncovmodel;k++){
9714: fscanf(ficpar,"%le",&delti3[i][j][k]);
9715: printf(" %le",delti3[i][j][k]);
9716: fprintf(ficparo," %le",delti3[i][j][k]);
9717: fprintf(ficlog," %le",delti3[i][j][k]);
9718: }
9719: fscanf(ficpar,"\n");
9720: numlinepar++;
9721: printf("\n");
9722: fprintf(ficparo,"\n");
9723: fprintf(ficlog,"\n");
1.126 brouard 9724: }
9725: }
9726: fflush(ficlog);
1.234 brouard 9727:
1.145 brouard 9728: /* Reads covariance matrix */
1.126 brouard 9729: delti=delti3[1][1];
1.220 brouard 9730:
9731:
1.126 brouard 9732: /* 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 9733:
1.126 brouard 9734: /* Reads comments: lines beginning with '#' */
9735: while((c=getc(ficpar))=='#' && c!= EOF){
9736: ungetc(c,ficpar);
9737: fgets(line, MAXLINE, ficpar);
9738: numlinepar++;
1.141 brouard 9739: fputs(line,stdout);
1.126 brouard 9740: fputs(line,ficparo);
9741: fputs(line,ficlog);
9742: }
9743: ungetc(c,ficpar);
1.220 brouard 9744:
1.126 brouard 9745: matcov=matrix(1,npar,1,npar);
1.203 brouard 9746: hess=matrix(1,npar,1,npar);
1.131 brouard 9747: for(i=1; i <=npar; i++)
9748: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 9749:
1.194 brouard 9750: /* Scans npar lines */
1.126 brouard 9751: for(i=1; i <=npar; i++){
1.226 brouard 9752: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 9753: if(count != 3){
1.226 brouard 9754: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 9755: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
9756: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 9757: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 9758: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
9759: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 9760: exit(1);
1.220 brouard 9761: }else{
1.226 brouard 9762: if(mle==1)
9763: printf("%1d%1d%d",i1,j1,jk);
9764: }
9765: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
9766: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 9767: for(j=1; j <=i; j++){
1.226 brouard 9768: fscanf(ficpar," %le",&matcov[i][j]);
9769: if(mle==1){
9770: printf(" %.5le",matcov[i][j]);
9771: }
9772: fprintf(ficlog," %.5le",matcov[i][j]);
9773: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 9774: }
9775: fscanf(ficpar,"\n");
9776: numlinepar++;
9777: if(mle==1)
1.220 brouard 9778: printf("\n");
1.126 brouard 9779: fprintf(ficlog,"\n");
9780: fprintf(ficparo,"\n");
9781: }
1.194 brouard 9782: /* End of read covariance matrix npar lines */
1.126 brouard 9783: for(i=1; i <=npar; i++)
9784: for(j=i+1;j<=npar;j++)
1.226 brouard 9785: matcov[i][j]=matcov[j][i];
1.126 brouard 9786:
9787: if(mle==1)
9788: printf("\n");
9789: fprintf(ficlog,"\n");
9790:
9791: fflush(ficlog);
9792:
9793: /*-------- Rewriting parameter file ----------*/
9794: strcpy(rfileres,"r"); /* "Rparameterfile */
9795: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
9796: strcat(rfileres,"."); /* */
9797: strcat(rfileres,optionfilext); /* Other files have txt extension */
9798: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 9799: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
9800: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 9801: }
9802: fprintf(ficres,"#%s\n",version);
9803: } /* End of mle != -3 */
1.218 brouard 9804:
1.186 brouard 9805: /* Main data
9806: */
1.126 brouard 9807: n= lastobs;
9808: num=lvector(1,n);
9809: moisnais=vector(1,n);
9810: annais=vector(1,n);
9811: moisdc=vector(1,n);
9812: andc=vector(1,n);
1.220 brouard 9813: weight=vector(1,n);
1.126 brouard 9814: agedc=vector(1,n);
9815: cod=ivector(1,n);
1.220 brouard 9816: for(i=1;i<=n;i++){
1.234 brouard 9817: num[i]=0;
9818: moisnais[i]=0;
9819: annais[i]=0;
9820: moisdc[i]=0;
9821: andc[i]=0;
9822: agedc[i]=0;
9823: cod[i]=0;
9824: weight[i]=1.0; /* Equal weights, 1 by default */
9825: }
1.126 brouard 9826: mint=matrix(1,maxwav,1,n);
9827: anint=matrix(1,maxwav,1,n);
1.131 brouard 9828: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 9829: tab=ivector(1,NCOVMAX);
1.144 brouard 9830: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 9831: 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 9832:
1.136 brouard 9833: /* Reads data from file datafile */
9834: if (readdata(datafile, firstobs, lastobs, &imx)==1)
9835: goto end;
9836:
9837: /* Calculation of the number of parameters from char model */
1.234 brouard 9838: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 9839: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
9840: k=3 V4 Tvar[k=3]= 4 (from V4)
9841: k=2 V1 Tvar[k=2]= 1 (from V1)
9842: k=1 Tvar[1]=2 (from V2)
1.234 brouard 9843: */
9844:
9845: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
9846: TvarsDind=ivector(1,NCOVMAX); /* */
9847: TvarsD=ivector(1,NCOVMAX); /* */
9848: TvarsQind=ivector(1,NCOVMAX); /* */
9849: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 9850: TvarF=ivector(1,NCOVMAX); /* */
9851: TvarFind=ivector(1,NCOVMAX); /* */
9852: TvarV=ivector(1,NCOVMAX); /* */
9853: TvarVind=ivector(1,NCOVMAX); /* */
9854: TvarA=ivector(1,NCOVMAX); /* */
9855: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 9856: TvarFD=ivector(1,NCOVMAX); /* */
9857: TvarFDind=ivector(1,NCOVMAX); /* */
9858: TvarFQ=ivector(1,NCOVMAX); /* */
9859: TvarFQind=ivector(1,NCOVMAX); /* */
9860: TvarVD=ivector(1,NCOVMAX); /* */
9861: TvarVDind=ivector(1,NCOVMAX); /* */
9862: TvarVQ=ivector(1,NCOVMAX); /* */
9863: TvarVQind=ivector(1,NCOVMAX); /* */
9864:
1.230 brouard 9865: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 9866: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 9867: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
9868: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
9869: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 9870: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
9871: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
9872: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
9873: */
9874: /* For model-covariate k tells which data-covariate to use but
9875: because this model-covariate is a construction we invent a new column
9876: ncovcol + k1
9877: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
9878: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 9879: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
9880: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 9881: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
9882: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 9883: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 9884: */
1.145 brouard 9885: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
9886: 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 9887: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
9888: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 9889: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 9890: 4 covariates (3 plus signs)
9891: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
9892: */
1.230 brouard 9893: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 9894: * individual dummy, fixed or varying:
9895: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
9896: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 9897: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
9898: * V1 df, V2 qf, V3 & V4 dv, V5 qv
9899: * Tmodelind[1]@9={9,0,3,2,}*/
9900: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
9901: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 9902: * individual quantitative, fixed or varying:
9903: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
9904: * 3, 1, 0, 0, 0, 0, 0, 0},
9905: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 9906: /* Main decodemodel */
9907:
1.187 brouard 9908:
1.223 brouard 9909: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 9910: goto end;
9911:
1.137 brouard 9912: if((double)(lastobs-imx)/(double)imx > 1.10){
9913: nbwarn++;
9914: 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);
9915: 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);
9916: }
1.136 brouard 9917: /* if(mle==1){*/
1.137 brouard 9918: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
9919: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 9920: }
9921:
9922: /*-calculation of age at interview from date of interview and age at death -*/
9923: agev=matrix(1,maxwav,1,imx);
9924:
9925: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
9926: goto end;
9927:
1.126 brouard 9928:
1.136 brouard 9929: agegomp=(int)agemin;
9930: free_vector(moisnais,1,n);
9931: free_vector(annais,1,n);
1.126 brouard 9932: /* free_matrix(mint,1,maxwav,1,n);
9933: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 9934: /* free_vector(moisdc,1,n); */
9935: /* free_vector(andc,1,n); */
1.145 brouard 9936: /* */
9937:
1.126 brouard 9938: wav=ivector(1,imx);
1.214 brouard 9939: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
9940: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
9941: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
9942: 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.*/
9943: bh=imatrix(1,lastpass-firstpass+2,1,imx);
9944: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 9945:
9946: /* Concatenates waves */
1.214 brouard 9947: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
9948: Death is a valid wave (if date is known).
9949: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
9950: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
9951: and mw[mi+1][i]. dh depends on stepm.
9952: */
9953:
1.126 brouard 9954: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.145 brouard 9955: /* */
9956:
1.215 brouard 9957: free_vector(moisdc,1,n);
9958: free_vector(andc,1,n);
9959:
1.126 brouard 9960: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
9961: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
9962: ncodemax[1]=1;
1.145 brouard 9963: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 9964: cptcoveff=0;
1.220 brouard 9965: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
9966: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 9967: }
9968:
9969: ncovcombmax=pow(2,cptcoveff);
9970: invalidvarcomb=ivector(1, ncovcombmax);
9971: for(i=1;i<ncovcombmax;i++)
9972: invalidvarcomb[i]=0;
9973:
1.211 brouard 9974: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 9975: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 9976: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 9977:
1.200 brouard 9978: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 9979: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 9980: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 9981: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
9982: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
9983: * (currently 0 or 1) in the data.
9984: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
9985: * corresponding modality (h,j).
9986: */
9987:
1.145 brouard 9988: h=0;
9989: /*if (cptcovn > 0) */
1.126 brouard 9990: m=pow(2,cptcoveff);
9991:
1.144 brouard 9992: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 9993: * For k=4 covariates, h goes from 1 to m=2**k
9994: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
9995: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 9996: * h\k 1 2 3 4
1.143 brouard 9997: *______________________________
9998: * 1 i=1 1 i=1 1 i=1 1 i=1 1
9999: * 2 2 1 1 1
10000: * 3 i=2 1 2 1 1
10001: * 4 2 2 1 1
10002: * 5 i=3 1 i=2 1 2 1
10003: * 6 2 1 2 1
10004: * 7 i=4 1 2 2 1
10005: * 8 2 2 2 1
1.197 brouard 10006: * 9 i=5 1 i=3 1 i=2 1 2
10007: * 10 2 1 1 2
10008: * 11 i=6 1 2 1 2
10009: * 12 2 2 1 2
10010: * 13 i=7 1 i=4 1 2 2
10011: * 14 2 1 2 2
10012: * 15 i=8 1 2 2 2
10013: * 16 2 2 2 2
1.143 brouard 10014: */
1.212 brouard 10015: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10016: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10017: * and the value of each covariate?
10018: * V1=1, V2=1, V3=2, V4=1 ?
10019: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10020: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10021: * In order to get the real value in the data, we use nbcode
10022: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10023: * We are keeping this crazy system in order to be able (in the future?)
10024: * to have more than 2 values (0 or 1) for a covariate.
10025: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10026: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10027: * bbbbbbbb
10028: * 76543210
10029: * h-1 00000101 (6-1=5)
1.219 brouard 10030: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10031: * &
10032: * 1 00000001 (1)
1.219 brouard 10033: * 00000000 = 1 & ((h-1) >> (k-1))
10034: * +1= 00000001 =1
1.211 brouard 10035: *
10036: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10037: * h' 1101 =2^3+2^2+0x2^1+2^0
10038: * >>k' 11
10039: * & 00000001
10040: * = 00000001
10041: * +1 = 00000010=2 = codtabm(14,3)
10042: * Reverse h=6 and m=16?
10043: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10044: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10045: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10046: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10047: * V3=decodtabm(14,3,2**4)=2
10048: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10049: *(h-1) >> (j-1) 0011 =13 >> 2
10050: * &1 000000001
10051: * = 000000001
10052: * +1= 000000010 =2
10053: * 2211
10054: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10055: * V3=2
1.220 brouard 10056: * codtabm and decodtabm are identical
1.211 brouard 10057: */
10058:
1.145 brouard 10059:
10060: free_ivector(Ndum,-1,NCOVMAX);
10061:
10062:
1.126 brouard 10063:
1.186 brouard 10064: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10065: strcpy(optionfilegnuplot,optionfilefiname);
10066: if(mle==-3)
1.201 brouard 10067: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10068: strcat(optionfilegnuplot,".gp");
10069:
10070: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10071: printf("Problem with file %s",optionfilegnuplot);
10072: }
10073: else{
1.204 brouard 10074: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10075: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10076: //fprintf(ficgp,"set missing 'NaNq'\n");
10077: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10078: }
10079: /* fclose(ficgp);*/
1.186 brouard 10080:
10081:
10082: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10083:
10084: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10085: if(mle==-3)
1.201 brouard 10086: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10087: strcat(optionfilehtm,".htm");
10088: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10089: printf("Problem with %s \n",optionfilehtm);
10090: exit(0);
1.126 brouard 10091: }
10092:
10093: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10094: strcat(optionfilehtmcov,"-cov.htm");
10095: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10096: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10097: }
10098: else{
10099: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10100: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10101: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10102: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10103: }
10104:
1.213 brouard 10105: 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 10106: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10107: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10108: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10109: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10110: \n\
10111: <hr size=\"2\" color=\"#EC5E5E\">\
10112: <ul><li><h4>Parameter files</h4>\n\
10113: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10114: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10115: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10116: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10117: - Date and time at start: %s</ul>\n",\
10118: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10119: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10120: fileres,fileres,\
10121: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10122: fflush(fichtm);
10123:
10124: strcpy(pathr,path);
10125: strcat(pathr,optionfilefiname);
1.184 brouard 10126: #ifdef WIN32
10127: _chdir(optionfilefiname); /* Move to directory named optionfile */
10128: #else
1.126 brouard 10129: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10130: #endif
10131:
1.126 brouard 10132:
1.220 brouard 10133: /* Calculates basic frequencies. Computes observed prevalence at single age
10134: and for any valid combination of covariates
1.126 brouard 10135: and prints on file fileres'p'. */
1.227 brouard 10136: freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
10137: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10138:
10139: fprintf(fichtm,"\n");
10140: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10141: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10142: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10143: imx,agemin,agemax,jmin,jmax,jmean);
10144: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10145: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10146: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10147: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10148: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10149:
1.126 brouard 10150: /* For Powell, parameters are in a vector p[] starting at p[1]
10151: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10152: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10153:
10154: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10155: /* For mortality only */
1.126 brouard 10156: if (mle==-3){
1.136 brouard 10157: ximort=matrix(1,NDIM,1,NDIM);
1.220 brouard 10158: for(i=1;i<=NDIM;i++)
10159: for(j=1;j<=NDIM;j++)
10160: ximort[i][j]=0.;
1.186 brouard 10161: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10162: cens=ivector(1,n);
10163: ageexmed=vector(1,n);
10164: agecens=vector(1,n);
10165: dcwave=ivector(1,n);
1.223 brouard 10166:
1.126 brouard 10167: for (i=1; i<=imx; i++){
10168: dcwave[i]=-1;
10169: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10170: if (s[m][i]>nlstate) {
10171: dcwave[i]=m;
10172: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10173: break;
10174: }
1.126 brouard 10175: }
1.226 brouard 10176:
1.126 brouard 10177: for (i=1; i<=imx; i++) {
10178: if (wav[i]>0){
1.226 brouard 10179: ageexmed[i]=agev[mw[1][i]][i];
10180: j=wav[i];
10181: agecens[i]=1.;
10182:
10183: if (ageexmed[i]> 1 && wav[i] > 0){
10184: agecens[i]=agev[mw[j][i]][i];
10185: cens[i]= 1;
10186: }else if (ageexmed[i]< 1)
10187: cens[i]= -1;
10188: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10189: cens[i]=0 ;
1.126 brouard 10190: }
10191: else cens[i]=-1;
10192: }
10193:
10194: for (i=1;i<=NDIM;i++) {
10195: for (j=1;j<=NDIM;j++)
1.226 brouard 10196: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10197: }
10198:
1.145 brouard 10199: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10200: /*printf("%lf %lf", p[1], p[2]);*/
10201:
10202:
1.136 brouard 10203: #ifdef GSL
10204: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10205: #else
1.126 brouard 10206: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10207: #endif
1.201 brouard 10208: strcpy(filerespow,"POW-MORT_");
10209: strcat(filerespow,fileresu);
1.126 brouard 10210: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10211: printf("Problem with resultfile: %s\n", filerespow);
10212: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10213: }
1.136 brouard 10214: #ifdef GSL
10215: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10216: #else
1.126 brouard 10217: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10218: #endif
1.126 brouard 10219: /* for (i=1;i<=nlstate;i++)
10220: for(j=1;j<=nlstate+ndeath;j++)
10221: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10222: */
10223: fprintf(ficrespow,"\n");
1.136 brouard 10224: #ifdef GSL
10225: /* gsl starts here */
10226: T = gsl_multimin_fminimizer_nmsimplex;
10227: gsl_multimin_fminimizer *sfm = NULL;
10228: gsl_vector *ss, *x;
10229: gsl_multimin_function minex_func;
10230:
10231: /* Initial vertex size vector */
10232: ss = gsl_vector_alloc (NDIM);
10233:
10234: if (ss == NULL){
10235: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10236: }
10237: /* Set all step sizes to 1 */
10238: gsl_vector_set_all (ss, 0.001);
10239:
10240: /* Starting point */
1.126 brouard 10241:
1.136 brouard 10242: x = gsl_vector_alloc (NDIM);
10243:
10244: if (x == NULL){
10245: gsl_vector_free(ss);
10246: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10247: }
10248:
10249: /* Initialize method and iterate */
10250: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10251: /* gsl_vector_set(x, 0, 0.0268); */
10252: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10253: gsl_vector_set(x, 0, p[1]);
10254: gsl_vector_set(x, 1, p[2]);
10255:
10256: minex_func.f = &gompertz_f;
10257: minex_func.n = NDIM;
10258: minex_func.params = (void *)&p; /* ??? */
10259:
10260: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10261: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10262:
10263: printf("Iterations beginning .....\n\n");
10264: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10265:
10266: iteri=0;
10267: while (rval == GSL_CONTINUE){
10268: iteri++;
10269: status = gsl_multimin_fminimizer_iterate(sfm);
10270:
10271: if (status) printf("error: %s\n", gsl_strerror (status));
10272: fflush(0);
10273:
10274: if (status)
10275: break;
10276:
10277: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10278: ssval = gsl_multimin_fminimizer_size (sfm);
10279:
10280: if (rval == GSL_SUCCESS)
10281: printf ("converged to a local maximum at\n");
10282:
10283: printf("%5d ", iteri);
10284: for (it = 0; it < NDIM; it++){
10285: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10286: }
10287: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10288: }
10289:
10290: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10291:
10292: gsl_vector_free(x); /* initial values */
10293: gsl_vector_free(ss); /* inital step size */
10294: for (it=0; it<NDIM; it++){
10295: p[it+1]=gsl_vector_get(sfm->x,it);
10296: fprintf(ficrespow," %.12lf", p[it]);
10297: }
10298: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10299: #endif
10300: #ifdef POWELL
10301: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10302: #endif
1.126 brouard 10303: fclose(ficrespow);
10304:
1.203 brouard 10305: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10306:
10307: for(i=1; i <=NDIM; i++)
10308: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10309: matcov[i][j]=matcov[j][i];
1.126 brouard 10310:
10311: printf("\nCovariance matrix\n ");
1.203 brouard 10312: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10313: for(i=1; i <=NDIM; i++) {
10314: for(j=1;j<=NDIM;j++){
1.220 brouard 10315: printf("%f ",matcov[i][j]);
10316: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10317: }
1.203 brouard 10318: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10319: }
10320:
10321: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10322: for (i=1;i<=NDIM;i++) {
1.126 brouard 10323: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10324: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10325: }
1.126 brouard 10326: lsurv=vector(1,AGESUP);
10327: lpop=vector(1,AGESUP);
10328: tpop=vector(1,AGESUP);
10329: lsurv[agegomp]=100000;
10330:
10331: for (k=agegomp;k<=AGESUP;k++) {
10332: agemortsup=k;
10333: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10334: }
10335:
10336: for (k=agegomp;k<agemortsup;k++)
10337: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10338:
10339: for (k=agegomp;k<agemortsup;k++){
10340: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10341: sumlpop=sumlpop+lpop[k];
10342: }
10343:
10344: tpop[agegomp]=sumlpop;
10345: for (k=agegomp;k<(agemortsup-3);k++){
10346: /* tpop[k+1]=2;*/
10347: tpop[k+1]=tpop[k]-lpop[k];
10348: }
10349:
10350:
10351: printf("\nAge lx qx dx Lx Tx e(x)\n");
10352: for (k=agegomp;k<(agemortsup-2);k++)
10353: 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]);
10354:
10355:
10356: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10357: ageminpar=50;
10358: agemaxpar=100;
1.194 brouard 10359: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10360: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10361: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10362: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10363: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10364: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10365: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10366: }else{
10367: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10368: 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 10369: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10370: }
1.201 brouard 10371: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10372: stepm, weightopt,\
10373: model,imx,p,matcov,agemortsup);
10374:
10375: free_vector(lsurv,1,AGESUP);
10376: free_vector(lpop,1,AGESUP);
10377: free_vector(tpop,1,AGESUP);
1.220 brouard 10378: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10379: free_ivector(cens,1,n);
10380: free_vector(agecens,1,n);
10381: free_ivector(dcwave,1,n);
1.220 brouard 10382: #ifdef GSL
1.136 brouard 10383: #endif
1.186 brouard 10384: } /* Endof if mle==-3 mortality only */
1.205 brouard 10385: /* Standard */
10386: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10387: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10388: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10389: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10390: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10391: for (k=1; k<=npar;k++)
10392: printf(" %d %8.5f",k,p[k]);
10393: printf("\n");
1.205 brouard 10394: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10395: /* mlikeli uses func not funcone */
10396: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10397: }
10398: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10399: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10400: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10401: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10402: }
10403: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10404: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10405: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10406: for (k=1; k<=npar;k++)
10407: printf(" %d %8.5f",k,p[k]);
10408: printf("\n");
10409:
10410: /*--------- results files --------------*/
1.224 brouard 10411: 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 10412:
10413:
10414: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10415: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10416: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10417: for(i=1,jk=1; i <=nlstate; i++){
10418: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10419: if (k != i) {
10420: printf("%d%d ",i,k);
10421: fprintf(ficlog,"%d%d ",i,k);
10422: fprintf(ficres,"%1d%1d ",i,k);
10423: for(j=1; j <=ncovmodel; j++){
10424: printf("%12.7f ",p[jk]);
10425: fprintf(ficlog,"%12.7f ",p[jk]);
10426: fprintf(ficres,"%12.7f ",p[jk]);
10427: jk++;
10428: }
10429: printf("\n");
10430: fprintf(ficlog,"\n");
10431: fprintf(ficres,"\n");
10432: }
1.126 brouard 10433: }
10434: }
1.203 brouard 10435: if(mle != 0){
10436: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10437: ftolhess=ftol; /* Usually correct */
1.203 brouard 10438: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10439: 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");
10440: 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");
10441: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10442: for(k=1; k <=(nlstate+ndeath); k++){
10443: if (k != i) {
10444: printf("%d%d ",i,k);
10445: fprintf(ficlog,"%d%d ",i,k);
10446: for(j=1; j <=ncovmodel; j++){
10447: 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]));
10448: 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]));
10449: jk++;
10450: }
10451: printf("\n");
10452: fprintf(ficlog,"\n");
10453: }
10454: }
1.193 brouard 10455: }
1.203 brouard 10456: } /* end of hesscov and Wald tests */
1.225 brouard 10457:
1.203 brouard 10458: /* */
1.126 brouard 10459: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
10460: printf("# Scales (for hessian or gradient estimation)\n");
10461: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
10462: for(i=1,jk=1; i <=nlstate; i++){
10463: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 10464: if (j!=i) {
10465: fprintf(ficres,"%1d%1d",i,j);
10466: printf("%1d%1d",i,j);
10467: fprintf(ficlog,"%1d%1d",i,j);
10468: for(k=1; k<=ncovmodel;k++){
10469: printf(" %.5e",delti[jk]);
10470: fprintf(ficlog," %.5e",delti[jk]);
10471: fprintf(ficres," %.5e",delti[jk]);
10472: jk++;
10473: }
10474: printf("\n");
10475: fprintf(ficlog,"\n");
10476: fprintf(ficres,"\n");
10477: }
1.126 brouard 10478: }
10479: }
10480:
10481: 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 10482: if(mle >= 1) /* To big for the screen */
1.126 brouard 10483: 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");
10484: 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");
10485: /* # 121 Var(a12)\n\ */
10486: /* # 122 Cov(b12,a12) Var(b12)\n\ */
10487: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
10488: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
10489: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
10490: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
10491: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
10492: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
10493:
10494:
10495: /* Just to have a covariance matrix which will be more understandable
10496: even is we still don't want to manage dictionary of variables
10497: */
10498: for(itimes=1;itimes<=2;itimes++){
10499: jj=0;
10500: for(i=1; i <=nlstate; i++){
1.225 brouard 10501: for(j=1; j <=nlstate+ndeath; j++){
10502: if(j==i) continue;
10503: for(k=1; k<=ncovmodel;k++){
10504: jj++;
10505: ca[0]= k+'a'-1;ca[1]='\0';
10506: if(itimes==1){
10507: if(mle>=1)
10508: printf("#%1d%1d%d",i,j,k);
10509: fprintf(ficlog,"#%1d%1d%d",i,j,k);
10510: fprintf(ficres,"#%1d%1d%d",i,j,k);
10511: }else{
10512: if(mle>=1)
10513: printf("%1d%1d%d",i,j,k);
10514: fprintf(ficlog,"%1d%1d%d",i,j,k);
10515: fprintf(ficres,"%1d%1d%d",i,j,k);
10516: }
10517: ll=0;
10518: for(li=1;li <=nlstate; li++){
10519: for(lj=1;lj <=nlstate+ndeath; lj++){
10520: if(lj==li) continue;
10521: for(lk=1;lk<=ncovmodel;lk++){
10522: ll++;
10523: if(ll<=jj){
10524: cb[0]= lk +'a'-1;cb[1]='\0';
10525: if(ll<jj){
10526: if(itimes==1){
10527: if(mle>=1)
10528: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10529: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10530: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10531: }else{
10532: if(mle>=1)
10533: printf(" %.5e",matcov[jj][ll]);
10534: fprintf(ficlog," %.5e",matcov[jj][ll]);
10535: fprintf(ficres," %.5e",matcov[jj][ll]);
10536: }
10537: }else{
10538: if(itimes==1){
10539: if(mle>=1)
10540: printf(" Var(%s%1d%1d)",ca,i,j);
10541: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
10542: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
10543: }else{
10544: if(mle>=1)
10545: printf(" %.7e",matcov[jj][ll]);
10546: fprintf(ficlog," %.7e",matcov[jj][ll]);
10547: fprintf(ficres," %.7e",matcov[jj][ll]);
10548: }
10549: }
10550: }
10551: } /* end lk */
10552: } /* end lj */
10553: } /* end li */
10554: if(mle>=1)
10555: printf("\n");
10556: fprintf(ficlog,"\n");
10557: fprintf(ficres,"\n");
10558: numlinepar++;
10559: } /* end k*/
10560: } /*end j */
1.126 brouard 10561: } /* end i */
10562: } /* end itimes */
10563:
10564: fflush(ficlog);
10565: fflush(ficres);
1.225 brouard 10566: while(fgets(line, MAXLINE, ficpar)) {
10567: /* If line starts with a # it is a comment */
10568: if (line[0] == '#') {
10569: numlinepar++;
10570: fputs(line,stdout);
10571: fputs(line,ficparo);
10572: fputs(line,ficlog);
10573: continue;
10574: }else
10575: break;
10576: }
10577:
1.209 brouard 10578: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
10579: /* ungetc(c,ficpar); */
10580: /* fgets(line, MAXLINE, ficpar); */
10581: /* fputs(line,stdout); */
10582: /* fputs(line,ficparo); */
10583: /* } */
10584: /* ungetc(c,ficpar); */
1.126 brouard 10585:
10586: estepm=0;
1.209 brouard 10587: 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 10588:
10589: if (num_filled != 6) {
10590: 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);
10591: 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);
10592: goto end;
10593: }
10594: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
10595: }
10596: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
10597: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
10598:
1.209 brouard 10599: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 10600: if (estepm==0 || estepm < stepm) estepm=stepm;
10601: if (fage <= 2) {
10602: bage = ageminpar;
10603: fage = agemaxpar;
10604: }
10605:
10606: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 10607: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
10608: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 10609:
1.186 brouard 10610: /* Other stuffs, more or less useful */
1.126 brouard 10611: while((c=getc(ficpar))=='#' && c!= EOF){
10612: ungetc(c,ficpar);
10613: fgets(line, MAXLINE, ficpar);
1.141 brouard 10614: fputs(line,stdout);
1.126 brouard 10615: fputs(line,ficparo);
10616: }
10617: ungetc(c,ficpar);
10618:
10619: 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);
10620: 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);
10621: 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);
10622: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
10623: 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);
10624:
10625: while((c=getc(ficpar))=='#' && c!= EOF){
10626: ungetc(c,ficpar);
10627: fgets(line, MAXLINE, ficpar);
1.141 brouard 10628: fputs(line,stdout);
1.126 brouard 10629: fputs(line,ficparo);
10630: }
10631: ungetc(c,ficpar);
10632:
10633:
10634: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
10635: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
10636:
10637: fscanf(ficpar,"pop_based=%d\n",&popbased);
1.193 brouard 10638: fprintf(ficlog,"pop_based=%d\n",popbased);
1.126 brouard 10639: fprintf(ficparo,"pop_based=%d\n",popbased);
10640: fprintf(ficres,"pop_based=%d\n",popbased);
10641:
10642: while((c=getc(ficpar))=='#' && c!= EOF){
10643: ungetc(c,ficpar);
10644: fgets(line, MAXLINE, ficpar);
1.141 brouard 10645: fputs(line,stdout);
1.126 brouard 10646: fputs(line,ficparo);
10647: }
10648: ungetc(c,ficpar);
10649:
10650: 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);
10651: 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);
10652: 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);
10653: 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);
10654: 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);
10655: /* day and month of proj2 are not used but only year anproj2.*/
10656:
1.217 brouard 10657: while((c=getc(ficpar))=='#' && c!= EOF){
10658: ungetc(c,ficpar);
10659: fgets(line, MAXLINE, ficpar);
10660: fputs(line,stdout);
10661: fputs(line,ficparo);
10662: }
10663: ungetc(c,ficpar);
10664:
10665: 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 10666: 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);
10667: 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);
10668: 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 10669: /* day and month of proj2 are not used but only year anproj2.*/
1.126 brouard 10670:
1.230 brouard 10671: /* Results */
1.235 brouard 10672: nresult=0;
1.230 brouard 10673: while(fgets(line, MAXLINE, ficpar)) {
10674: /* If line starts with a # it is a comment */
10675: if (line[0] == '#') {
10676: numlinepar++;
10677: fputs(line,stdout);
10678: fputs(line,ficparo);
10679: fputs(line,ficlog);
10680: continue;
10681: }else
10682: break;
10683: }
10684: while((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
10685: if (num_filled == 0)
10686: resultline[0]='\0';
10687: else if (num_filled != 1){
10688: printf("ERROR %d: result line should be at minimum 'result=' %s\n",num_filled, line);
10689: }
1.235 brouard 10690: nresult++; /* Sum of resultlines */
10691: printf("Result %d: result=%s\n",nresult, resultline);
10692: if(nresult > MAXRESULTLINES){
10693: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
10694: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
10695: goto end;
10696: }
10697: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
1.230 brouard 10698: while(fgets(line, MAXLINE, ficpar)) {
10699: /* If line starts with a # it is a comment */
10700: if (line[0] == '#') {
10701: numlinepar++;
10702: fputs(line,stdout);
10703: fputs(line,ficparo);
10704: fputs(line,ficlog);
10705: continue;
10706: }else
10707: break;
10708: }
10709: if (feof(ficpar))
10710: break;
10711: else{ /* Processess output results for this combination of covariate values */
10712: }
10713: }
10714:
10715:
1.126 brouard 10716:
1.230 brouard 10717: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 10718: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 10719:
10720: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 10721: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 10722: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 10723: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10724: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 10725: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 10726: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10727: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10728: }else{
1.218 brouard 10729: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 10730: }
10731: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.225 brouard 10732: model,imx,jmin,jmax,jmean,rfileres,popforecast,prevfcast,backcast, estepm, \
10733: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 10734:
1.225 brouard 10735: /*------------ free_vector -------------*/
10736: /* chdir(path); */
1.220 brouard 10737:
1.215 brouard 10738: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
10739: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
10740: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
10741: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 10742: free_lvector(num,1,n);
10743: free_vector(agedc,1,n);
10744: /*free_matrix(covar,0,NCOVMAX,1,n);*/
10745: /*free_matrix(covar,1,NCOVMAX,1,n);*/
10746: fclose(ficparo);
10747: fclose(ficres);
1.220 brouard 10748:
10749:
1.186 brouard 10750: /* Other results (useful)*/
1.220 brouard 10751:
10752:
1.126 brouard 10753: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 10754: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
10755: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 10756: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 10757: fclose(ficrespl);
10758:
10759: /*------------- h Pij x at various ages ------------*/
1.180 brouard 10760: /*#include "hpijx.h"*/
10761: hPijx(p, bage, fage);
1.145 brouard 10762: fclose(ficrespij);
1.227 brouard 10763:
1.220 brouard 10764: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 10765: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 10766: k=1;
1.126 brouard 10767: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 10768:
1.219 brouard 10769: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 10770: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 10771: for(i=1;i<=AGESUP;i++)
1.219 brouard 10772: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 10773: for(k=1;k<=ncovcombmax;k++)
10774: probs[i][j][k]=0.;
1.219 brouard 10775: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
10776: if (mobilav!=0 ||mobilavproj !=0 ) {
10777: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 10778: for(i=1;i<=AGESUP;i++)
10779: for(j=1;j<=nlstate;j++)
10780: for(k=1;k<=ncovcombmax;k++)
10781: mobaverages[i][j][k]=0.;
1.219 brouard 10782: mobaverage=mobaverages;
10783: if (mobilav!=0) {
1.235 brouard 10784: printf("Movingaveraging observed prevalence\n");
1.227 brouard 10785: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
10786: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
10787: printf(" Error in movingaverage mobilav=%d\n",mobilav);
10788: }
1.219 brouard 10789: }
10790: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
10791: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
10792: else if (mobilavproj !=0) {
1.235 brouard 10793: printf("Movingaveraging projected observed prevalence\n");
1.227 brouard 10794: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
10795: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
10796: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
10797: }
1.219 brouard 10798: }
10799: }/* end if moving average */
1.227 brouard 10800:
1.126 brouard 10801: /*---------- Forecasting ------------------*/
10802: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
10803: if(prevfcast==1){
10804: /* if(stepm ==1){*/
1.225 brouard 10805: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 10806: }
1.217 brouard 10807: if(backcast==1){
1.219 brouard 10808: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10809: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10810: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10811:
10812: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
10813:
10814: bprlim=matrix(1,nlstate,1,nlstate);
10815: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
10816: fclose(ficresplb);
10817:
1.222 brouard 10818: hBijx(p, bage, fage, mobaverage);
10819: fclose(ficrespijb);
1.219 brouard 10820: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
10821:
10822: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 10823: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 10824: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10825: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10826: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10827: }
1.217 brouard 10828:
1.186 brouard 10829:
10830: /* ------ Other prevalence ratios------------ */
1.126 brouard 10831:
1.215 brouard 10832: free_ivector(wav,1,imx);
10833: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
10834: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
10835: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 10836:
10837:
1.127 brouard 10838: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 10839:
1.201 brouard 10840: strcpy(filerese,"E_");
10841: strcat(filerese,fileresu);
1.126 brouard 10842: if((ficreseij=fopen(filerese,"w"))==NULL) {
10843: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
10844: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
10845: }
1.208 brouard 10846: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
10847: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.219 brouard 10848:
1.235 brouard 10849: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
10850: if (cptcovn < 1){i1=1;}
10851:
10852: for(nres=1; nres <= nresult; nres++) /* For each resultline */
10853: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
10854: if(TKresult[nres]!= k)
10855: continue;
1.219 brouard 10856: fprintf(ficreseij,"\n#****** ");
1.235 brouard 10857: printf("\n#****** ");
1.225 brouard 10858: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 10859: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 10860: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10861: }
10862: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
10863: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10864: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 10865: }
10866: fprintf(ficreseij,"******\n");
1.235 brouard 10867: printf("******\n");
1.219 brouard 10868:
10869: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10870: oldm=oldms;savm=savms;
1.235 brouard 10871: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 10872:
1.219 brouard 10873: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 10874: }
10875: fclose(ficreseij);
1.208 brouard 10876: printf("done evsij\n");fflush(stdout);
10877: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 10878:
1.227 brouard 10879: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 10880:
10881:
1.201 brouard 10882: strcpy(filerest,"T_");
10883: strcat(filerest,fileresu);
1.127 brouard 10884: if((ficrest=fopen(filerest,"w"))==NULL) {
10885: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
10886: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
10887: }
1.208 brouard 10888: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
10889: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 10890:
1.126 brouard 10891:
1.201 brouard 10892: strcpy(fileresstde,"STDE_");
10893: strcat(fileresstde,fileresu);
1.126 brouard 10894: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 10895: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
10896: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 10897: }
1.227 brouard 10898: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
10899: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 10900:
1.201 brouard 10901: strcpy(filerescve,"CVE_");
10902: strcat(filerescve,fileresu);
1.126 brouard 10903: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 10904: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
10905: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 10906: }
1.227 brouard 10907: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
10908: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 10909:
1.201 brouard 10910: strcpy(fileresv,"V_");
10911: strcat(fileresv,fileresu);
1.126 brouard 10912: if((ficresvij=fopen(fileresv,"w"))==NULL) {
10913: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
10914: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
10915: }
1.227 brouard 10916: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
10917: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 10918:
1.145 brouard 10919: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
10920: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
10921:
1.235 brouard 10922: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
10923: if (cptcovn < 1){i1=1;}
10924:
10925: for(nres=1; nres <= nresult; nres++) /* For each resultline */
10926: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
10927: if(TKresult[nres]!= k)
10928: continue;
10929: printf("\n#****** Selected:");
10930: fprintf(ficrest,"\n#****** Selected:");
10931: fprintf(ficlog,"\n#****** Selected:");
1.227 brouard 10932: for(j=1;j<=cptcoveff;j++){
10933: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10934: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10935: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10936: }
1.235 brouard 10937: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
10938: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10939: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10940: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10941: }
1.208 brouard 10942: fprintf(ficrest,"******\n");
1.227 brouard 10943: fprintf(ficlog,"******\n");
10944: printf("******\n");
1.208 brouard 10945:
10946: fprintf(ficresstdeij,"\n#****** ");
10947: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 10948: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 10949: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10950: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 10951: }
1.235 brouard 10952: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
10953: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10954: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10955: }
1.208 brouard 10956: fprintf(ficresstdeij,"******\n");
10957: fprintf(ficrescveij,"******\n");
10958:
10959: fprintf(ficresvij,"\n#****** ");
1.225 brouard 10960: for(j=1;j<=cptcoveff;j++)
1.227 brouard 10961: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 10962: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
10963: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
10964: }
1.208 brouard 10965: fprintf(ficresvij,"******\n");
10966:
10967: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10968: oldm=oldms;savm=savms;
1.235 brouard 10969: printf(" cvevsij ");
10970: fprintf(ficlog, " cvevsij ");
10971: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 10972: printf(" end cvevsij \n ");
10973: fprintf(ficlog, " end cvevsij \n ");
10974:
10975: /*
10976: */
10977: /* goto endfree; */
10978:
10979: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10980: pstamp(ficrest);
10981:
10982:
10983: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 10984: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
10985: cptcod= 0; /* To be deleted */
10986: printf("varevsij vpopbased=%d \n",vpopbased);
10987: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 10988: 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 10989: 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 ");
10990: if(vpopbased==1)
10991: 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);
10992: else
10993: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
10994: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
10995: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
10996: fprintf(ficrest,"\n");
10997: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
10998: epj=vector(1,nlstate+1);
10999: printf("Computing age specific period (stable) prevalences in each health state \n");
11000: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11001: for(age=bage; age <=fage ;age++){
1.235 brouard 11002: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11003: if (vpopbased==1) {
11004: if(mobilav ==0){
11005: for(i=1; i<=nlstate;i++)
11006: prlim[i][i]=probs[(int)age][i][k];
11007: }else{ /* mobilav */
11008: for(i=1; i<=nlstate;i++)
11009: prlim[i][i]=mobaverage[(int)age][i][k];
11010: }
11011: }
1.219 brouard 11012:
1.227 brouard 11013: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11014: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11015: /* printf(" age %4.0f ",age); */
11016: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11017: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11018: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11019: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11020: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11021: }
11022: epj[nlstate+1] +=epj[j];
11023: }
11024: /* printf(" age %4.0f \n",age); */
1.219 brouard 11025:
1.227 brouard 11026: for(i=1, vepp=0.;i <=nlstate;i++)
11027: for(j=1;j <=nlstate;j++)
11028: vepp += vareij[i][j][(int)age];
11029: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11030: for(j=1;j <=nlstate;j++){
11031: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11032: }
11033: fprintf(ficrest,"\n");
11034: }
1.208 brouard 11035: } /* End vpopbased */
11036: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11037: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11038: free_vector(epj,1,nlstate+1);
1.235 brouard 11039: printf("done selection\n");fflush(stdout);
11040: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11041:
1.145 brouard 11042: /*}*/
1.235 brouard 11043: } /* End k selection */
1.227 brouard 11044:
11045: printf("done State-specific expectancies\n");fflush(stdout);
11046: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11047:
1.126 brouard 11048: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11049:
1.201 brouard 11050: strcpy(fileresvpl,"VPL_");
11051: strcat(fileresvpl,fileresu);
1.126 brouard 11052: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11053: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11054: exit(0);
11055: }
1.208 brouard 11056: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11057: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11058:
1.145 brouard 11059: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11060: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11061:
1.235 brouard 11062: i1=pow(2,cptcoveff);
11063: if (cptcovn < 1){i1=1;}
11064:
11065: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11066: for(k=1; k<=i1;k++){
11067: if(TKresult[nres]!= k)
11068: continue;
1.227 brouard 11069: fprintf(ficresvpl,"\n#****** ");
11070: printf("\n#****** ");
11071: fprintf(ficlog,"\n#****** ");
11072: for(j=1;j<=cptcoveff;j++) {
11073: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11074: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11075: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11076: }
1.235 brouard 11077: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11078: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11079: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11080: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11081: }
1.227 brouard 11082: fprintf(ficresvpl,"******\n");
11083: printf("******\n");
11084: fprintf(ficlog,"******\n");
11085:
11086: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11087: oldm=oldms;savm=savms;
1.235 brouard 11088: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11089: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11090: /*}*/
1.126 brouard 11091: }
1.227 brouard 11092:
1.126 brouard 11093: fclose(ficresvpl);
1.208 brouard 11094: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11095: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11096:
11097: free_vector(weight,1,n);
11098: free_imatrix(Tvard,1,NCOVMAX,1,2);
11099: free_imatrix(s,1,maxwav+1,1,n);
11100: free_matrix(anint,1,maxwav,1,n);
11101: free_matrix(mint,1,maxwav,1,n);
11102: free_ivector(cod,1,n);
11103: free_ivector(tab,1,NCOVMAX);
11104: fclose(ficresstdeij);
11105: fclose(ficrescveij);
11106: fclose(ficresvij);
11107: fclose(ficrest);
11108: fclose(ficpar);
11109:
11110:
1.126 brouard 11111: /*---------- End : free ----------------*/
1.219 brouard 11112: if (mobilav!=0 ||mobilavproj !=0)
11113: 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 11114: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11115: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11116: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11117: } /* mle==-3 arrives here for freeing */
1.227 brouard 11118: /* endfree:*/
11119: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11120: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11121: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11122: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11123: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11124: free_matrix(coqvar,1,maxwav,1,n);
11125: free_matrix(covar,0,NCOVMAX,1,n);
11126: free_matrix(matcov,1,npar,1,npar);
11127: free_matrix(hess,1,npar,1,npar);
11128: /*free_vector(delti,1,npar);*/
11129: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11130: free_matrix(agev,1,maxwav,1,imx);
11131: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11132:
11133: free_ivector(ncodemax,1,NCOVMAX);
11134: free_ivector(ncodemaxwundef,1,NCOVMAX);
11135: free_ivector(Dummy,-1,NCOVMAX);
11136: free_ivector(Fixed,-1,NCOVMAX);
11137: free_ivector(Typevar,-1,NCOVMAX);
11138: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11139: free_ivector(TvarsQ,1,NCOVMAX);
11140: free_ivector(TvarsQind,1,NCOVMAX);
11141: free_ivector(TvarsD,1,NCOVMAX);
11142: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11143: free_ivector(TvarFD,1,NCOVMAX);
11144: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11145: free_ivector(TvarF,1,NCOVMAX);
11146: free_ivector(TvarFind,1,NCOVMAX);
11147: free_ivector(TvarV,1,NCOVMAX);
11148: free_ivector(TvarVind,1,NCOVMAX);
11149: free_ivector(TvarA,1,NCOVMAX);
11150: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11151: free_ivector(TvarFQ,1,NCOVMAX);
11152: free_ivector(TvarFQind,1,NCOVMAX);
11153: free_ivector(TvarVD,1,NCOVMAX);
11154: free_ivector(TvarVDind,1,NCOVMAX);
11155: free_ivector(TvarVQ,1,NCOVMAX);
11156: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11157: free_ivector(Tvarsel,1,NCOVMAX);
11158: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11159: free_ivector(Tposprod,1,NCOVMAX);
11160: free_ivector(Tprod,1,NCOVMAX);
11161: free_ivector(Tvaraff,1,NCOVMAX);
11162: free_ivector(invalidvarcomb,1,ncovcombmax);
11163: free_ivector(Tage,1,NCOVMAX);
11164: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11165: free_ivector(TmodelInvind,1,NCOVMAX);
11166: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11167:
11168: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11169: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11170: fflush(fichtm);
11171: fflush(ficgp);
11172:
1.227 brouard 11173:
1.126 brouard 11174: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11175: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11176: 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 11177: }else{
11178: printf("End of Imach\n");
11179: fprintf(ficlog,"End of Imach\n");
11180: }
11181: printf("See log file on %s\n",filelog);
11182: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11183: /*(void) gettimeofday(&end_time,&tzp);*/
11184: rend_time = time(NULL);
11185: end_time = *localtime(&rend_time);
11186: /* tml = *localtime(&end_time.tm_sec); */
11187: strcpy(strtend,asctime(&end_time));
1.126 brouard 11188: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11189: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11190: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11191:
1.157 brouard 11192: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11193: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11194: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11195: /* printf("Total time was %d uSec.\n", total_usecs);*/
11196: /* if(fileappend(fichtm,optionfilehtm)){ */
11197: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11198: fclose(fichtm);
11199: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11200: fclose(fichtmcov);
11201: fclose(ficgp);
11202: fclose(ficlog);
11203: /*------ End -----------*/
1.227 brouard 11204:
11205:
11206: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11207: #ifdef WIN32
1.227 brouard 11208: if (_chdir(pathcd) != 0)
11209: printf("Can't move to directory %s!\n",path);
11210: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11211: #else
1.227 brouard 11212: if(chdir(pathcd) != 0)
11213: printf("Can't move to directory %s!\n", path);
11214: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11215: #endif
1.126 brouard 11216: printf("Current directory %s!\n",pathcd);
11217: /*strcat(plotcmd,CHARSEPARATOR);*/
11218: sprintf(plotcmd,"gnuplot");
1.157 brouard 11219: #ifdef _WIN32
1.126 brouard 11220: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11221: #endif
11222: if(!stat(plotcmd,&info)){
1.158 brouard 11223: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11224: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11225: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11226: }else
11227: strcpy(pplotcmd,plotcmd);
1.157 brouard 11228: #ifdef __unix
1.126 brouard 11229: strcpy(plotcmd,GNUPLOTPROGRAM);
11230: if(!stat(plotcmd,&info)){
1.158 brouard 11231: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11232: }else
11233: strcpy(pplotcmd,plotcmd);
11234: #endif
11235: }else
11236: strcpy(pplotcmd,plotcmd);
11237:
11238: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 11239: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 11240:
1.126 brouard 11241: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 11242: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 11243: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 11244: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 11245: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 11246: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 11247: }
1.158 brouard 11248: printf(" Successful, please wait...");
1.126 brouard 11249: while (z[0] != 'q') {
11250: /* chdir(path); */
1.154 brouard 11251: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 11252: scanf("%s",z);
11253: /* if (z[0] == 'c') system("./imach"); */
11254: if (z[0] == 'e') {
1.158 brouard 11255: #ifdef __APPLE__
1.152 brouard 11256: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 11257: #elif __linux
11258: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 11259: #else
1.152 brouard 11260: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 11261: #endif
11262: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
11263: system(pplotcmd);
1.126 brouard 11264: }
11265: else if (z[0] == 'g') system(plotcmd);
11266: else if (z[0] == 'q') exit(0);
11267: }
1.227 brouard 11268: end:
1.126 brouard 11269: while (z[0] != 'q') {
1.195 brouard 11270: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 11271: scanf("%s",z);
11272: }
11273: }
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