Annotation of imach/src/imach.c, revision 1.233
1.233 ! brouard 1: /* $Id: imach.c,v 1.232 2016/08/22 14:20:21 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.233 ! brouard 4: Revision 1.232 2016/08/22 14:20:21 brouard
! 5: Summary: not working
! 6:
1.232 brouard 7: Revision 1.231 2016/08/22 07:17:15 brouard
8: Summary: not working
9:
1.231 brouard 10: Revision 1.230 2016/08/22 06:55:53 brouard
11: Summary: Not working
12:
1.230 brouard 13: Revision 1.229 2016/07/23 09:45:53 brouard
14: Summary: Completing for func too
15:
1.229 brouard 16: Revision 1.228 2016/07/22 17:45:30 brouard
17: Summary: Fixing some arrays, still debugging
18:
1.227 brouard 19: Revision 1.226 2016/07/12 18:42:34 brouard
20: Summary: temp
21:
1.226 brouard 22: Revision 1.225 2016/07/12 08:40:03 brouard
23: Summary: saving but not running
24:
1.225 brouard 25: Revision 1.224 2016/07/01 13:16:01 brouard
26: Summary: Fixes
27:
1.224 brouard 28: Revision 1.223 2016/02/19 09:23:35 brouard
29: Summary: temporary
30:
1.223 brouard 31: Revision 1.222 2016/02/17 08:14:50 brouard
32: Summary: Probably last 0.98 stable version 0.98r6
33:
1.222 brouard 34: Revision 1.221 2016/02/15 23:35:36 brouard
35: Summary: minor bug
36:
1.220 brouard 37: Revision 1.219 2016/02/15 00:48:12 brouard
38: *** empty log message ***
39:
1.219 brouard 40: Revision 1.218 2016/02/12 11:29:23 brouard
41: Summary: 0.99 Back projections
42:
1.218 brouard 43: Revision 1.217 2015/12/23 17:18:31 brouard
44: Summary: Experimental backcast
45:
1.217 brouard 46: Revision 1.216 2015/12/18 17:32:11 brouard
47: Summary: 0.98r4 Warning and status=-2
48:
49: Version 0.98r4 is now:
50: - displaying an error when status is -1, date of interview unknown and date of death known;
51: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
52: Older changes concerning s=-2, dating from 2005 have been supersed.
53:
1.216 brouard 54: Revision 1.215 2015/12/16 08:52:24 brouard
55: Summary: 0.98r4 working
56:
1.215 brouard 57: Revision 1.214 2015/12/16 06:57:54 brouard
58: Summary: temporary not working
59:
1.214 brouard 60: Revision 1.213 2015/12/11 18:22:17 brouard
61: Summary: 0.98r4
62:
1.213 brouard 63: Revision 1.212 2015/11/21 12:47:24 brouard
64: Summary: minor typo
65:
1.212 brouard 66: Revision 1.211 2015/11/21 12:41:11 brouard
67: Summary: 0.98r3 with some graph of projected cross-sectional
68:
69: Author: Nicolas Brouard
70:
1.211 brouard 71: Revision 1.210 2015/11/18 17:41:20 brouard
72: Summary: Start working on projected prevalences
73:
1.210 brouard 74: Revision 1.209 2015/11/17 22:12:03 brouard
75: Summary: Adding ftolpl parameter
76: Author: N Brouard
77:
78: We had difficulties to get smoothed confidence intervals. It was due
79: to the period prevalence which wasn't computed accurately. The inner
80: parameter ftolpl is now an outer parameter of the .imach parameter
81: file after estepm. If ftolpl is small 1.e-4 and estepm too,
82: computation are long.
83:
1.209 brouard 84: Revision 1.208 2015/11/17 14:31:57 brouard
85: Summary: temporary
86:
1.208 brouard 87: Revision 1.207 2015/10/27 17:36:57 brouard
88: *** empty log message ***
89:
1.207 brouard 90: Revision 1.206 2015/10/24 07:14:11 brouard
91: *** empty log message ***
92:
1.206 brouard 93: Revision 1.205 2015/10/23 15:50:53 brouard
94: Summary: 0.98r3 some clarification for graphs on likelihood contributions
95:
1.205 brouard 96: Revision 1.204 2015/10/01 16:20:26 brouard
97: Summary: Some new graphs of contribution to likelihood
98:
1.204 brouard 99: Revision 1.203 2015/09/30 17:45:14 brouard
100: Summary: looking at better estimation of the hessian
101:
102: Also a better criteria for convergence to the period prevalence And
103: therefore adding the number of years needed to converge. (The
104: prevalence in any alive state shold sum to one
105:
1.203 brouard 106: Revision 1.202 2015/09/22 19:45:16 brouard
107: Summary: Adding some overall graph on contribution to likelihood. Might change
108:
1.202 brouard 109: Revision 1.201 2015/09/15 17:34:58 brouard
110: Summary: 0.98r0
111:
112: - Some new graphs like suvival functions
113: - Some bugs fixed like model=1+age+V2.
114:
1.201 brouard 115: Revision 1.200 2015/09/09 16:53:55 brouard
116: Summary: Big bug thanks to Flavia
117:
118: Even model=1+age+V2. did not work anymore
119:
1.200 brouard 120: Revision 1.199 2015/09/07 14:09:23 brouard
121: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
122:
1.199 brouard 123: Revision 1.198 2015/09/03 07:14:39 brouard
124: Summary: 0.98q5 Flavia
125:
1.198 brouard 126: Revision 1.197 2015/09/01 18:24:39 brouard
127: *** empty log message ***
128:
1.197 brouard 129: Revision 1.196 2015/08/18 23:17:52 brouard
130: Summary: 0.98q5
131:
1.196 brouard 132: Revision 1.195 2015/08/18 16:28:39 brouard
133: Summary: Adding a hack for testing purpose
134:
135: After reading the title, ftol and model lines, if the comment line has
136: a q, starting with #q, the answer at the end of the run is quit. It
137: permits to run test files in batch with ctest. The former workaround was
138: $ echo q | imach foo.imach
139:
1.195 brouard 140: Revision 1.194 2015/08/18 13:32:00 brouard
141: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
142:
1.194 brouard 143: Revision 1.193 2015/08/04 07:17:42 brouard
144: Summary: 0.98q4
145:
1.193 brouard 146: Revision 1.192 2015/07/16 16:49:02 brouard
147: Summary: Fixing some outputs
148:
1.192 brouard 149: Revision 1.191 2015/07/14 10:00:33 brouard
150: Summary: Some fixes
151:
1.191 brouard 152: Revision 1.190 2015/05/05 08:51:13 brouard
153: Summary: Adding digits in output parameters (7 digits instead of 6)
154:
155: Fix 1+age+.
156:
1.190 brouard 157: Revision 1.189 2015/04/30 14:45:16 brouard
158: Summary: 0.98q2
159:
1.189 brouard 160: Revision 1.188 2015/04/30 08:27:53 brouard
161: *** empty log message ***
162:
1.188 brouard 163: Revision 1.187 2015/04/29 09:11:15 brouard
164: *** empty log message ***
165:
1.187 brouard 166: Revision 1.186 2015/04/23 12:01:52 brouard
167: Summary: V1*age is working now, version 0.98q1
168:
169: Some codes had been disabled in order to simplify and Vn*age was
170: working in the optimization phase, ie, giving correct MLE parameters,
171: but, as usual, outputs were not correct and program core dumped.
172:
1.186 brouard 173: Revision 1.185 2015/03/11 13:26:42 brouard
174: Summary: Inclusion of compile and links command line for Intel Compiler
175:
1.185 brouard 176: Revision 1.184 2015/03/11 11:52:39 brouard
177: Summary: Back from Windows 8. Intel Compiler
178:
1.184 brouard 179: Revision 1.183 2015/03/10 20:34:32 brouard
180: Summary: 0.98q0, trying with directest, mnbrak fixed
181:
182: We use directest instead of original Powell test; probably no
183: incidence on the results, but better justifications;
184: We fixed Numerical Recipes mnbrak routine which was wrong and gave
185: wrong results.
186:
1.183 brouard 187: Revision 1.182 2015/02/12 08:19:57 brouard
188: Summary: Trying to keep directest which seems simpler and more general
189: Author: Nicolas Brouard
190:
1.182 brouard 191: Revision 1.181 2015/02/11 23:22:24 brouard
192: Summary: Comments on Powell added
193:
194: Author:
195:
1.181 brouard 196: Revision 1.180 2015/02/11 17:33:45 brouard
197: Summary: Finishing move from main to function (hpijx and prevalence_limit)
198:
1.180 brouard 199: Revision 1.179 2015/01/04 09:57:06 brouard
200: Summary: back to OS/X
201:
1.179 brouard 202: Revision 1.178 2015/01/04 09:35:48 brouard
203: *** empty log message ***
204:
1.178 brouard 205: Revision 1.177 2015/01/03 18:40:56 brouard
206: Summary: Still testing ilc32 on OSX
207:
1.177 brouard 208: Revision 1.176 2015/01/03 16:45:04 brouard
209: *** empty log message ***
210:
1.176 brouard 211: Revision 1.175 2015/01/03 16:33:42 brouard
212: *** empty log message ***
213:
1.175 brouard 214: Revision 1.174 2015/01/03 16:15:49 brouard
215: Summary: Still in cross-compilation
216:
1.174 brouard 217: Revision 1.173 2015/01/03 12:06:26 brouard
218: Summary: trying to detect cross-compilation
219:
1.173 brouard 220: Revision 1.172 2014/12/27 12:07:47 brouard
221: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
222:
1.172 brouard 223: Revision 1.171 2014/12/23 13:26:59 brouard
224: Summary: Back from Visual C
225:
226: Still problem with utsname.h on Windows
227:
1.171 brouard 228: Revision 1.170 2014/12/23 11:17:12 brouard
229: Summary: Cleaning some \%% back to %%
230:
231: The escape was mandatory for a specific compiler (which one?), but too many warnings.
232:
1.170 brouard 233: Revision 1.169 2014/12/22 23:08:31 brouard
234: Summary: 0.98p
235:
236: Outputs some informations on compiler used, OS etc. Testing on different platforms.
237:
1.169 brouard 238: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 239: Summary: update
1.169 brouard 240:
1.168 brouard 241: Revision 1.167 2014/12/22 13:50:56 brouard
242: Summary: Testing uname and compiler version and if compiled 32 or 64
243:
244: Testing on Linux 64
245:
1.167 brouard 246: Revision 1.166 2014/12/22 11:40:47 brouard
247: *** empty log message ***
248:
1.166 brouard 249: Revision 1.165 2014/12/16 11:20:36 brouard
250: Summary: After compiling on Visual C
251:
252: * imach.c (Module): Merging 1.61 to 1.162
253:
1.165 brouard 254: Revision 1.164 2014/12/16 10:52:11 brouard
255: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
256:
257: * imach.c (Module): Merging 1.61 to 1.162
258:
1.164 brouard 259: Revision 1.163 2014/12/16 10:30:11 brouard
260: * imach.c (Module): Merging 1.61 to 1.162
261:
1.163 brouard 262: Revision 1.162 2014/09/25 11:43:39 brouard
263: Summary: temporary backup 0.99!
264:
1.162 brouard 265: Revision 1.1 2014/09/16 11:06:58 brouard
266: Summary: With some code (wrong) for nlopt
267:
268: Author:
269:
270: Revision 1.161 2014/09/15 20:41:41 brouard
271: Summary: Problem with macro SQR on Intel compiler
272:
1.161 brouard 273: Revision 1.160 2014/09/02 09:24:05 brouard
274: *** empty log message ***
275:
1.160 brouard 276: Revision 1.159 2014/09/01 10:34:10 brouard
277: Summary: WIN32
278: Author: Brouard
279:
1.159 brouard 280: Revision 1.158 2014/08/27 17:11:51 brouard
281: *** empty log message ***
282:
1.158 brouard 283: Revision 1.157 2014/08/27 16:26:55 brouard
284: Summary: Preparing windows Visual studio version
285: Author: Brouard
286:
287: In order to compile on Visual studio, time.h is now correct and time_t
288: and tm struct should be used. difftime should be used but sometimes I
289: just make the differences in raw time format (time(&now).
290: Trying to suppress #ifdef LINUX
291: Add xdg-open for __linux in order to open default browser.
292:
1.157 brouard 293: Revision 1.156 2014/08/25 20:10:10 brouard
294: *** empty log message ***
295:
1.156 brouard 296: Revision 1.155 2014/08/25 18:32:34 brouard
297: Summary: New compile, minor changes
298: Author: Brouard
299:
1.155 brouard 300: Revision 1.154 2014/06/20 17:32:08 brouard
301: Summary: Outputs now all graphs of convergence to period prevalence
302:
1.154 brouard 303: Revision 1.153 2014/06/20 16:45:46 brouard
304: Summary: If 3 live state, convergence to period prevalence on same graph
305: Author: Brouard
306:
1.153 brouard 307: Revision 1.152 2014/06/18 17:54:09 brouard
308: Summary: open browser, use gnuplot on same dir than imach if not found in the path
309:
1.152 brouard 310: Revision 1.151 2014/06/18 16:43:30 brouard
311: *** empty log message ***
312:
1.151 brouard 313: Revision 1.150 2014/06/18 16:42:35 brouard
314: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
315: Author: brouard
316:
1.150 brouard 317: Revision 1.149 2014/06/18 15:51:14 brouard
318: Summary: Some fixes in parameter files errors
319: Author: Nicolas Brouard
320:
1.149 brouard 321: Revision 1.148 2014/06/17 17:38:48 brouard
322: Summary: Nothing new
323: Author: Brouard
324:
325: Just a new packaging for OS/X version 0.98nS
326:
1.148 brouard 327: Revision 1.147 2014/06/16 10:33:11 brouard
328: *** empty log message ***
329:
1.147 brouard 330: Revision 1.146 2014/06/16 10:20:28 brouard
331: Summary: Merge
332: Author: Brouard
333:
334: Merge, before building revised version.
335:
1.146 brouard 336: Revision 1.145 2014/06/10 21:23:15 brouard
337: Summary: Debugging with valgrind
338: Author: Nicolas Brouard
339:
340: Lot of changes in order to output the results with some covariates
341: After the Edimburgh REVES conference 2014, it seems mandatory to
342: improve the code.
343: No more memory valgrind error but a lot has to be done in order to
344: continue the work of splitting the code into subroutines.
345: Also, decodemodel has been improved. Tricode is still not
346: optimal. nbcode should be improved. Documentation has been added in
347: the source code.
348:
1.144 brouard 349: Revision 1.143 2014/01/26 09:45:38 brouard
350: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
351:
352: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
353: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
354:
1.143 brouard 355: Revision 1.142 2014/01/26 03:57:36 brouard
356: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
357:
358: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
359:
1.142 brouard 360: Revision 1.141 2014/01/26 02:42:01 brouard
361: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
362:
1.141 brouard 363: Revision 1.140 2011/09/02 10:37:54 brouard
364: Summary: times.h is ok with mingw32 now.
365:
1.140 brouard 366: Revision 1.139 2010/06/14 07:50:17 brouard
367: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
368: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
369:
1.139 brouard 370: Revision 1.138 2010/04/30 18:19:40 brouard
371: *** empty log message ***
372:
1.138 brouard 373: Revision 1.137 2010/04/29 18:11:38 brouard
374: (Module): Checking covariates for more complex models
375: than V1+V2. A lot of change to be done. Unstable.
376:
1.137 brouard 377: Revision 1.136 2010/04/26 20:30:53 brouard
378: (Module): merging some libgsl code. Fixing computation
379: of likelione (using inter/intrapolation if mle = 0) in order to
380: get same likelihood as if mle=1.
381: Some cleaning of code and comments added.
382:
1.136 brouard 383: Revision 1.135 2009/10/29 15:33:14 brouard
384: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
385:
1.135 brouard 386: Revision 1.134 2009/10/29 13:18:53 brouard
387: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
388:
1.134 brouard 389: Revision 1.133 2009/07/06 10:21:25 brouard
390: just nforces
391:
1.133 brouard 392: Revision 1.132 2009/07/06 08:22:05 brouard
393: Many tings
394:
1.132 brouard 395: Revision 1.131 2009/06/20 16:22:47 brouard
396: Some dimensions resccaled
397:
1.131 brouard 398: Revision 1.130 2009/05/26 06:44:34 brouard
399: (Module): Max Covariate is now set to 20 instead of 8. A
400: lot of cleaning with variables initialized to 0. Trying to make
401: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
402:
1.130 brouard 403: Revision 1.129 2007/08/31 13:49:27 lievre
404: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
405:
1.129 lievre 406: Revision 1.128 2006/06/30 13:02:05 brouard
407: (Module): Clarifications on computing e.j
408:
1.128 brouard 409: Revision 1.127 2006/04/28 18:11:50 brouard
410: (Module): Yes the sum of survivors was wrong since
411: imach-114 because nhstepm was no more computed in the age
412: loop. Now we define nhstepma in the age loop.
413: (Module): In order to speed up (in case of numerous covariates) we
414: compute health expectancies (without variances) in a first step
415: and then all the health expectancies with variances or standard
416: deviation (needs data from the Hessian matrices) which slows the
417: computation.
418: In the future we should be able to stop the program is only health
419: expectancies and graph are needed without standard deviations.
420:
1.127 brouard 421: Revision 1.126 2006/04/28 17:23:28 brouard
422: (Module): Yes the sum of survivors was wrong since
423: imach-114 because nhstepm was no more computed in the age
424: loop. Now we define nhstepma in the age loop.
425: Version 0.98h
426:
1.126 brouard 427: Revision 1.125 2006/04/04 15:20:31 lievre
428: Errors in calculation of health expectancies. Age was not initialized.
429: Forecasting file added.
430:
431: Revision 1.124 2006/03/22 17:13:53 lievre
432: Parameters are printed with %lf instead of %f (more numbers after the comma).
433: The log-likelihood is printed in the log file
434:
435: Revision 1.123 2006/03/20 10:52:43 brouard
436: * imach.c (Module): <title> changed, corresponds to .htm file
437: name. <head> headers where missing.
438:
439: * imach.c (Module): Weights can have a decimal point as for
440: English (a comma might work with a correct LC_NUMERIC environment,
441: otherwise the weight is truncated).
442: Modification of warning when the covariates values are not 0 or
443: 1.
444: Version 0.98g
445:
446: Revision 1.122 2006/03/20 09:45:41 brouard
447: (Module): Weights can have a decimal point as for
448: English (a comma might work with a correct LC_NUMERIC environment,
449: otherwise the weight is truncated).
450: Modification of warning when the covariates values are not 0 or
451: 1.
452: Version 0.98g
453:
454: Revision 1.121 2006/03/16 17:45:01 lievre
455: * imach.c (Module): Comments concerning covariates added
456:
457: * imach.c (Module): refinements in the computation of lli if
458: status=-2 in order to have more reliable computation if stepm is
459: not 1 month. Version 0.98f
460:
461: Revision 1.120 2006/03/16 15:10:38 lievre
462: (Module): refinements in the computation of lli if
463: status=-2 in order to have more reliable computation if stepm is
464: not 1 month. Version 0.98f
465:
466: Revision 1.119 2006/03/15 17:42:26 brouard
467: (Module): Bug if status = -2, the loglikelihood was
468: computed as likelihood omitting the logarithm. Version O.98e
469:
470: Revision 1.118 2006/03/14 18:20:07 brouard
471: (Module): varevsij Comments added explaining the second
472: table of variances if popbased=1 .
473: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
474: (Module): Function pstamp added
475: (Module): Version 0.98d
476:
477: Revision 1.117 2006/03/14 17:16:22 brouard
478: (Module): varevsij Comments added explaining the second
479: table of variances if popbased=1 .
480: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
481: (Module): Function pstamp added
482: (Module): Version 0.98d
483:
484: Revision 1.116 2006/03/06 10:29:27 brouard
485: (Module): Variance-covariance wrong links and
486: varian-covariance of ej. is needed (Saito).
487:
488: Revision 1.115 2006/02/27 12:17:45 brouard
489: (Module): One freematrix added in mlikeli! 0.98c
490:
491: Revision 1.114 2006/02/26 12:57:58 brouard
492: (Module): Some improvements in processing parameter
493: filename with strsep.
494:
495: Revision 1.113 2006/02/24 14:20:24 brouard
496: (Module): Memory leaks checks with valgrind and:
497: datafile was not closed, some imatrix were not freed and on matrix
498: allocation too.
499:
500: Revision 1.112 2006/01/30 09:55:26 brouard
501: (Module): Back to gnuplot.exe instead of wgnuplot.exe
502:
503: Revision 1.111 2006/01/25 20:38:18 brouard
504: (Module): Lots of cleaning and bugs added (Gompertz)
505: (Module): Comments can be added in data file. Missing date values
506: can be a simple dot '.'.
507:
508: Revision 1.110 2006/01/25 00:51:50 brouard
509: (Module): Lots of cleaning and bugs added (Gompertz)
510:
511: Revision 1.109 2006/01/24 19:37:15 brouard
512: (Module): Comments (lines starting with a #) are allowed in data.
513:
514: Revision 1.108 2006/01/19 18:05:42 lievre
515: Gnuplot problem appeared...
516: To be fixed
517:
518: Revision 1.107 2006/01/19 16:20:37 brouard
519: Test existence of gnuplot in imach path
520:
521: Revision 1.106 2006/01/19 13:24:36 brouard
522: Some cleaning and links added in html output
523:
524: Revision 1.105 2006/01/05 20:23:19 lievre
525: *** empty log message ***
526:
527: Revision 1.104 2005/09/30 16:11:43 lievre
528: (Module): sump fixed, loop imx fixed, and simplifications.
529: (Module): If the status is missing at the last wave but we know
530: that the person is alive, then we can code his/her status as -2
531: (instead of missing=-1 in earlier versions) and his/her
532: contributions to the likelihood is 1 - Prob of dying from last
533: health status (= 1-p13= p11+p12 in the easiest case of somebody in
534: the healthy state at last known wave). Version is 0.98
535:
536: Revision 1.103 2005/09/30 15:54:49 lievre
537: (Module): sump fixed, loop imx fixed, and simplifications.
538:
539: Revision 1.102 2004/09/15 17:31:30 brouard
540: Add the possibility to read data file including tab characters.
541:
542: Revision 1.101 2004/09/15 10:38:38 brouard
543: Fix on curr_time
544:
545: Revision 1.100 2004/07/12 18:29:06 brouard
546: Add version for Mac OS X. Just define UNIX in Makefile
547:
548: Revision 1.99 2004/06/05 08:57:40 brouard
549: *** empty log message ***
550:
551: Revision 1.98 2004/05/16 15:05:56 brouard
552: New version 0.97 . First attempt to estimate force of mortality
553: directly from the data i.e. without the need of knowing the health
554: state at each age, but using a Gompertz model: log u =a + b*age .
555: This is the basic analysis of mortality and should be done before any
556: other analysis, in order to test if the mortality estimated from the
557: cross-longitudinal survey is different from the mortality estimated
558: from other sources like vital statistic data.
559:
560: The same imach parameter file can be used but the option for mle should be -3.
561:
1.133 brouard 562: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 563: former routines in order to include the new code within the former code.
564:
565: The output is very simple: only an estimate of the intercept and of
566: the slope with 95% confident intervals.
567:
568: Current limitations:
569: A) Even if you enter covariates, i.e. with the
570: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
571: B) There is no computation of Life Expectancy nor Life Table.
572:
573: Revision 1.97 2004/02/20 13:25:42 lievre
574: Version 0.96d. Population forecasting command line is (temporarily)
575: suppressed.
576:
577: Revision 1.96 2003/07/15 15:38:55 brouard
578: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
579: rewritten within the same printf. Workaround: many printfs.
580:
581: Revision 1.95 2003/07/08 07:54:34 brouard
582: * imach.c (Repository):
583: (Repository): Using imachwizard code to output a more meaningful covariance
584: matrix (cov(a12,c31) instead of numbers.
585:
586: Revision 1.94 2003/06/27 13:00:02 brouard
587: Just cleaning
588:
589: Revision 1.93 2003/06/25 16:33:55 brouard
590: (Module): On windows (cygwin) function asctime_r doesn't
591: exist so I changed back to asctime which exists.
592: (Module): Version 0.96b
593:
594: Revision 1.92 2003/06/25 16:30:45 brouard
595: (Module): On windows (cygwin) function asctime_r doesn't
596: exist so I changed back to asctime which exists.
597:
598: Revision 1.91 2003/06/25 15:30:29 brouard
599: * imach.c (Repository): Duplicated warning errors corrected.
600: (Repository): Elapsed time after each iteration is now output. It
601: helps to forecast when convergence will be reached. Elapsed time
602: is stamped in powell. We created a new html file for the graphs
603: concerning matrix of covariance. It has extension -cov.htm.
604:
605: Revision 1.90 2003/06/24 12:34:15 brouard
606: (Module): Some bugs corrected for windows. Also, when
607: mle=-1 a template is output in file "or"mypar.txt with the design
608: of the covariance matrix to be input.
609:
610: Revision 1.89 2003/06/24 12:30:52 brouard
611: (Module): Some bugs corrected for windows. Also, when
612: mle=-1 a template is output in file "or"mypar.txt with the design
613: of the covariance matrix to be input.
614:
615: Revision 1.88 2003/06/23 17:54:56 brouard
616: * 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.
617:
618: Revision 1.87 2003/06/18 12:26:01 brouard
619: Version 0.96
620:
621: Revision 1.86 2003/06/17 20:04:08 brouard
622: (Module): Change position of html and gnuplot routines and added
623: routine fileappend.
624:
625: Revision 1.85 2003/06/17 13:12:43 brouard
626: * imach.c (Repository): Check when date of death was earlier that
627: current date of interview. It may happen when the death was just
628: prior to the death. In this case, dh was negative and likelihood
629: was wrong (infinity). We still send an "Error" but patch by
630: assuming that the date of death was just one stepm after the
631: interview.
632: (Repository): Because some people have very long ID (first column)
633: we changed int to long in num[] and we added a new lvector for
634: memory allocation. But we also truncated to 8 characters (left
635: truncation)
636: (Repository): No more line truncation errors.
637:
638: Revision 1.84 2003/06/13 21:44:43 brouard
639: * imach.c (Repository): Replace "freqsummary" at a correct
640: place. It differs from routine "prevalence" which may be called
641: many times. Probs is memory consuming and must be used with
642: parcimony.
643: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
644:
645: Revision 1.83 2003/06/10 13:39:11 lievre
646: *** empty log message ***
647:
648: Revision 1.82 2003/06/05 15:57:20 brouard
649: Add log in imach.c and fullversion number is now printed.
650:
651: */
652: /*
653: Interpolated Markov Chain
654:
655: Short summary of the programme:
656:
1.227 brouard 657: This program computes Healthy Life Expectancies or State-specific
658: (if states aren't health statuses) Expectancies from
659: cross-longitudinal data. Cross-longitudinal data consist in:
660:
661: -1- a first survey ("cross") where individuals from different ages
662: are interviewed on their health status or degree of disability (in
663: the case of a health survey which is our main interest)
664:
665: -2- at least a second wave of interviews ("longitudinal") which
666: measure each change (if any) in individual health status. Health
667: expectancies are computed from the time spent in each health state
668: according to a model. More health states you consider, more time is
669: necessary to reach the Maximum Likelihood of the parameters involved
670: in the model. The simplest model is the multinomial logistic model
671: where pij is the probability to be observed in state j at the second
672: wave conditional to be observed in state i at the first
673: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
674: etc , where 'age' is age and 'sex' is a covariate. If you want to
675: have a more complex model than "constant and age", you should modify
676: the program where the markup *Covariates have to be included here
677: again* invites you to do it. More covariates you add, slower the
1.126 brouard 678: convergence.
679:
680: The advantage of this computer programme, compared to a simple
681: multinomial logistic model, is clear when the delay between waves is not
682: identical for each individual. Also, if a individual missed an
683: intermediate interview, the information is lost, but taken into
684: account using an interpolation or extrapolation.
685:
686: hPijx is the probability to be observed in state i at age x+h
687: conditional to the observed state i at age x. The delay 'h' can be
688: split into an exact number (nh*stepm) of unobserved intermediate
689: states. This elementary transition (by month, quarter,
690: semester or year) is modelled as a multinomial logistic. The hPx
691: matrix is simply the matrix product of nh*stepm elementary matrices
692: and the contribution of each individual to the likelihood is simply
693: hPijx.
694:
695: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 696: of the life expectancies. It also computes the period (stable) prevalence.
697:
698: Back prevalence and projections:
1.227 brouard 699:
700: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
701: double agemaxpar, double ftolpl, int *ncvyearp, double
702: dateprev1,double dateprev2, int firstpass, int lastpass, int
703: mobilavproj)
704:
705: Computes the back prevalence limit for any combination of
706: covariate values k at any age between ageminpar and agemaxpar and
707: returns it in **bprlim. In the loops,
708:
709: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
710: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
711:
712: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 713: Computes for any combination of covariates k and any age between bage and fage
714: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
715: oldm=oldms;savm=savms;
1.227 brouard 716:
717: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 718: Computes the transition matrix starting at age 'age' over
719: 'nhstepm*hstepm*stepm' months (i.e. until
720: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 721: nhstepm*hstepm matrices.
722:
723: Returns p3mat[i][j][h] after calling
724: p3mat[i][j][h]=matprod2(newm,
725: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
726: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
727: oldm);
1.226 brouard 728:
729: Important routines
730:
731: - func (or funcone), computes logit (pij) distinguishing
732: o fixed variables (single or product dummies or quantitative);
733: o varying variables by:
734: (1) wave (single, product dummies, quantitative),
735: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
736: % fixed dummy (treated) or quantitative (not done because time-consuming);
737: % varying dummy (not done) or quantitative (not done);
738: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
739: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
740: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
741: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
742: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 743:
1.226 brouard 744:
745:
1.133 brouard 746: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
747: Institut national d'études démographiques, Paris.
1.126 brouard 748: This software have been partly granted by Euro-REVES, a concerted action
749: from the European Union.
750: It is copyrighted identically to a GNU software product, ie programme and
751: software can be distributed freely for non commercial use. Latest version
752: can be accessed at http://euroreves.ined.fr/imach .
753:
754: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
755: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
756:
757: **********************************************************************/
758: /*
759: main
760: read parameterfile
761: read datafile
762: concatwav
763: freqsummary
764: if (mle >= 1)
765: mlikeli
766: print results files
767: if mle==1
768: computes hessian
769: read end of parameter file: agemin, agemax, bage, fage, estepm
770: begin-prev-date,...
771: open gnuplot file
772: open html file
1.145 brouard 773: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
774: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
775: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
776: freexexit2 possible for memory heap.
777:
778: h Pij x | pij_nom ficrestpij
779: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
780: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
781: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
782:
783: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
784: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
785: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
786: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
787: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
788:
1.126 brouard 789: forecasting if prevfcast==1 prevforecast call prevalence()
790: health expectancies
791: Variance-covariance of DFLE
792: prevalence()
793: movingaverage()
794: varevsij()
795: if popbased==1 varevsij(,popbased)
796: total life expectancies
797: Variance of period (stable) prevalence
798: end
799: */
800:
1.187 brouard 801: /* #define DEBUG */
802: /* #define DEBUGBRENT */
1.203 brouard 803: /* #define DEBUGLINMIN */
804: /* #define DEBUGHESS */
805: #define DEBUGHESSIJ
1.224 brouard 806: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 807: #define POWELL /* Instead of NLOPT */
1.224 brouard 808: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 809: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
810: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 811:
812: #include <math.h>
813: #include <stdio.h>
814: #include <stdlib.h>
815: #include <string.h>
1.226 brouard 816: #include <ctype.h>
1.159 brouard 817:
818: #ifdef _WIN32
819: #include <io.h>
1.172 brouard 820: #include <windows.h>
821: #include <tchar.h>
1.159 brouard 822: #else
1.126 brouard 823: #include <unistd.h>
1.159 brouard 824: #endif
1.126 brouard 825:
826: #include <limits.h>
827: #include <sys/types.h>
1.171 brouard 828:
829: #if defined(__GNUC__)
830: #include <sys/utsname.h> /* Doesn't work on Windows */
831: #endif
832:
1.126 brouard 833: #include <sys/stat.h>
834: #include <errno.h>
1.159 brouard 835: /* extern int errno; */
1.126 brouard 836:
1.157 brouard 837: /* #ifdef LINUX */
838: /* #include <time.h> */
839: /* #include "timeval.h" */
840: /* #else */
841: /* #include <sys/time.h> */
842: /* #endif */
843:
1.126 brouard 844: #include <time.h>
845:
1.136 brouard 846: #ifdef GSL
847: #include <gsl/gsl_errno.h>
848: #include <gsl/gsl_multimin.h>
849: #endif
850:
1.167 brouard 851:
1.162 brouard 852: #ifdef NLOPT
853: #include <nlopt.h>
854: typedef struct {
855: double (* function)(double [] );
856: } myfunc_data ;
857: #endif
858:
1.126 brouard 859: /* #include <libintl.h> */
860: /* #define _(String) gettext (String) */
861:
1.141 brouard 862: #define MAXLINE 1024 /* Was 256. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 863:
864: #define GNUPLOTPROGRAM "gnuplot"
865: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
866: #define FILENAMELENGTH 132
867:
868: #define GLOCK_ERROR_NOPATH -1 /* empty path */
869: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
870:
1.144 brouard 871: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
872: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 873:
874: #define NINTERVMAX 8
1.144 brouard 875: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
876: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
877: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 878: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 879: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
880: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 881: #define MAXN 20000
1.144 brouard 882: #define YEARM 12. /**< Number of months per year */
1.218 brouard 883: /* #define AGESUP 130 */
884: #define AGESUP 150
885: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 886: #define AGEBASE 40
1.194 brouard 887: #define AGEOVERFLOW 1.e20
1.164 brouard 888: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 889: #ifdef _WIN32
890: #define DIRSEPARATOR '\\'
891: #define CHARSEPARATOR "\\"
892: #define ODIRSEPARATOR '/'
893: #else
1.126 brouard 894: #define DIRSEPARATOR '/'
895: #define CHARSEPARATOR "/"
896: #define ODIRSEPARATOR '\\'
897: #endif
898:
1.233 ! brouard 899: /* $Id: imach.c,v 1.232 2016/08/22 14:20:21 brouard Exp $ */
1.126 brouard 900: /* $State: Exp $ */
1.196 brouard 901: #include "version.h"
902: char version[]=__IMACH_VERSION__;
1.224 brouard 903: 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";
1.233 ! brouard 904: char fullversion[]="$Revision: 1.232 $ $Date: 2016/08/22 14:20:21 $";
1.126 brouard 905: char strstart[80];
906: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 907: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 908: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 909: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
910: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
911: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 912: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
913: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 914: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
915: int cptcovprodnoage=0; /**< Number of covariate products without age */
916: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 ! brouard 917: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
! 918: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 919: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
920:
921: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 922: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 923: int ntveff=0; /**< ntveff number of effective time varying variables */
924: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 925: int cptcov=0; /* Working variable */
1.218 brouard 926: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 927: int npar=NPARMAX;
928: int nlstate=2; /* Number of live states */
929: int ndeath=1; /* Number of dead states */
1.130 brouard 930: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 931: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 932: int popbased=0;
933:
934: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 935: int maxwav=0; /* Maxim number of waves */
936: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
937: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
938: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 939: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 940: int mle=1, weightopt=0;
1.126 brouard 941: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
942: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
943: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
944: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 945: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 946: int selected(int kvar); /* Is covariate kvar selected for printing results */
947:
1.130 brouard 948: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 949: double **matprod2(); /* test */
1.126 brouard 950: double **oldm, **newm, **savm; /* Working pointers to matrices */
951: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 952: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
953:
1.136 brouard 954: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 955: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 956: FILE *ficlog, *ficrespow;
1.130 brouard 957: int globpr=0; /* Global variable for printing or not */
1.126 brouard 958: double fretone; /* Only one call to likelihood */
1.130 brouard 959: long ipmx=0; /* Number of contributions */
1.126 brouard 960: double sw; /* Sum of weights */
961: char filerespow[FILENAMELENGTH];
962: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
963: FILE *ficresilk;
964: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
965: FILE *ficresprobmorprev;
966: FILE *fichtm, *fichtmcov; /* Html File */
967: FILE *ficreseij;
968: char filerese[FILENAMELENGTH];
969: FILE *ficresstdeij;
970: char fileresstde[FILENAMELENGTH];
971: FILE *ficrescveij;
972: char filerescve[FILENAMELENGTH];
973: FILE *ficresvij;
974: char fileresv[FILENAMELENGTH];
975: FILE *ficresvpl;
976: char fileresvpl[FILENAMELENGTH];
977: char title[MAXLINE];
1.217 brouard 978: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 979: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
980: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
981: char command[FILENAMELENGTH];
982: int outcmd=0;
983:
1.217 brouard 984: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 985: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 986: char filelog[FILENAMELENGTH]; /* Log file */
987: char filerest[FILENAMELENGTH];
988: char fileregp[FILENAMELENGTH];
989: char popfile[FILENAMELENGTH];
990:
991: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
992:
1.157 brouard 993: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
994: /* struct timezone tzp; */
995: /* extern int gettimeofday(); */
996: struct tm tml, *gmtime(), *localtime();
997:
998: extern time_t time();
999:
1000: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1001: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1002: struct tm tm;
1003:
1.126 brouard 1004: char strcurr[80], strfor[80];
1005:
1006: char *endptr;
1007: long lval;
1008: double dval;
1009:
1010: #define NR_END 1
1011: #define FREE_ARG char*
1012: #define FTOL 1.0e-10
1013:
1014: #define NRANSI
1015: #define ITMAX 200
1016:
1017: #define TOL 2.0e-4
1018:
1019: #define CGOLD 0.3819660
1020: #define ZEPS 1.0e-10
1021: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1022:
1023: #define GOLD 1.618034
1024: #define GLIMIT 100.0
1025: #define TINY 1.0e-20
1026:
1027: static double maxarg1,maxarg2;
1028: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1029: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1030:
1031: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1032: #define rint(a) floor(a+0.5)
1.166 brouard 1033: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1034: #define mytinydouble 1.0e-16
1.166 brouard 1035: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1036: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1037: /* static double dsqrarg; */
1038: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1039: static double sqrarg;
1040: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1041: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1042: int agegomp= AGEGOMP;
1043:
1044: int imx;
1045: int stepm=1;
1046: /* Stepm, step in month: minimum step interpolation*/
1047:
1048: int estepm;
1049: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1050:
1051: int m,nb;
1052: long *num;
1.197 brouard 1053: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1054: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1055: covariate for which somebody answered excluding
1056: undefined. Usually 2: 0 and 1. */
1057: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1058: covariate for which somebody answered including
1059: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1060: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1061: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1062: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1063: double *ageexmed,*agecens;
1064: double dateintmean=0;
1065:
1066: double *weight;
1067: int **s; /* Status */
1.141 brouard 1068: double *agedc;
1.145 brouard 1069: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1070: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1071: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1072: double **coqvar; /* Fixed quantitative covariate iqv */
1073: double ***cotvar; /* Time varying covariate itv */
1074: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1075: double idx;
1076: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.232 brouard 1077: 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 */
1078: 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 */
1079: 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 */
1080: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1081: 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 */
1082: 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 1083: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1084: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1085: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1086: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1087: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1088: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1089: 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 */
1090: 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 */
1091:
1.230 brouard 1092: int *Tvarsel; /**< Selected covariates for output */
1093: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1094: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1095: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1096: 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 1097: int *Tage;
1.227 brouard 1098: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1099: 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 1100: 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*/
1101: 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 1102: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1103: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1104: int **Tvard;
1105: int *Tprod;/**< Gives the k position of the k1 product */
1106: int *Tposprod; /**< Gives the k1 product from the k position */
1107: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
1108: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1109: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1110: */
1111: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1112: double *lsurv, *lpop, *tpop;
1113:
1.231 brouard 1114: #define FD 1; /* Fixed dummy covariate */
1115: #define FQ 2; /* Fixed quantitative covariate */
1116: #define FP 3; /* Fixed product covariate */
1117: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1118: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1119: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1120: #define VD 10; /* Varying dummy covariate */
1121: #define VQ 11; /* Varying quantitative covariate */
1122: #define VP 12; /* Varying product covariate */
1123: #define VPDD 13; /* Varying product dummy*dummy covariate */
1124: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1125: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1126: #define APFD 16; /* Age product * fixed dummy covariate */
1127: #define APFQ 17; /* Age product * fixed quantitative covariate */
1128: #define APVD 18; /* Age product * varying dummy covariate */
1129: #define APVQ 19; /* Age product * varying quantitative covariate */
1130:
1131: #define FTYPE 1; /* Fixed covariate */
1132: #define VTYPE 2; /* Varying covariate (loop in wave) */
1133: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1134:
1135: struct kmodel{
1136: int maintype; /* main type */
1137: int subtype; /* subtype */
1138: };
1139: struct kmodel modell[NCOVMAX];
1140:
1.143 brouard 1141: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1142: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1143:
1144: /**************** split *************************/
1145: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1146: {
1147: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1148: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1149: */
1150: char *ss; /* pointer */
1.186 brouard 1151: int l1=0, l2=0; /* length counters */
1.126 brouard 1152:
1153: l1 = strlen(path ); /* length of path */
1154: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1155: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1156: if ( ss == NULL ) { /* no directory, so determine current directory */
1157: strcpy( name, path ); /* we got the fullname name because no directory */
1158: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1159: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1160: /* get current working directory */
1161: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1162: #ifdef WIN32
1163: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1164: #else
1165: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1166: #endif
1.126 brouard 1167: return( GLOCK_ERROR_GETCWD );
1168: }
1169: /* got dirc from getcwd*/
1170: printf(" DIRC = %s \n",dirc);
1.205 brouard 1171: } else { /* strip directory from path */
1.126 brouard 1172: ss++; /* after this, the filename */
1173: l2 = strlen( ss ); /* length of filename */
1174: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1175: strcpy( name, ss ); /* save file name */
1176: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1177: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1178: printf(" DIRC2 = %s \n",dirc);
1179: }
1180: /* We add a separator at the end of dirc if not exists */
1181: l1 = strlen( dirc ); /* length of directory */
1182: if( dirc[l1-1] != DIRSEPARATOR ){
1183: dirc[l1] = DIRSEPARATOR;
1184: dirc[l1+1] = 0;
1185: printf(" DIRC3 = %s \n",dirc);
1186: }
1187: ss = strrchr( name, '.' ); /* find last / */
1188: if (ss >0){
1189: ss++;
1190: strcpy(ext,ss); /* save extension */
1191: l1= strlen( name);
1192: l2= strlen(ss)+1;
1193: strncpy( finame, name, l1-l2);
1194: finame[l1-l2]= 0;
1195: }
1196:
1197: return( 0 ); /* we're done */
1198: }
1199:
1200:
1201: /******************************************/
1202:
1203: void replace_back_to_slash(char *s, char*t)
1204: {
1205: int i;
1206: int lg=0;
1207: i=0;
1208: lg=strlen(t);
1209: for(i=0; i<= lg; i++) {
1210: (s[i] = t[i]);
1211: if (t[i]== '\\') s[i]='/';
1212: }
1213: }
1214:
1.132 brouard 1215: char *trimbb(char *out, char *in)
1.137 brouard 1216: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1217: char *s;
1218: s=out;
1219: while (*in != '\0'){
1.137 brouard 1220: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1221: in++;
1222: }
1223: *out++ = *in++;
1224: }
1225: *out='\0';
1226: return s;
1227: }
1228:
1.187 brouard 1229: /* char *substrchaine(char *out, char *in, char *chain) */
1230: /* { */
1231: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1232: /* char *s, *t; */
1233: /* t=in;s=out; */
1234: /* while ((*in != *chain) && (*in != '\0')){ */
1235: /* *out++ = *in++; */
1236: /* } */
1237:
1238: /* /\* *in matches *chain *\/ */
1239: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1240: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1241: /* } */
1242: /* in--; chain--; */
1243: /* while ( (*in != '\0')){ */
1244: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1245: /* *out++ = *in++; */
1246: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1247: /* } */
1248: /* *out='\0'; */
1249: /* out=s; */
1250: /* return out; */
1251: /* } */
1252: char *substrchaine(char *out, char *in, char *chain)
1253: {
1254: /* Substract chain 'chain' from 'in', return and output 'out' */
1255: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1256:
1257: char *strloc;
1258:
1259: strcpy (out, in);
1260: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1261: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1262: if(strloc != NULL){
1263: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1264: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1265: /* strcpy (strloc, strloc +strlen(chain));*/
1266: }
1267: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1268: return out;
1269: }
1270:
1271:
1.145 brouard 1272: char *cutl(char *blocc, char *alocc, char *in, char occ)
1273: {
1.187 brouard 1274: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1275: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1276: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1277: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1278: */
1.160 brouard 1279: char *s, *t;
1.145 brouard 1280: t=in;s=in;
1281: while ((*in != occ) && (*in != '\0')){
1282: *alocc++ = *in++;
1283: }
1284: if( *in == occ){
1285: *(alocc)='\0';
1286: s=++in;
1287: }
1288:
1289: if (s == t) {/* occ not found */
1290: *(alocc-(in-s))='\0';
1291: in=s;
1292: }
1293: while ( *in != '\0'){
1294: *blocc++ = *in++;
1295: }
1296:
1297: *blocc='\0';
1298: return t;
1299: }
1.137 brouard 1300: char *cutv(char *blocc, char *alocc, char *in, char occ)
1301: {
1.187 brouard 1302: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1303: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1304: gives blocc="abcdef2ghi" and alocc="j".
1305: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1306: */
1307: char *s, *t;
1308: t=in;s=in;
1309: while (*in != '\0'){
1310: while( *in == occ){
1311: *blocc++ = *in++;
1312: s=in;
1313: }
1314: *blocc++ = *in++;
1315: }
1316: if (s == t) /* occ not found */
1317: *(blocc-(in-s))='\0';
1318: else
1319: *(blocc-(in-s)-1)='\0';
1320: in=s;
1321: while ( *in != '\0'){
1322: *alocc++ = *in++;
1323: }
1324:
1325: *alocc='\0';
1326: return s;
1327: }
1328:
1.126 brouard 1329: int nbocc(char *s, char occ)
1330: {
1331: int i,j=0;
1332: int lg=20;
1333: i=0;
1334: lg=strlen(s);
1335: for(i=0; i<= lg; i++) {
1336: if (s[i] == occ ) j++;
1337: }
1338: return j;
1339: }
1340:
1.137 brouard 1341: /* void cutv(char *u,char *v, char*t, char occ) */
1342: /* { */
1343: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1344: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1345: /* gives u="abcdef2ghi" and v="j" *\/ */
1346: /* int i,lg,j,p=0; */
1347: /* i=0; */
1348: /* lg=strlen(t); */
1349: /* for(j=0; j<=lg-1; j++) { */
1350: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1351: /* } */
1.126 brouard 1352:
1.137 brouard 1353: /* for(j=0; j<p; j++) { */
1354: /* (u[j] = t[j]); */
1355: /* } */
1356: /* u[p]='\0'; */
1.126 brouard 1357:
1.137 brouard 1358: /* for(j=0; j<= lg; j++) { */
1359: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1360: /* } */
1361: /* } */
1.126 brouard 1362:
1.160 brouard 1363: #ifdef _WIN32
1364: char * strsep(char **pp, const char *delim)
1365: {
1366: char *p, *q;
1367:
1368: if ((p = *pp) == NULL)
1369: return 0;
1370: if ((q = strpbrk (p, delim)) != NULL)
1371: {
1372: *pp = q + 1;
1373: *q = '\0';
1374: }
1375: else
1376: *pp = 0;
1377: return p;
1378: }
1379: #endif
1380:
1.126 brouard 1381: /********************** nrerror ********************/
1382:
1383: void nrerror(char error_text[])
1384: {
1385: fprintf(stderr,"ERREUR ...\n");
1386: fprintf(stderr,"%s\n",error_text);
1387: exit(EXIT_FAILURE);
1388: }
1389: /*********************** vector *******************/
1390: double *vector(int nl, int nh)
1391: {
1392: double *v;
1393: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1394: if (!v) nrerror("allocation failure in vector");
1395: return v-nl+NR_END;
1396: }
1397:
1398: /************************ free vector ******************/
1399: void free_vector(double*v, int nl, int nh)
1400: {
1401: free((FREE_ARG)(v+nl-NR_END));
1402: }
1403:
1404: /************************ivector *******************************/
1405: int *ivector(long nl,long nh)
1406: {
1407: int *v;
1408: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1409: if (!v) nrerror("allocation failure in ivector");
1410: return v-nl+NR_END;
1411: }
1412:
1413: /******************free ivector **************************/
1414: void free_ivector(int *v, long nl, long nh)
1415: {
1416: free((FREE_ARG)(v+nl-NR_END));
1417: }
1418:
1419: /************************lvector *******************************/
1420: long *lvector(long nl,long nh)
1421: {
1422: long *v;
1423: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1424: if (!v) nrerror("allocation failure in ivector");
1425: return v-nl+NR_END;
1426: }
1427:
1428: /******************free lvector **************************/
1429: void free_lvector(long *v, long nl, long nh)
1430: {
1431: free((FREE_ARG)(v+nl-NR_END));
1432: }
1433:
1434: /******************* imatrix *******************************/
1435: int **imatrix(long nrl, long nrh, long ncl, long nch)
1436: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1437: {
1438: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1439: int **m;
1440:
1441: /* allocate pointers to rows */
1442: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1443: if (!m) nrerror("allocation failure 1 in matrix()");
1444: m += NR_END;
1445: m -= nrl;
1446:
1447:
1448: /* allocate rows and set pointers to them */
1449: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1450: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1451: m[nrl] += NR_END;
1452: m[nrl] -= ncl;
1453:
1454: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1455:
1456: /* return pointer to array of pointers to rows */
1457: return m;
1458: }
1459:
1460: /****************** free_imatrix *************************/
1461: void free_imatrix(m,nrl,nrh,ncl,nch)
1462: int **m;
1463: long nch,ncl,nrh,nrl;
1464: /* free an int matrix allocated by imatrix() */
1465: {
1466: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1467: free((FREE_ARG) (m+nrl-NR_END));
1468: }
1469:
1470: /******************* matrix *******************************/
1471: double **matrix(long nrl, long nrh, long ncl, long nch)
1472: {
1473: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1474: double **m;
1475:
1476: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1477: if (!m) nrerror("allocation failure 1 in matrix()");
1478: m += NR_END;
1479: m -= nrl;
1480:
1481: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1482: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1483: m[nrl] += NR_END;
1484: m[nrl] -= ncl;
1485:
1486: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1487: return m;
1.145 brouard 1488: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1489: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1490: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1491: */
1492: }
1493:
1494: /*************************free matrix ************************/
1495: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1496: {
1497: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1498: free((FREE_ARG)(m+nrl-NR_END));
1499: }
1500:
1501: /******************* ma3x *******************************/
1502: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1503: {
1504: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1505: double ***m;
1506:
1507: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1508: if (!m) nrerror("allocation failure 1 in matrix()");
1509: m += NR_END;
1510: m -= nrl;
1511:
1512: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1513: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1514: m[nrl] += NR_END;
1515: m[nrl] -= ncl;
1516:
1517: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1518:
1519: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1520: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1521: m[nrl][ncl] += NR_END;
1522: m[nrl][ncl] -= nll;
1523: for (j=ncl+1; j<=nch; j++)
1524: m[nrl][j]=m[nrl][j-1]+nlay;
1525:
1526: for (i=nrl+1; i<=nrh; i++) {
1527: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1528: for (j=ncl+1; j<=nch; j++)
1529: m[i][j]=m[i][j-1]+nlay;
1530: }
1531: return m;
1532: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1533: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1534: */
1535: }
1536:
1537: /*************************free ma3x ************************/
1538: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1539: {
1540: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1541: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1542: free((FREE_ARG)(m+nrl-NR_END));
1543: }
1544:
1545: /*************** function subdirf ***********/
1546: char *subdirf(char fileres[])
1547: {
1548: /* Caution optionfilefiname is hidden */
1549: strcpy(tmpout,optionfilefiname);
1550: strcat(tmpout,"/"); /* Add to the right */
1551: strcat(tmpout,fileres);
1552: return tmpout;
1553: }
1554:
1555: /*************** function subdirf2 ***********/
1556: char *subdirf2(char fileres[], char *preop)
1557: {
1558:
1559: /* Caution optionfilefiname is hidden */
1560: strcpy(tmpout,optionfilefiname);
1561: strcat(tmpout,"/");
1562: strcat(tmpout,preop);
1563: strcat(tmpout,fileres);
1564: return tmpout;
1565: }
1566:
1567: /*************** function subdirf3 ***********/
1568: char *subdirf3(char fileres[], char *preop, char *preop2)
1569: {
1570:
1571: /* Caution optionfilefiname is hidden */
1572: strcpy(tmpout,optionfilefiname);
1573: strcat(tmpout,"/");
1574: strcat(tmpout,preop);
1575: strcat(tmpout,preop2);
1576: strcat(tmpout,fileres);
1577: return tmpout;
1578: }
1.213 brouard 1579:
1580: /*************** function subdirfext ***********/
1581: char *subdirfext(char fileres[], char *preop, char *postop)
1582: {
1583:
1584: strcpy(tmpout,preop);
1585: strcat(tmpout,fileres);
1586: strcat(tmpout,postop);
1587: return tmpout;
1588: }
1.126 brouard 1589:
1.213 brouard 1590: /*************** function subdirfext3 ***********/
1591: char *subdirfext3(char fileres[], char *preop, char *postop)
1592: {
1593:
1594: /* Caution optionfilefiname is hidden */
1595: strcpy(tmpout,optionfilefiname);
1596: strcat(tmpout,"/");
1597: strcat(tmpout,preop);
1598: strcat(tmpout,fileres);
1599: strcat(tmpout,postop);
1600: return tmpout;
1601: }
1602:
1.162 brouard 1603: char *asc_diff_time(long time_sec, char ascdiff[])
1604: {
1605: long sec_left, days, hours, minutes;
1606: days = (time_sec) / (60*60*24);
1607: sec_left = (time_sec) % (60*60*24);
1608: hours = (sec_left) / (60*60) ;
1609: sec_left = (sec_left) %(60*60);
1610: minutes = (sec_left) /60;
1611: sec_left = (sec_left) % (60);
1612: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1613: return ascdiff;
1614: }
1615:
1.126 brouard 1616: /***************** f1dim *************************/
1617: extern int ncom;
1618: extern double *pcom,*xicom;
1619: extern double (*nrfunc)(double []);
1620:
1621: double f1dim(double x)
1622: {
1623: int j;
1624: double f;
1625: double *xt;
1626:
1627: xt=vector(1,ncom);
1628: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1629: f=(*nrfunc)(xt);
1630: free_vector(xt,1,ncom);
1631: return f;
1632: }
1633:
1634: /*****************brent *************************/
1635: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1636: {
1637: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1638: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1639: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1640: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1641: * returned function value.
1642: */
1.126 brouard 1643: int iter;
1644: double a,b,d,etemp;
1.159 brouard 1645: double fu=0,fv,fw,fx;
1.164 brouard 1646: double ftemp=0.;
1.126 brouard 1647: double p,q,r,tol1,tol2,u,v,w,x,xm;
1648: double e=0.0;
1649:
1650: a=(ax < cx ? ax : cx);
1651: b=(ax > cx ? ax : cx);
1652: x=w=v=bx;
1653: fw=fv=fx=(*f)(x);
1654: for (iter=1;iter<=ITMAX;iter++) {
1655: xm=0.5*(a+b);
1656: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1657: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1658: printf(".");fflush(stdout);
1659: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1660: #ifdef DEBUGBRENT
1.126 brouard 1661: 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);
1662: 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);
1663: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1664: #endif
1665: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1666: *xmin=x;
1667: return fx;
1668: }
1669: ftemp=fu;
1670: if (fabs(e) > tol1) {
1671: r=(x-w)*(fx-fv);
1672: q=(x-v)*(fx-fw);
1673: p=(x-v)*q-(x-w)*r;
1674: q=2.0*(q-r);
1675: if (q > 0.0) p = -p;
1676: q=fabs(q);
1677: etemp=e;
1678: e=d;
1679: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1680: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1681: else {
1.224 brouard 1682: d=p/q;
1683: u=x+d;
1684: if (u-a < tol2 || b-u < tol2)
1685: d=SIGN(tol1,xm-x);
1.126 brouard 1686: }
1687: } else {
1688: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1689: }
1690: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1691: fu=(*f)(u);
1692: if (fu <= fx) {
1693: if (u >= x) a=x; else b=x;
1694: SHFT(v,w,x,u)
1.183 brouard 1695: SHFT(fv,fw,fx,fu)
1696: } else {
1697: if (u < x) a=u; else b=u;
1698: if (fu <= fw || w == x) {
1.224 brouard 1699: v=w;
1700: w=u;
1701: fv=fw;
1702: fw=fu;
1.183 brouard 1703: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1704: v=u;
1705: fv=fu;
1.183 brouard 1706: }
1707: }
1.126 brouard 1708: }
1709: nrerror("Too many iterations in brent");
1710: *xmin=x;
1711: return fx;
1712: }
1713:
1714: /****************** mnbrak ***********************/
1715:
1716: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1717: double (*func)(double))
1.183 brouard 1718: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1719: the downhill direction (defined by the function as evaluated at the initial points) and returns
1720: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1721: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1722: */
1.126 brouard 1723: double ulim,u,r,q, dum;
1724: double fu;
1.187 brouard 1725:
1726: double scale=10.;
1727: int iterscale=0;
1728:
1729: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1730: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1731:
1732:
1733: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1734: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1735: /* *bx = *ax - (*ax - *bx)/scale; */
1736: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1737: /* } */
1738:
1.126 brouard 1739: if (*fb > *fa) {
1740: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1741: SHFT(dum,*fb,*fa,dum)
1742: }
1.126 brouard 1743: *cx=(*bx)+GOLD*(*bx-*ax);
1744: *fc=(*func)(*cx);
1.183 brouard 1745: #ifdef DEBUG
1.224 brouard 1746: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1747: 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 1748: #endif
1.224 brouard 1749: 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 1750: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1751: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1752: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1753: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1754: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1755: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1756: fu=(*func)(u);
1.163 brouard 1757: #ifdef DEBUG
1758: /* f(x)=A(x-u)**2+f(u) */
1759: double A, fparabu;
1760: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1761: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1762: 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);
1763: 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 1764: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1765: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1766: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1767: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1768: #endif
1.184 brouard 1769: #ifdef MNBRAKORIGINAL
1.183 brouard 1770: #else
1.191 brouard 1771: /* if (fu > *fc) { */
1772: /* #ifdef DEBUG */
1773: /* printf("mnbrak4 fu > fc \n"); */
1774: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1775: /* #endif */
1776: /* /\* 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 *\\/ *\/ */
1777: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1778: /* dum=u; /\* Shifting c and u *\/ */
1779: /* u = *cx; */
1780: /* *cx = dum; */
1781: /* dum = fu; */
1782: /* fu = *fc; */
1783: /* *fc =dum; */
1784: /* } else { /\* end *\/ */
1785: /* #ifdef DEBUG */
1786: /* printf("mnbrak3 fu < fc \n"); */
1787: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1788: /* #endif */
1789: /* dum=u; /\* Shifting c and u *\/ */
1790: /* u = *cx; */
1791: /* *cx = dum; */
1792: /* dum = fu; */
1793: /* fu = *fc; */
1794: /* *fc =dum; */
1795: /* } */
1.224 brouard 1796: #ifdef DEBUGMNBRAK
1797: double A, fparabu;
1798: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1799: fparabu= *fa - A*(*ax-u)*(*ax-u);
1800: 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);
1801: 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 1802: #endif
1.191 brouard 1803: dum=u; /* Shifting c and u */
1804: u = *cx;
1805: *cx = dum;
1806: dum = fu;
1807: fu = *fc;
1808: *fc =dum;
1.183 brouard 1809: #endif
1.162 brouard 1810: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1811: #ifdef DEBUG
1.224 brouard 1812: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1813: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1814: #endif
1.126 brouard 1815: fu=(*func)(u);
1816: if (fu < *fc) {
1.183 brouard 1817: #ifdef DEBUG
1.224 brouard 1818: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1819: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1820: #endif
1821: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1822: SHFT(*fb,*fc,fu,(*func)(u))
1823: #ifdef DEBUG
1824: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1825: #endif
1826: }
1.162 brouard 1827: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1828: #ifdef DEBUG
1.224 brouard 1829: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1830: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1831: #endif
1.126 brouard 1832: u=ulim;
1833: fu=(*func)(u);
1.183 brouard 1834: } else { /* u could be left to b (if r > q parabola has a maximum) */
1835: #ifdef DEBUG
1.224 brouard 1836: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1837: 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 1838: #endif
1.126 brouard 1839: u=(*cx)+GOLD*(*cx-*bx);
1840: fu=(*func)(u);
1.224 brouard 1841: #ifdef DEBUG
1842: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1843: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1844: #endif
1.183 brouard 1845: } /* end tests */
1.126 brouard 1846: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1847: SHFT(*fa,*fb,*fc,fu)
1848: #ifdef DEBUG
1.224 brouard 1849: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1850: 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 1851: #endif
1852: } /* 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 1853: }
1854:
1855: /*************** linmin ************************/
1.162 brouard 1856: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1857: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1858: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1859: the value of func at the returned location p . This is actually all accomplished by calling the
1860: routines mnbrak and brent .*/
1.126 brouard 1861: int ncom;
1862: double *pcom,*xicom;
1863: double (*nrfunc)(double []);
1864:
1.224 brouard 1865: #ifdef LINMINORIGINAL
1.126 brouard 1866: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1867: #else
1868: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1869: #endif
1.126 brouard 1870: {
1871: double brent(double ax, double bx, double cx,
1872: double (*f)(double), double tol, double *xmin);
1873: double f1dim(double x);
1874: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
1875: double *fc, double (*func)(double));
1876: int j;
1877: double xx,xmin,bx,ax;
1878: double fx,fb,fa;
1.187 brouard 1879:
1.203 brouard 1880: #ifdef LINMINORIGINAL
1881: #else
1882: double scale=10., axs, xxs; /* Scale added for infinity */
1883: #endif
1884:
1.126 brouard 1885: ncom=n;
1886: pcom=vector(1,n);
1887: xicom=vector(1,n);
1888: nrfunc=func;
1889: for (j=1;j<=n;j++) {
1890: pcom[j]=p[j];
1.202 brouard 1891: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 1892: }
1.187 brouard 1893:
1.203 brouard 1894: #ifdef LINMINORIGINAL
1895: xx=1.;
1896: #else
1897: axs=0.0;
1898: xxs=1.;
1899: do{
1900: xx= xxs;
1901: #endif
1.187 brouard 1902: ax=0.;
1903: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
1904: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
1905: /* 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)) */
1906: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
1907: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
1908: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
1909: /* 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 1910: #ifdef LINMINORIGINAL
1911: #else
1912: if (fx != fx){
1.224 brouard 1913: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
1914: printf("|");
1915: fprintf(ficlog,"|");
1.203 brouard 1916: #ifdef DEBUGLINMIN
1.224 brouard 1917: 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 1918: #endif
1919: }
1.224 brouard 1920: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 1921: #endif
1922:
1.191 brouard 1923: #ifdef DEBUGLINMIN
1924: 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 1925: 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 1926: #endif
1.224 brouard 1927: #ifdef LINMINORIGINAL
1928: #else
1929: if(fb == fx){ /* Flat function in the direction */
1930: xmin=xx;
1931: *flat=1;
1932: }else{
1933: *flat=0;
1934: #endif
1935: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 1936: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
1937: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
1938: /* fmin = f(p[j] + xmin * xi[j]) */
1939: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
1940: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 1941: #ifdef DEBUG
1.224 brouard 1942: 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);
1943: 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);
1944: #endif
1945: #ifdef LINMINORIGINAL
1946: #else
1947: }
1.126 brouard 1948: #endif
1.191 brouard 1949: #ifdef DEBUGLINMIN
1950: printf("linmin end ");
1.202 brouard 1951: fprintf(ficlog,"linmin end ");
1.191 brouard 1952: #endif
1.126 brouard 1953: for (j=1;j<=n;j++) {
1.203 brouard 1954: #ifdef LINMINORIGINAL
1955: xi[j] *= xmin;
1956: #else
1957: #ifdef DEBUGLINMIN
1958: if(xxs <1.0)
1959: printf(" before xi[%d]=%12.8f", j,xi[j]);
1960: #endif
1961: 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) */
1962: #ifdef DEBUGLINMIN
1963: if(xxs <1.0)
1964: 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 );
1965: #endif
1966: #endif
1.187 brouard 1967: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 1968: }
1.191 brouard 1969: #ifdef DEBUGLINMIN
1.203 brouard 1970: printf("\n");
1.191 brouard 1971: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 1972: 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 1973: for (j=1;j<=n;j++) {
1.202 brouard 1974: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
1975: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
1976: if(j % ncovmodel == 0){
1.191 brouard 1977: printf("\n");
1.202 brouard 1978: fprintf(ficlog,"\n");
1979: }
1.191 brouard 1980: }
1.203 brouard 1981: #else
1.191 brouard 1982: #endif
1.126 brouard 1983: free_vector(xicom,1,n);
1984: free_vector(pcom,1,n);
1985: }
1986:
1987:
1988: /*************** powell ************************/
1.162 brouard 1989: /*
1990: Minimization of a function func of n variables. Input consists of an initial starting point
1991: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
1992: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
1993: such that failure to decrease by more than this amount on one iteration signals doneness. On
1994: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
1995: function value at p , and iter is the number of iterations taken. The routine linmin is used.
1996: */
1.224 brouard 1997: #ifdef LINMINORIGINAL
1998: #else
1999: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2000: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2001: #endif
1.126 brouard 2002: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2003: double (*func)(double []))
2004: {
1.224 brouard 2005: #ifdef LINMINORIGINAL
2006: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2007: double (*func)(double []));
1.224 brouard 2008: #else
2009: void linmin(double p[], double xi[], int n, double *fret,
2010: double (*func)(double []),int *flat);
2011: #endif
1.126 brouard 2012: int i,ibig,j;
2013: double del,t,*pt,*ptt,*xit;
1.181 brouard 2014: double directest;
1.126 brouard 2015: double fp,fptt;
2016: double *xits;
2017: int niterf, itmp;
1.224 brouard 2018: #ifdef LINMINORIGINAL
2019: #else
2020:
2021: flatdir=ivector(1,n);
2022: for (j=1;j<=n;j++) flatdir[j]=0;
2023: #endif
1.126 brouard 2024:
2025: pt=vector(1,n);
2026: ptt=vector(1,n);
2027: xit=vector(1,n);
2028: xits=vector(1,n);
2029: *fret=(*func)(p);
2030: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2031: rcurr_time = time(NULL);
1.126 brouard 2032: for (*iter=1;;++(*iter)) {
1.187 brouard 2033: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2034: ibig=0;
2035: del=0.0;
1.157 brouard 2036: rlast_time=rcurr_time;
2037: /* (void) gettimeofday(&curr_time,&tzp); */
2038: rcurr_time = time(NULL);
2039: curr_time = *localtime(&rcurr_time);
2040: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2041: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2042: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2043: for (i=1;i<=n;i++) {
1.126 brouard 2044: printf(" %d %.12f",i, p[i]);
2045: fprintf(ficlog," %d %.12lf",i, p[i]);
2046: fprintf(ficrespow," %.12lf", p[i]);
2047: }
2048: printf("\n");
2049: fprintf(ficlog,"\n");
2050: fprintf(ficrespow,"\n");fflush(ficrespow);
2051: if(*iter <=3){
1.157 brouard 2052: tml = *localtime(&rcurr_time);
2053: strcpy(strcurr,asctime(&tml));
2054: rforecast_time=rcurr_time;
1.126 brouard 2055: itmp = strlen(strcurr);
2056: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.224 brouard 2057: strcurr[itmp-1]='\0';
1.162 brouard 2058: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2059: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2060: for(niterf=10;niterf<=30;niterf+=10){
1.224 brouard 2061: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2062: forecast_time = *localtime(&rforecast_time);
2063: strcpy(strfor,asctime(&forecast_time));
2064: itmp = strlen(strfor);
2065: if(strfor[itmp-1]=='\n')
2066: strfor[itmp-1]='\0';
2067: 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);
2068: 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 2069: }
2070: }
1.187 brouard 2071: for (i=1;i<=n;i++) { /* For each direction i */
2072: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2073: fptt=(*fret);
2074: #ifdef DEBUG
1.203 brouard 2075: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2076: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2077: #endif
1.203 brouard 2078: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2079: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2080: #ifdef LINMINORIGINAL
1.188 brouard 2081: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2082: #else
2083: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2084: flatdir[i]=flat; /* Function is vanishing in that direction i */
2085: #endif
2086: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2087: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2088: /* because that direction will be replaced unless the gain del is small */
2089: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2090: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2091: /* with the new direction. */
2092: del=fabs(fptt-(*fret));
2093: ibig=i;
1.126 brouard 2094: }
2095: #ifdef DEBUG
2096: printf("%d %.12e",i,(*fret));
2097: fprintf(ficlog,"%d %.12e",i,(*fret));
2098: for (j=1;j<=n;j++) {
1.224 brouard 2099: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2100: printf(" x(%d)=%.12e",j,xit[j]);
2101: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2102: }
2103: for(j=1;j<=n;j++) {
1.225 brouard 2104: printf(" p(%d)=%.12e",j,p[j]);
2105: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2106: }
2107: printf("\n");
2108: fprintf(ficlog,"\n");
2109: #endif
1.187 brouard 2110: } /* end loop on each direction i */
2111: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2112: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2113: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2114: for(j=1;j<=n;j++) {
1.225 brouard 2115: if(flatdir[j] >0){
2116: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2117: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2118: }
2119: /* printf("\n"); */
2120: /* fprintf(ficlog,"\n"); */
2121: }
1.182 brouard 2122: if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /* Did we reach enough precision? */
1.188 brouard 2123: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2124: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2125: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2126: /* decreased of more than 3.84 */
2127: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2128: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2129: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2130:
1.188 brouard 2131: /* Starting the program with initial values given by a former maximization will simply change */
2132: /* the scales of the directions and the directions, because the are reset to canonical directions */
2133: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2134: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2135: #ifdef DEBUG
2136: int k[2],l;
2137: k[0]=1;
2138: k[1]=-1;
2139: printf("Max: %.12e",(*func)(p));
2140: fprintf(ficlog,"Max: %.12e",(*func)(p));
2141: for (j=1;j<=n;j++) {
2142: printf(" %.12e",p[j]);
2143: fprintf(ficlog," %.12e",p[j]);
2144: }
2145: printf("\n");
2146: fprintf(ficlog,"\n");
2147: for(l=0;l<=1;l++) {
2148: for (j=1;j<=n;j++) {
2149: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2150: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2151: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2152: }
2153: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2154: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2155: }
2156: #endif
2157:
1.224 brouard 2158: #ifdef LINMINORIGINAL
2159: #else
2160: free_ivector(flatdir,1,n);
2161: #endif
1.126 brouard 2162: free_vector(xit,1,n);
2163: free_vector(xits,1,n);
2164: free_vector(ptt,1,n);
2165: free_vector(pt,1,n);
2166: return;
1.192 brouard 2167: } /* enough precision */
1.126 brouard 2168: if (*iter == ITMAX) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2169: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2170: ptt[j]=2.0*p[j]-pt[j];
2171: xit[j]=p[j]-pt[j];
2172: pt[j]=p[j];
2173: }
1.181 brouard 2174: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2175: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2176: if (*iter <=4) {
1.225 brouard 2177: #else
2178: #endif
1.224 brouard 2179: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2180: #else
1.161 brouard 2181: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2182: #endif
1.162 brouard 2183: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2184: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2185: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2186: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2187: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2188: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2189: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2190: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2191: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2192: /* Even if f3 <f1, directest can be negative and t >0 */
2193: /* mu² and del² are equal when f3=f1 */
2194: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2195: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2196: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2197: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2198: #ifdef NRCORIGINAL
2199: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2200: #else
2201: 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 2202: t= t- del*SQR(fp-fptt);
1.183 brouard 2203: #endif
1.202 brouard 2204: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2205: #ifdef DEBUG
1.181 brouard 2206: 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);
2207: 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 2208: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2209: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2210: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2211: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2212: 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);
2213: 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);
2214: #endif
1.183 brouard 2215: #ifdef POWELLORIGINAL
2216: if (t < 0.0) { /* Then we use it for new direction */
2217: #else
1.182 brouard 2218: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2219: 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 2220: 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 2221: 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 2222: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2223: }
1.181 brouard 2224: if (directest < 0.0) { /* Then we use it for new direction */
2225: #endif
1.191 brouard 2226: #ifdef DEBUGLINMIN
1.224 brouard 2227: printf("Before linmin in direction P%d-P0\n",n);
2228: for (j=1;j<=n;j++) {
2229: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2230: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2231: if(j % ncovmodel == 0){
2232: printf("\n");
2233: fprintf(ficlog,"\n");
2234: }
2235: }
2236: #endif
2237: #ifdef LINMINORIGINAL
2238: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2239: #else
2240: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2241: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2242: #endif
1.224 brouard 2243:
1.191 brouard 2244: #ifdef DEBUGLINMIN
1.224 brouard 2245: for (j=1;j<=n;j++) {
2246: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2247: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2248: if(j % ncovmodel == 0){
2249: printf("\n");
2250: fprintf(ficlog,"\n");
2251: }
2252: }
2253: #endif
2254: for (j=1;j<=n;j++) {
2255: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2256: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2257: }
2258: #ifdef LINMINORIGINAL
2259: #else
1.225 brouard 2260: for (j=1, flatd=0;j<=n;j++) {
2261: if(flatdir[j]>0)
2262: flatd++;
2263: }
2264: if(flatd >0){
2265: printf("%d flat directions\n",flatd);
2266: fprintf(ficlog,"%d flat directions\n",flatd);
2267: for (j=1;j<=n;j++) {
2268: if(flatdir[j]>0){
2269: printf("%d ",j);
2270: fprintf(ficlog,"%d ",j);
2271: }
2272: }
2273: printf("\n");
2274: fprintf(ficlog,"\n");
2275: }
1.191 brouard 2276: #endif
1.224 brouard 2277: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2278: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2279:
1.126 brouard 2280: #ifdef DEBUG
1.224 brouard 2281: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2282: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2283: for(j=1;j<=n;j++){
2284: printf(" %lf",xit[j]);
2285: fprintf(ficlog," %lf",xit[j]);
2286: }
2287: printf("\n");
2288: fprintf(ficlog,"\n");
1.126 brouard 2289: #endif
1.192 brouard 2290: } /* end of t or directest negative */
1.224 brouard 2291: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2292: #else
1.162 brouard 2293: } /* end if (fptt < fp) */
1.192 brouard 2294: #endif
1.225 brouard 2295: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.224 brouard 2296: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2297: #else
1.224 brouard 2298: #endif
1.192 brouard 2299: } /* loop iteration */
1.126 brouard 2300: }
2301:
2302: /**** Prevalence limit (stable or period prevalence) ****************/
2303:
1.203 brouard 2304: double **prevalim(double **prlim, int nlstate, double x[], double age, double **oldm, double **savm, double ftolpl, int *ncvyear, int ij)
1.126 brouard 2305: {
1.218 brouard 2306: /* Computes the prevalence limit in each live state at age x and for covariate ij by left multiplying the unit
1.203 brouard 2307: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2308: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2309: /* Wx is row vector: population in state 1, population in state 2, population dead */
2310: /* or prevalence in state 1, prevalence in state 2, 0 */
2311: /* newm is the matrix after multiplications, its rows are identical at a factor */
2312: /* Initial matrix pimij */
2313: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2314: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2315: /* 0, 0 , 1} */
2316: /*
2317: * and after some iteration: */
2318: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2319: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2320: /* 0, 0 , 1} */
2321: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2322: /* {0.51571254859325999, 0.4842874514067399, */
2323: /* 0.51326036147820708, 0.48673963852179264} */
2324: /* If we start from prlim again, prlim tends to a constant matrix */
2325:
1.126 brouard 2326: int i, ii,j,k;
1.209 brouard 2327: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2328: /* double **matprod2(); */ /* test */
1.218 brouard 2329: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2330: double **newm;
1.209 brouard 2331: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2332: int ncvloop=0;
1.169 brouard 2333:
1.209 brouard 2334: min=vector(1,nlstate);
2335: max=vector(1,nlstate);
2336: meandiff=vector(1,nlstate);
2337:
1.218 brouard 2338: /* Starting with matrix unity */
1.126 brouard 2339: for (ii=1;ii<=nlstate+ndeath;ii++)
2340: for (j=1;j<=nlstate+ndeath;j++){
2341: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2342: }
1.169 brouard 2343:
2344: cov[1]=1.;
2345:
2346: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2347: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2348: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2349: ncvloop++;
1.126 brouard 2350: newm=savm;
2351: /* Covariates have to be included here again */
1.138 brouard 2352: cov[2]=agefin;
1.187 brouard 2353: if(nagesqr==1)
2354: cov[3]= agefin*agefin;;
1.138 brouard 2355: for (k=1; k<=cptcovn;k++) {
1.200 brouard 2356: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.218 brouard 2357: /* Here comes the value of the covariate 'ij' */
1.200 brouard 2358: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
1.198 brouard 2359: /* 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])]); */
1.138 brouard 2360: }
1.186 brouard 2361: /*wrong? for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
1.200 brouard 2362: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]*cov[2]; */
2363: for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2];
1.186 brouard 2364: for (k=1; k<=cptcovprod;k++) /* Useless */
1.200 brouard 2365: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
2366: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
1.138 brouard 2367:
2368: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2369: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2370: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2371: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2372: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2373: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2374: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2375:
1.126 brouard 2376: savm=oldm;
2377: oldm=newm;
1.209 brouard 2378:
2379: for(j=1; j<=nlstate; j++){
2380: max[j]=0.;
2381: min[j]=1.;
2382: }
2383: for(i=1;i<=nlstate;i++){
2384: sumnew=0;
2385: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2386: for(j=1; j<=nlstate; j++){
2387: prlim[i][j]= newm[i][j]/(1-sumnew);
2388: max[j]=FMAX(max[j],prlim[i][j]);
2389: min[j]=FMIN(min[j],prlim[i][j]);
2390: }
2391: }
2392:
1.126 brouard 2393: maxmax=0.;
1.209 brouard 2394: for(j=1; j<=nlstate; j++){
2395: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2396: maxmax=FMAX(maxmax,meandiff[j]);
2397: /* 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 2398: } /* j loop */
1.203 brouard 2399: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2400: /* 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 2401: if(maxmax < ftolpl){
1.209 brouard 2402: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2403: free_vector(min,1,nlstate);
2404: free_vector(max,1,nlstate);
2405: free_vector(meandiff,1,nlstate);
1.126 brouard 2406: return prlim;
2407: }
1.169 brouard 2408: } /* age loop */
1.208 brouard 2409: /* After some age loop it doesn't converge */
1.209 brouard 2410: 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 2411: 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 2412: /* 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); */
2413: free_vector(min,1,nlstate);
2414: free_vector(max,1,nlstate);
2415: free_vector(meandiff,1,nlstate);
1.208 brouard 2416:
1.169 brouard 2417: return prlim; /* should not reach here */
1.126 brouard 2418: }
2419:
1.217 brouard 2420:
2421: /**** Back Prevalence limit (stable or period prevalence) ****************/
2422:
1.218 brouard 2423: /* 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) */
2424: /* 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) */
2425: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij)
1.217 brouard 2426: {
1.218 brouard 2427: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2428: matrix by transitions matrix until convergence is reached with precision ftolpl */
2429: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2430: /* Wx is row vector: population in state 1, population in state 2, population dead */
2431: /* or prevalence in state 1, prevalence in state 2, 0 */
2432: /* newm is the matrix after multiplications, its rows are identical at a factor */
2433: /* Initial matrix pimij */
2434: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2435: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2436: /* 0, 0 , 1} */
2437: /*
2438: * and after some iteration: */
2439: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2440: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2441: /* 0, 0 , 1} */
2442: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2443: /* {0.51571254859325999, 0.4842874514067399, */
2444: /* 0.51326036147820708, 0.48673963852179264} */
2445: /* If we start from prlim again, prlim tends to a constant matrix */
2446:
2447: int i, ii,j,k;
2448: double *min, *max, *meandiff, maxmax,sumnew=0.;
2449: /* double **matprod2(); */ /* test */
2450: double **out, cov[NCOVMAX+1], **bmij();
2451: double **newm;
1.218 brouard 2452: double **dnewm, **doldm, **dsavm; /* for use */
2453: double **oldm, **savm; /* for use */
2454:
1.217 brouard 2455: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2456: int ncvloop=0;
2457:
2458: min=vector(1,nlstate);
2459: max=vector(1,nlstate);
2460: meandiff=vector(1,nlstate);
2461:
1.218 brouard 2462: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2463: oldm=oldms; savm=savms;
2464:
2465: /* Starting with matrix unity */
2466: for (ii=1;ii<=nlstate+ndeath;ii++)
2467: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2468: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2469: }
2470:
2471: cov[1]=1.;
2472:
2473: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2474: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2475: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2476: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2477: ncvloop++;
1.218 brouard 2478: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2479: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2480: /* Covariates have to be included here again */
2481: cov[2]=agefin;
2482: if(nagesqr==1)
2483: cov[3]= agefin*agefin;;
2484: for (k=1; k<=cptcovn;k++) {
2485: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
2486: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
2487: /* 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])]); */
2488: }
2489: /*wrong? for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
2490: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]*cov[2]; */
2491: for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2];
2492: for (k=1; k<=cptcovprod;k++) /* Useless */
2493: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
2494: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2495:
2496: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2497: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2498: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2499: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2500: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2501: /* ij should be linked to the correct index of cov */
2502: /* age and covariate values ij are in 'cov', but we need to pass
2503: * ij for the observed prevalence at age and status and covariate
2504: * number: prevacurrent[(int)agefin][ii][ij]
2505: */
2506: /* 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 *\/ */
2507: /* 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 *\/ */
2508: 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 2509: savm=oldm;
2510: oldm=newm;
2511: for(j=1; j<=nlstate; j++){
2512: max[j]=0.;
2513: min[j]=1.;
2514: }
2515: for(j=1; j<=nlstate; j++){
2516: for(i=1;i<=nlstate;i++){
1.218 brouard 2517: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2518: bprlim[i][j]= newm[i][j];
2519: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2520: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2521: }
2522: }
1.218 brouard 2523:
1.217 brouard 2524: maxmax=0.;
2525: for(i=1; i<=nlstate; i++){
2526: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2527: maxmax=FMAX(maxmax,meandiff[i]);
2528: /* 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); */
2529: } /* j loop */
2530: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2531: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2532: if(maxmax < ftolpl){
1.220 brouard 2533: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2534: free_vector(min,1,nlstate);
2535: free_vector(max,1,nlstate);
2536: free_vector(meandiff,1,nlstate);
2537: return bprlim;
2538: }
2539: } /* age loop */
2540: /* After some age loop it doesn't converge */
2541: 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\
2542: 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);
2543: /* 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); */
2544: free_vector(min,1,nlstate);
2545: free_vector(max,1,nlstate);
2546: free_vector(meandiff,1,nlstate);
2547:
2548: return bprlim; /* should not reach here */
2549: }
2550:
1.126 brouard 2551: /*************** transition probabilities ***************/
2552:
2553: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2554: {
1.138 brouard 2555: /* According to parameters values stored in x and the covariate's values stored in cov,
2556: computes the probability to be observed in state j being in state i by appying the
2557: model to the ncovmodel covariates (including constant and age).
2558: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2559: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2560: ncth covariate in the global vector x is given by the formula:
2561: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2562: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2563: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2564: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2565: Outputs ps[i][j] the probability to be observed in j being in j according to
2566: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2567: */
2568: double s1, lnpijopii;
1.126 brouard 2569: /*double t34;*/
1.164 brouard 2570: int i,j, nc, ii, jj;
1.126 brouard 2571:
1.223 brouard 2572: for(i=1; i<= nlstate; i++){
2573: for(j=1; j<i;j++){
2574: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2575: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2576: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2577: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2578: }
2579: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2580: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2581: }
2582: for(j=i+1; j<=nlstate+ndeath;j++){
2583: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2584: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2585: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2586: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2587: }
2588: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2589: }
2590: }
1.218 brouard 2591:
1.223 brouard 2592: for(i=1; i<= nlstate; i++){
2593: s1=0;
2594: for(j=1; j<i; j++){
2595: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2596: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2597: }
2598: for(j=i+1; j<=nlstate+ndeath; j++){
2599: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2600: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2601: }
2602: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2603: ps[i][i]=1./(s1+1.);
2604: /* Computing other pijs */
2605: for(j=1; j<i; j++)
2606: ps[i][j]= exp(ps[i][j])*ps[i][i];
2607: for(j=i+1; j<=nlstate+ndeath; j++)
2608: ps[i][j]= exp(ps[i][j])*ps[i][i];
2609: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2610: } /* end i */
1.218 brouard 2611:
1.223 brouard 2612: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2613: for(jj=1; jj<= nlstate+ndeath; jj++){
2614: ps[ii][jj]=0;
2615: ps[ii][ii]=1;
2616: }
2617: }
1.218 brouard 2618:
2619:
1.223 brouard 2620: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2621: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2622: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2623: /* } */
2624: /* printf("\n "); */
2625: /* } */
2626: /* printf("\n ");printf("%lf ",cov[2]);*/
2627: /*
2628: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2629: goto end;*/
1.223 brouard 2630: return ps;
1.126 brouard 2631: }
2632:
1.218 brouard 2633: /*************** backward transition probabilities ***************/
2634:
2635: /* 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 ) */
2636: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2637: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2638: {
1.222 brouard 2639: /* Computes the backward probability at age agefin and covariate ij
2640: * and returns in **ps as well as **bmij.
2641: */
1.218 brouard 2642: int i, ii, j,k;
1.222 brouard 2643:
2644: double **out, **pmij();
2645: double sumnew=0.;
1.218 brouard 2646: double agefin;
1.222 brouard 2647:
2648: double **dnewm, **dsavm, **doldm;
2649: double **bbmij;
2650:
1.218 brouard 2651: doldm=ddoldms; /* global pointers */
1.222 brouard 2652: dnewm=ddnewms;
2653: dsavm=ddsavms;
2654:
2655: agefin=cov[2];
2656: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2657: the observed prevalence (with this covariate ij) */
2658: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2659: /* We do have the matrix Px in savm and we need pij */
2660: for (j=1;j<=nlstate+ndeath;j++){
2661: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2662: for (ii=1;ii<=nlstate;ii++){
2663: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2664: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2665: for (ii=1;ii<=nlstate+ndeath;ii++){
2666: if(sumnew >= 1.e-10){
2667: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2668: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2669: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2670: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2671: /* }else */
2672: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2673: }else{
2674: 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);
2675: }
2676: } /*End ii */
2677: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2678: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2679: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2680: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2681: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2682: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2683: /* left Product of this matrix by diag matrix of prevalences (savm) */
2684: for (j=1;j<=nlstate+ndeath;j++){
2685: for (ii=1;ii<=nlstate+ndeath;ii++){
2686: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2687: }
2688: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2689: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2690: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2691: /* end bmij */
2692: return ps;
1.218 brouard 2693: }
1.217 brouard 2694: /*************** transition probabilities ***************/
2695:
1.218 brouard 2696: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2697: {
2698: /* According to parameters values stored in x and the covariate's values stored in cov,
2699: computes the probability to be observed in state j being in state i by appying the
2700: model to the ncovmodel covariates (including constant and age).
2701: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2702: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2703: ncth covariate in the global vector x is given by the formula:
2704: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2705: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2706: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2707: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2708: Outputs ps[i][j] the probability to be observed in j being in j according to
2709: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2710: */
2711: double s1, lnpijopii;
2712: /*double t34;*/
2713: int i,j, nc, ii, jj;
2714:
1.218 brouard 2715: for(i=1; i<= nlstate; i++){
2716: for(j=1; j<i;j++){
2717: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2718: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2719: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2720: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2721: }
2722: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2723: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2724: }
2725: for(j=i+1; j<=nlstate+ndeath;j++){
2726: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2727: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2728: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2729: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2730: }
2731: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2732: }
2733: }
2734:
2735: for(i=1; i<= nlstate; i++){
2736: s1=0;
2737: for(j=1; j<i; j++){
2738: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2739: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2740: }
2741: for(j=i+1; j<=nlstate+ndeath; j++){
2742: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2743: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2744: }
2745: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2746: ps[i][i]=1./(s1+1.);
2747: /* Computing other pijs */
2748: for(j=1; j<i; j++)
2749: ps[i][j]= exp(ps[i][j])*ps[i][i];
2750: for(j=i+1; j<=nlstate+ndeath; j++)
2751: ps[i][j]= exp(ps[i][j])*ps[i][i];
2752: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2753: } /* end i */
2754:
2755: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2756: for(jj=1; jj<= nlstate+ndeath; jj++){
2757: ps[ii][jj]=0;
2758: ps[ii][ii]=1;
2759: }
2760: }
2761: /* Added for backcast */ /* Transposed matrix too */
2762: for(jj=1; jj<= nlstate+ndeath; jj++){
2763: s1=0.;
2764: for(ii=1; ii<= nlstate+ndeath; ii++){
2765: s1+=ps[ii][jj];
2766: }
2767: for(ii=1; ii<= nlstate; ii++){
2768: ps[ii][jj]=ps[ii][jj]/s1;
2769: }
2770: }
2771: /* Transposition */
2772: for(jj=1; jj<= nlstate+ndeath; jj++){
2773: for(ii=jj; ii<= nlstate+ndeath; ii++){
2774: s1=ps[ii][jj];
2775: ps[ii][jj]=ps[jj][ii];
2776: ps[jj][ii]=s1;
2777: }
2778: }
2779: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2780: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2781: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2782: /* } */
2783: /* printf("\n "); */
2784: /* } */
2785: /* printf("\n ");printf("%lf ",cov[2]);*/
2786: /*
2787: for(i=1; i<= npar; i++) printf("%f ",x[i]);
2788: goto end;*/
2789: return ps;
1.217 brouard 2790: }
2791:
2792:
1.126 brouard 2793: /**************** Product of 2 matrices ******************/
2794:
1.145 brouard 2795: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 2796: {
2797: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
2798: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
2799: /* in, b, out are matrice of pointers which should have been initialized
2800: before: only the contents of out is modified. The function returns
2801: a pointer to pointers identical to out */
1.145 brouard 2802: int i, j, k;
1.126 brouard 2803: for(i=nrl; i<= nrh; i++)
1.145 brouard 2804: for(k=ncolol; k<=ncoloh; k++){
2805: out[i][k]=0.;
2806: for(j=ncl; j<=nch; j++)
2807: out[i][k] +=in[i][j]*b[j][k];
2808: }
1.126 brouard 2809: return out;
2810: }
2811:
2812:
2813: /************* Higher Matrix Product ***************/
2814:
2815: double ***hpxij(double ***po, int nhstepm, double age, int hstepm, double *x, int nlstate, int stepm, double **oldm, double **savm, int ij )
2816: {
1.218 brouard 2817: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 2818: 'nhstepm*hstepm*stepm' months (i.e. until
2819: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
2820: nhstepm*hstepm matrices.
2821: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
2822: (typically every 2 years instead of every month which is too big
2823: for the memory).
2824: Model is determined by parameters x and covariates have to be
2825: included manually here.
2826:
2827: */
2828:
2829: int i, j, d, h, k;
1.131 brouard 2830: double **out, cov[NCOVMAX+1];
1.126 brouard 2831: double **newm;
1.187 brouard 2832: double agexact;
1.214 brouard 2833: double agebegin, ageend;
1.126 brouard 2834:
2835: /* Hstepm could be zero and should return the unit matrix */
2836: for (i=1;i<=nlstate+ndeath;i++)
2837: for (j=1;j<=nlstate+ndeath;j++){
2838: oldm[i][j]=(i==j ? 1.0 : 0.0);
2839: po[i][j][0]=(i==j ? 1.0 : 0.0);
2840: }
2841: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2842: for(h=1; h <=nhstepm; h++){
2843: for(d=1; d <=hstepm; d++){
2844: newm=savm;
2845: /* Covariates have to be included here again */
2846: cov[1]=1.;
1.214 brouard 2847: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 2848: cov[2]=agexact;
2849: if(nagesqr==1)
1.227 brouard 2850: cov[3]= agexact*agexact;
1.131 brouard 2851: for (k=1; k<=cptcovn;k++)
1.227 brouard 2852: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
2853: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.186 brouard 2854: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.227 brouard 2855: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
2856: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2857: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.145 brouard 2858: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.227 brouard 2859: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
2860: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])]*nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
2861:
2862:
1.126 brouard 2863: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
2864: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 2865: /* right multiplication of oldm by the current matrix */
1.126 brouard 2866: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
2867: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 2868: /* if((int)age == 70){ */
2869: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
2870: /* for(i=1; i<=nlstate+ndeath; i++) { */
2871: /* printf("%d pmmij ",i); */
2872: /* for(j=1;j<=nlstate+ndeath;j++) { */
2873: /* printf("%f ",pmmij[i][j]); */
2874: /* } */
2875: /* printf(" oldm "); */
2876: /* for(j=1;j<=nlstate+ndeath;j++) { */
2877: /* printf("%f ",oldm[i][j]); */
2878: /* } */
2879: /* printf("\n"); */
2880: /* } */
2881: /* } */
1.126 brouard 2882: savm=oldm;
2883: oldm=newm;
2884: }
2885: for(i=1; i<=nlstate+ndeath; i++)
2886: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 2887: po[i][j][h]=newm[i][j];
2888: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 2889: }
1.128 brouard 2890: /*printf("h=%d ",h);*/
1.126 brouard 2891: } /* end h */
1.218 brouard 2892: /* printf("\n H=%d \n",h); */
1.126 brouard 2893: return po;
2894: }
2895:
1.217 brouard 2896: /************* Higher Back Matrix Product ***************/
1.218 brouard 2897: /* 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 2898: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 2899: {
1.218 brouard 2900: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 2901: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 2902: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
2903: nhstepm*hstepm matrices.
2904: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
2905: (typically every 2 years instead of every month which is too big
1.217 brouard 2906: for the memory).
1.218 brouard 2907: Model is determined by parameters x and covariates have to be
2908: included manually here.
1.217 brouard 2909:
1.222 brouard 2910: */
1.217 brouard 2911:
2912: int i, j, d, h, k;
2913: double **out, cov[NCOVMAX+1];
2914: double **newm;
2915: double agexact;
2916: double agebegin, ageend;
1.222 brouard 2917: double **oldm, **savm;
1.217 brouard 2918:
1.222 brouard 2919: oldm=oldms;savm=savms;
1.217 brouard 2920: /* Hstepm could be zero and should return the unit matrix */
2921: for (i=1;i<=nlstate+ndeath;i++)
2922: for (j=1;j<=nlstate+ndeath;j++){
2923: oldm[i][j]=(i==j ? 1.0 : 0.0);
2924: po[i][j][0]=(i==j ? 1.0 : 0.0);
2925: }
2926: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2927: for(h=1; h <=nhstepm; h++){
2928: for(d=1; d <=hstepm; d++){
2929: newm=savm;
2930: /* Covariates have to be included here again */
2931: cov[1]=1.;
2932: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
2933: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
2934: cov[2]=agexact;
2935: if(nagesqr==1)
1.222 brouard 2936: cov[3]= agexact*agexact;
1.218 brouard 2937: for (k=1; k<=cptcovn;k++)
1.222 brouard 2938: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
2939: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 2940: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 2941: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
2942: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2943: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 2944: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 2945: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
2946: /* 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 2947:
2948:
1.217 brouard 2949: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
2950: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 2951: /* Careful transposed matrix */
1.222 brouard 2952: /* age is in cov[2] */
1.218 brouard 2953: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 2954: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 2955: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 2956: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 2957: /* if((int)age == 70){ */
2958: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
2959: /* for(i=1; i<=nlstate+ndeath; i++) { */
2960: /* printf("%d pmmij ",i); */
2961: /* for(j=1;j<=nlstate+ndeath;j++) { */
2962: /* printf("%f ",pmmij[i][j]); */
2963: /* } */
2964: /* printf(" oldm "); */
2965: /* for(j=1;j<=nlstate+ndeath;j++) { */
2966: /* printf("%f ",oldm[i][j]); */
2967: /* } */
2968: /* printf("\n"); */
2969: /* } */
2970: /* } */
2971: savm=oldm;
2972: oldm=newm;
2973: }
2974: for(i=1; i<=nlstate+ndeath; i++)
2975: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 2976: po[i][j][h]=newm[i][j];
2977: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 2978: }
2979: /*printf("h=%d ",h);*/
2980: } /* end h */
1.222 brouard 2981: /* printf("\n H=%d \n",h); */
1.217 brouard 2982: return po;
2983: }
2984:
2985:
1.162 brouard 2986: #ifdef NLOPT
2987: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
2988: double fret;
2989: double *xt;
2990: int j;
2991: myfunc_data *d2 = (myfunc_data *) pd;
2992: /* xt = (p1-1); */
2993: xt=vector(1,n);
2994: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
2995:
2996: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
2997: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
2998: printf("Function = %.12lf ",fret);
2999: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3000: printf("\n");
3001: free_vector(xt,1,n);
3002: return fret;
3003: }
3004: #endif
1.126 brouard 3005:
3006: /*************** log-likelihood *************/
3007: double func( double *x)
3008: {
1.226 brouard 3009: int i, ii, j, k, mi, d, kk;
3010: int ioffset=0;
3011: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3012: double **out;
3013: double lli; /* Individual log likelihood */
3014: int s1, s2;
1.228 brouard 3015: 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 3016: double bbh, survp;
3017: long ipmx;
3018: double agexact;
3019: /*extern weight */
3020: /* We are differentiating ll according to initial status */
3021: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3022: /*for(i=1;i<imx;i++)
3023: printf(" %d\n",s[4][i]);
3024: */
1.162 brouard 3025:
1.226 brouard 3026: ++countcallfunc;
1.162 brouard 3027:
1.226 brouard 3028: cov[1]=1.;
1.126 brouard 3029:
1.226 brouard 3030: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3031: ioffset=0;
1.226 brouard 3032: if(mle==1){
3033: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3034: /* Computes the values of the ncovmodel covariates of the model
3035: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3036: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3037: to be observed in j being in i according to the model.
3038: */
3039: ioffset=2+nagesqr+cptcovage;
1.233 ! brouard 3040: /* Fixed */
! 3041: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
! 3042: 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)*/
! 3043: }
1.226 brouard 3044: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3045: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3046: has been calculated etc */
3047: /* For an individual i, wav[i] gives the number of effective waves */
3048: /* We compute the contribution to Likelihood of each effective transition
3049: mw[mi][i] is real wave of the mi th effectve wave */
3050: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3051: s2=s[mw[mi+1][i]][i];
3052: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3053: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3054: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3055: */
3056: for(mi=1; mi<= wav[i]-1; mi++){
1.233 ! brouard 3057: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
! 3058: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i];
1.231 brouard 3059: }
3060: for (ii=1;ii<=nlstate+ndeath;ii++)
3061: for (j=1;j<=nlstate+ndeath;j++){
3062: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3063: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3064: }
3065: for(d=0; d<dh[mi][i]; d++){
3066: newm=savm;
3067: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3068: cov[2]=agexact;
3069: if(nagesqr==1)
3070: cov[3]= agexact*agexact; /* Should be changed here */
3071: for (kk=1; kk<=cptcovage;kk++) {
3072: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
3073: }
3074: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3075: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3076: savm=oldm;
3077: oldm=newm;
3078: } /* end mult */
1.224 brouard 3079:
1.231 brouard 3080: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3081: /* But now since version 0.9 we anticipate for bias at large stepm.
3082: * If stepm is larger than one month (smallest stepm) and if the exact delay
3083: * (in months) between two waves is not a multiple of stepm, we rounded to
3084: * the nearest (and in case of equal distance, to the lowest) interval but now
3085: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3086: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3087: * probability in order to take into account the bias as a fraction of the way
3088: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3089: * -stepm/2 to stepm/2 .
3090: * For stepm=1 the results are the same as for previous versions of Imach.
3091: * For stepm > 1 the results are less biased than in previous versions.
3092: */
3093: s1=s[mw[mi][i]][i];
3094: s2=s[mw[mi+1][i]][i];
3095: bbh=(double)bh[mi][i]/(double)stepm;
3096: /* bias bh is positive if real duration
3097: * is higher than the multiple of stepm and negative otherwise.
3098: */
3099: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3100: if( s2 > nlstate){
3101: /* i.e. if s2 is a death state and if the date of death is known
3102: then the contribution to the likelihood is the probability to
3103: die between last step unit time and current step unit time,
3104: which is also equal to probability to die before dh
3105: minus probability to die before dh-stepm .
3106: In version up to 0.92 likelihood was computed
3107: as if date of death was unknown. Death was treated as any other
3108: health state: the date of the interview describes the actual state
3109: and not the date of a change in health state. The former idea was
3110: to consider that at each interview the state was recorded
3111: (healthy, disable or death) and IMaCh was corrected; but when we
3112: introduced the exact date of death then we should have modified
3113: the contribution of an exact death to the likelihood. This new
3114: contribution is smaller and very dependent of the step unit
3115: stepm. It is no more the probability to die between last interview
3116: and month of death but the probability to survive from last
3117: interview up to one month before death multiplied by the
3118: probability to die within a month. Thanks to Chris
3119: Jackson for correcting this bug. Former versions increased
3120: mortality artificially. The bad side is that we add another loop
3121: which slows down the processing. The difference can be up to 10%
3122: lower mortality.
3123: */
3124: /* If, at the beginning of the maximization mostly, the
3125: cumulative probability or probability to be dead is
3126: constant (ie = 1) over time d, the difference is equal to
3127: 0. out[s1][3] = savm[s1][3]: probability, being at state
3128: s1 at precedent wave, to be dead a month before current
3129: wave is equal to probability, being at state s1 at
3130: precedent wave, to be dead at mont of the current
3131: wave. Then the observed probability (that this person died)
3132: is null according to current estimated parameter. In fact,
3133: it should be very low but not zero otherwise the log go to
3134: infinity.
3135: */
1.183 brouard 3136: /* #ifdef INFINITYORIGINAL */
3137: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3138: /* #else */
3139: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3140: /* lli=log(mytinydouble); */
3141: /* else */
3142: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3143: /* #endif */
1.226 brouard 3144: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3145:
1.226 brouard 3146: } else if ( s2==-1 ) { /* alive */
3147: for (j=1,survp=0. ; j<=nlstate; j++)
3148: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3149: /*survp += out[s1][j]; */
3150: lli= log(survp);
3151: }
3152: else if (s2==-4) {
3153: for (j=3,survp=0. ; j<=nlstate; j++)
3154: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3155: lli= log(survp);
3156: }
3157: else if (s2==-5) {
3158: for (j=1,survp=0. ; j<=2; j++)
3159: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3160: lli= log(survp);
3161: }
3162: else{
3163: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3164: /* 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 */
3165: }
3166: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3167: /*if(lli ==000.0)*/
3168: /*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); */
3169: ipmx +=1;
3170: sw += weight[i];
3171: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3172: /* if (lli < log(mytinydouble)){ */
3173: /* 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); */
3174: /* 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]); */
3175: /* } */
3176: } /* end of wave */
3177: } /* end of individual */
3178: } else if(mle==2){
3179: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3180: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3181: for(mi=1; mi<= wav[i]-1; mi++){
3182: for (ii=1;ii<=nlstate+ndeath;ii++)
3183: for (j=1;j<=nlstate+ndeath;j++){
3184: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3185: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3186: }
3187: for(d=0; d<=dh[mi][i]; d++){
3188: newm=savm;
3189: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3190: cov[2]=agexact;
3191: if(nagesqr==1)
3192: cov[3]= agexact*agexact;
3193: for (kk=1; kk<=cptcovage;kk++) {
3194: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3195: }
3196: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3197: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3198: savm=oldm;
3199: oldm=newm;
3200: } /* end mult */
3201:
3202: s1=s[mw[mi][i]][i];
3203: s2=s[mw[mi+1][i]][i];
3204: bbh=(double)bh[mi][i]/(double)stepm;
3205: 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 */
3206: ipmx +=1;
3207: sw += weight[i];
3208: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3209: } /* end of wave */
3210: } /* end of individual */
3211: } else if(mle==3){ /* exponential inter-extrapolation */
3212: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3213: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3214: for(mi=1; mi<= wav[i]-1; mi++){
3215: for (ii=1;ii<=nlstate+ndeath;ii++)
3216: for (j=1;j<=nlstate+ndeath;j++){
3217: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3218: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3219: }
3220: for(d=0; d<dh[mi][i]; d++){
3221: newm=savm;
3222: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3223: cov[2]=agexact;
3224: if(nagesqr==1)
3225: cov[3]= agexact*agexact;
3226: for (kk=1; kk<=cptcovage;kk++) {
3227: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3228: }
3229: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3230: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3231: savm=oldm;
3232: oldm=newm;
3233: } /* end mult */
3234:
3235: s1=s[mw[mi][i]][i];
3236: s2=s[mw[mi+1][i]][i];
3237: bbh=(double)bh[mi][i]/(double)stepm;
3238: 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 */
3239: ipmx +=1;
3240: sw += weight[i];
3241: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3242: } /* end of wave */
3243: } /* end of individual */
3244: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3245: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3246: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3247: for(mi=1; mi<= wav[i]-1; mi++){
3248: for (ii=1;ii<=nlstate+ndeath;ii++)
3249: for (j=1;j<=nlstate+ndeath;j++){
3250: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3251: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3252: }
3253: for(d=0; d<dh[mi][i]; d++){
3254: newm=savm;
3255: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3256: cov[2]=agexact;
3257: if(nagesqr==1)
3258: cov[3]= agexact*agexact;
3259: for (kk=1; kk<=cptcovage;kk++) {
3260: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3261: }
1.126 brouard 3262:
1.226 brouard 3263: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3264: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3265: savm=oldm;
3266: oldm=newm;
3267: } /* end mult */
3268:
3269: s1=s[mw[mi][i]][i];
3270: s2=s[mw[mi+1][i]][i];
3271: if( s2 > nlstate){
3272: lli=log(out[s1][s2] - savm[s1][s2]);
3273: } else if ( s2==-1 ) { /* alive */
3274: for (j=1,survp=0. ; j<=nlstate; j++)
3275: survp += out[s1][j];
3276: lli= log(survp);
3277: }else{
3278: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3279: }
3280: ipmx +=1;
3281: sw += weight[i];
3282: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3283: /* 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 3284: } /* end of wave */
3285: } /* end of individual */
3286: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3287: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3288: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3289: for(mi=1; mi<= wav[i]-1; mi++){
3290: for (ii=1;ii<=nlstate+ndeath;ii++)
3291: for (j=1;j<=nlstate+ndeath;j++){
3292: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3293: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3294: }
3295: for(d=0; d<dh[mi][i]; d++){
3296: newm=savm;
3297: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3298: cov[2]=agexact;
3299: if(nagesqr==1)
3300: cov[3]= agexact*agexact;
3301: for (kk=1; kk<=cptcovage;kk++) {
3302: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3303: }
1.126 brouard 3304:
1.226 brouard 3305: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3306: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3307: savm=oldm;
3308: oldm=newm;
3309: } /* end mult */
3310:
3311: s1=s[mw[mi][i]][i];
3312: s2=s[mw[mi+1][i]][i];
3313: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3314: ipmx +=1;
3315: sw += weight[i];
3316: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3317: /*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]);*/
3318: } /* end of wave */
3319: } /* end of individual */
3320: } /* End of if */
3321: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3322: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3323: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3324: return -l;
1.126 brouard 3325: }
3326:
3327: /*************** log-likelihood *************/
3328: double funcone( double *x)
3329: {
1.228 brouard 3330: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3331: int i, ii, j, k, mi, d, kk;
1.228 brouard 3332: int ioffset=0;
1.131 brouard 3333: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3334: double **out;
3335: double lli; /* Individual log likelihood */
3336: double llt;
3337: int s1, s2;
1.228 brouard 3338: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3339:
1.126 brouard 3340: double bbh, survp;
1.187 brouard 3341: double agexact;
1.214 brouard 3342: double agebegin, ageend;
1.126 brouard 3343: /*extern weight */
3344: /* We are differentiating ll according to initial status */
3345: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3346: /*for(i=1;i<imx;i++)
3347: printf(" %d\n",s[4][i]);
3348: */
3349: cov[1]=1.;
3350:
3351: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3352: ioffset=0;
3353: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.225 brouard 3354: ioffset=2+nagesqr+cptcovage;
1.232 brouard 3355: /* Fixed */
1.224 brouard 3356: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3357: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3358: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3359: 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)*/
3360: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3361: /* cov[2+6]=covar[Tvar[6]][i]; */
3362: /* cov[2+6]=covar[2][i]; V2 */
3363: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3364: /* cov[2+7]=covar[Tvar[7]][i]; */
3365: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3366: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3367: /* cov[2+9]=covar[Tvar[9]][i]; */
3368: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3369: }
1.232 brouard 3370: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3371: /* 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?)*\/ */
3372: /* } */
1.231 brouard 3373: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3374: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3375: /* } */
1.225 brouard 3376:
1.233 ! brouard 3377:
! 3378: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3379: /* Wave varying (but not age varying) */
3380: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.233 ! brouard 3381: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i];
1.232 brouard 3382: }
3383: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.231 brouard 3384: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3385: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
1.232 brouard 3386: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3387: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
1.231 brouard 3388: /* 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 3389: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
3390: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3391: /* /\* 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]); *\/ */
3392: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
3393: /* } */
1.126 brouard 3394: for (ii=1;ii<=nlstate+ndeath;ii++)
1.231 brouard 3395: for (j=1;j<=nlstate+ndeath;j++){
3396: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3397: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3398: }
1.214 brouard 3399:
3400: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3401: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3402: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.231 brouard 3403: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3404: and mw[mi+1][i]. dh depends on stepm.*/
3405: newm=savm;
3406: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3407: cov[2]=agexact;
3408: if(nagesqr==1)
3409: cov[3]= agexact*agexact;
3410: for (kk=1; kk<=cptcovage;kk++) {
3411: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3412: }
3413: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3414: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3415: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3416: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3417: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3418: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3419: savm=oldm;
3420: oldm=newm;
1.126 brouard 3421: } /* end mult */
3422:
3423: s1=s[mw[mi][i]][i];
3424: s2=s[mw[mi+1][i]][i];
1.217 brouard 3425: /* if(s2==-1){ */
3426: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3427: /* /\* exit(1); *\/ */
3428: /* } */
1.126 brouard 3429: bbh=(double)bh[mi][i]/(double)stepm;
3430: /* bias is positive if real duration
3431: * is higher than the multiple of stepm and negative otherwise.
3432: */
3433: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.232 brouard 3434: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3435: } else if ( s2==-1 ) { /* alive */
1.232 brouard 3436: for (j=1,survp=0. ; j<=nlstate; j++)
3437: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3438: lli= log(survp);
1.126 brouard 3439: }else if (mle==1){
1.232 brouard 3440: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3441: } else if(mle==2){
1.232 brouard 3442: 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 3443: } else if(mle==3){ /* exponential inter-extrapolation */
1.232 brouard 3444: 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 3445: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.232 brouard 3446: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3447: } else{ /* mle=0 back to 1 */
1.232 brouard 3448: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3449: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3450: } /* End of if */
3451: ipmx +=1;
3452: sw += weight[i];
3453: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3454: /*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 3455: if(globpr){
1.232 brouard 3456: fprintf(ficresilk,"%9ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3457: %11.6f %11.6f %11.6f ", \
1.232 brouard 3458: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3459: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3460: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3461: llt +=ll[k]*gipmx/gsw;
3462: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3463: }
3464: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3465: }
1.232 brouard 3466: } /* end of wave */
3467: } /* end of individual */
3468: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3469: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3470: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3471: if(globpr==0){ /* First time we count the contributions and weights */
3472: gipmx=ipmx;
3473: gsw=sw;
3474: }
3475: return -l;
1.126 brouard 3476: }
3477:
3478:
3479: /*************** function likelione ***********/
3480: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3481: {
3482: /* This routine should help understanding what is done with
3483: the selection of individuals/waves and
3484: to check the exact contribution to the likelihood.
3485: Plotting could be done.
3486: */
3487: int k;
3488:
3489: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3490: strcpy(fileresilk,"ILK_");
1.202 brouard 3491: strcat(fileresilk,fileresu);
1.126 brouard 3492: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3493: printf("Problem with resultfile: %s\n", fileresilk);
3494: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3495: }
1.214 brouard 3496: 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");
3497: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3498: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3499: for(k=1; k<=nlstate; k++)
3500: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3501: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3502: }
3503:
3504: *fretone=(*funcone)(p);
3505: if(*globpri !=0){
3506: fclose(ficresilk);
1.205 brouard 3507: if (mle ==0)
3508: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3509: else if(mle >=1)
3510: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3511: 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 3512:
1.208 brouard 3513:
3514: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3515: 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 3516: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3517: }
1.207 brouard 3518: 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 3519: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3520: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3521: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3522: fflush(fichtm);
1.205 brouard 3523: }
1.126 brouard 3524: return;
3525: }
3526:
3527:
3528: /*********** Maximum Likelihood Estimation ***************/
3529:
3530: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3531: {
1.165 brouard 3532: int i,j, iter=0;
1.126 brouard 3533: double **xi;
3534: double fret;
3535: double fretone; /* Only one call to likelihood */
3536: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3537:
3538: #ifdef NLOPT
3539: int creturn;
3540: nlopt_opt opt;
3541: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3542: double *lb;
3543: double minf; /* the minimum objective value, upon return */
3544: double * p1; /* Shifted parameters from 0 instead of 1 */
3545: myfunc_data dinst, *d = &dinst;
3546: #endif
3547:
3548:
1.126 brouard 3549: xi=matrix(1,npar,1,npar);
3550: for (i=1;i<=npar;i++)
3551: for (j=1;j<=npar;j++)
3552: xi[i][j]=(i==j ? 1.0 : 0.0);
3553: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3554: strcpy(filerespow,"POW_");
1.126 brouard 3555: strcat(filerespow,fileres);
3556: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3557: printf("Problem with resultfile: %s\n", filerespow);
3558: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3559: }
3560: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3561: for (i=1;i<=nlstate;i++)
3562: for(j=1;j<=nlstate+ndeath;j++)
3563: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3564: fprintf(ficrespow,"\n");
1.162 brouard 3565: #ifdef POWELL
1.126 brouard 3566: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3567: #endif
1.126 brouard 3568:
1.162 brouard 3569: #ifdef NLOPT
3570: #ifdef NEWUOA
3571: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3572: #else
3573: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3574: #endif
3575: lb=vector(0,npar-1);
3576: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3577: nlopt_set_lower_bounds(opt, lb);
3578: nlopt_set_initial_step1(opt, 0.1);
3579:
3580: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3581: d->function = func;
3582: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3583: nlopt_set_min_objective(opt, myfunc, d);
3584: nlopt_set_xtol_rel(opt, ftol);
3585: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3586: printf("nlopt failed! %d\n",creturn);
3587: }
3588: else {
3589: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3590: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3591: iter=1; /* not equal */
3592: }
3593: nlopt_destroy(opt);
3594: #endif
1.126 brouard 3595: free_matrix(xi,1,npar,1,npar);
3596: fclose(ficrespow);
1.203 brouard 3597: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3598: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3599: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3600:
3601: }
3602:
3603: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3604: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3605: {
3606: double **a,**y,*x,pd;
1.203 brouard 3607: /* double **hess; */
1.164 brouard 3608: int i, j;
1.126 brouard 3609: int *indx;
3610:
3611: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3612: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3613: void lubksb(double **a, int npar, int *indx, double b[]) ;
3614: void ludcmp(double **a, int npar, int *indx, double *d) ;
3615: double gompertz(double p[]);
1.203 brouard 3616: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3617:
3618: printf("\nCalculation of the hessian matrix. Wait...\n");
3619: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3620: for (i=1;i<=npar;i++){
1.203 brouard 3621: printf("%d-",i);fflush(stdout);
3622: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3623:
3624: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3625:
3626: /* printf(" %f ",p[i]);
3627: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3628: }
3629:
3630: for (i=1;i<=npar;i++) {
3631: for (j=1;j<=npar;j++) {
3632: if (j>i) {
1.203 brouard 3633: printf(".%d-%d",i,j);fflush(stdout);
3634: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3635: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3636:
3637: hess[j][i]=hess[i][j];
3638: /*printf(" %lf ",hess[i][j]);*/
3639: }
3640: }
3641: }
3642: printf("\n");
3643: fprintf(ficlog,"\n");
3644:
3645: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3646: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3647:
3648: a=matrix(1,npar,1,npar);
3649: y=matrix(1,npar,1,npar);
3650: x=vector(1,npar);
3651: indx=ivector(1,npar);
3652: for (i=1;i<=npar;i++)
3653: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3654: ludcmp(a,npar,indx,&pd);
3655:
3656: for (j=1;j<=npar;j++) {
3657: for (i=1;i<=npar;i++) x[i]=0;
3658: x[j]=1;
3659: lubksb(a,npar,indx,x);
3660: for (i=1;i<=npar;i++){
3661: matcov[i][j]=x[i];
3662: }
3663: }
3664:
3665: printf("\n#Hessian matrix#\n");
3666: fprintf(ficlog,"\n#Hessian matrix#\n");
3667: for (i=1;i<=npar;i++) {
3668: for (j=1;j<=npar;j++) {
1.203 brouard 3669: printf("%.6e ",hess[i][j]);
3670: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3671: }
3672: printf("\n");
3673: fprintf(ficlog,"\n");
3674: }
3675:
1.203 brouard 3676: /* printf("\n#Covariance matrix#\n"); */
3677: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3678: /* for (i=1;i<=npar;i++) { */
3679: /* for (j=1;j<=npar;j++) { */
3680: /* printf("%.6e ",matcov[i][j]); */
3681: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3682: /* } */
3683: /* printf("\n"); */
3684: /* fprintf(ficlog,"\n"); */
3685: /* } */
3686:
1.126 brouard 3687: /* Recompute Inverse */
1.203 brouard 3688: /* for (i=1;i<=npar;i++) */
3689: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3690: /* ludcmp(a,npar,indx,&pd); */
3691:
3692: /* printf("\n#Hessian matrix recomputed#\n"); */
3693:
3694: /* for (j=1;j<=npar;j++) { */
3695: /* for (i=1;i<=npar;i++) x[i]=0; */
3696: /* x[j]=1; */
3697: /* lubksb(a,npar,indx,x); */
3698: /* for (i=1;i<=npar;i++){ */
3699: /* y[i][j]=x[i]; */
3700: /* printf("%.3e ",y[i][j]); */
3701: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3702: /* } */
3703: /* printf("\n"); */
3704: /* fprintf(ficlog,"\n"); */
3705: /* } */
3706:
3707: /* Verifying the inverse matrix */
3708: #ifdef DEBUGHESS
3709: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3710:
1.203 brouard 3711: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3712: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3713:
3714: for (j=1;j<=npar;j++) {
3715: for (i=1;i<=npar;i++){
1.203 brouard 3716: printf("%.2f ",y[i][j]);
3717: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3718: }
3719: printf("\n");
3720: fprintf(ficlog,"\n");
3721: }
1.203 brouard 3722: #endif
1.126 brouard 3723:
3724: free_matrix(a,1,npar,1,npar);
3725: free_matrix(y,1,npar,1,npar);
3726: free_vector(x,1,npar);
3727: free_ivector(indx,1,npar);
1.203 brouard 3728: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3729:
3730:
3731: }
3732:
3733: /*************** hessian matrix ****************/
3734: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3735: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3736: int i;
3737: int l=1, lmax=20;
1.203 brouard 3738: double k1,k2, res, fx;
1.132 brouard 3739: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3740: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3741: int k=0,kmax=10;
3742: double l1;
3743:
3744: fx=func(x);
3745: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3746: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3747: l1=pow(10,l);
3748: delts=delt;
3749: for(k=1 ; k <kmax; k=k+1){
3750: delt = delta*(l1*k);
3751: p2[theta]=x[theta] +delt;
1.145 brouard 3752: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 3753: p2[theta]=x[theta]-delt;
3754: k2=func(p2)-fx;
3755: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 3756: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 3757:
1.203 brouard 3758: #ifdef DEBUGHESSII
1.126 brouard 3759: 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);
3760: 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);
3761: #endif
3762: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
3763: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
3764: k=kmax;
3765: }
3766: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 3767: k=kmax; l=lmax*10;
1.126 brouard 3768: }
3769: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
3770: delts=delt;
3771: }
1.203 brouard 3772: } /* End loop k */
1.126 brouard 3773: }
3774: delti[theta]=delts;
3775: return res;
3776:
3777: }
3778:
1.203 brouard 3779: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 3780: {
3781: int i;
1.164 brouard 3782: int l=1, lmax=20;
1.126 brouard 3783: double k1,k2,k3,k4,res,fx;
1.132 brouard 3784: double p2[MAXPARM+1];
1.203 brouard 3785: int k, kmax=1;
3786: double v1, v2, cv12, lc1, lc2;
1.208 brouard 3787:
3788: int firstime=0;
1.203 brouard 3789:
1.126 brouard 3790: fx=func(x);
1.203 brouard 3791: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 3792: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 3793: p2[thetai]=x[thetai]+delti[thetai]*k;
3794: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 3795: k1=func(p2)-fx;
3796:
1.203 brouard 3797: p2[thetai]=x[thetai]+delti[thetai]*k;
3798: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 3799: k2=func(p2)-fx;
3800:
1.203 brouard 3801: p2[thetai]=x[thetai]-delti[thetai]*k;
3802: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 3803: k3=func(p2)-fx;
3804:
1.203 brouard 3805: p2[thetai]=x[thetai]-delti[thetai]*k;
3806: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 3807: k4=func(p2)-fx;
1.203 brouard 3808: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
3809: if(k1*k2*k3*k4 <0.){
1.208 brouard 3810: firstime=1;
1.203 brouard 3811: kmax=kmax+10;
1.208 brouard 3812: }
3813: if(kmax >=10 || firstime ==1){
1.218 brouard 3814: 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);
3815: 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 3816: 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);
3817: 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);
3818: }
3819: #ifdef DEBUGHESSIJ
3820: v1=hess[thetai][thetai];
3821: v2=hess[thetaj][thetaj];
3822: cv12=res;
3823: /* Computing eigen value of Hessian matrix */
3824: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
3825: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
3826: if ((lc2 <0) || (lc1 <0) ){
3827: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
3828: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
3829: 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);
3830: 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);
3831: }
1.126 brouard 3832: #endif
3833: }
3834: return res;
3835: }
3836:
1.203 brouard 3837: /* Not done yet: Was supposed to fix if not exactly at the maximum */
3838: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
3839: /* { */
3840: /* int i; */
3841: /* int l=1, lmax=20; */
3842: /* double k1,k2,k3,k4,res,fx; */
3843: /* double p2[MAXPARM+1]; */
3844: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
3845: /* int k=0,kmax=10; */
3846: /* double l1; */
3847:
3848: /* fx=func(x); */
3849: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
3850: /* l1=pow(10,l); */
3851: /* delts=delt; */
3852: /* for(k=1 ; k <kmax; k=k+1){ */
3853: /* delt = delti*(l1*k); */
3854: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
3855: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
3856: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
3857: /* k1=func(p2)-fx; */
3858:
3859: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
3860: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
3861: /* k2=func(p2)-fx; */
3862:
3863: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
3864: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
3865: /* k3=func(p2)-fx; */
3866:
3867: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
3868: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
3869: /* k4=func(p2)-fx; */
3870: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
3871: /* #ifdef DEBUGHESSIJ */
3872: /* 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); */
3873: /* 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); */
3874: /* #endif */
3875: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
3876: /* k=kmax; */
3877: /* } */
3878: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
3879: /* k=kmax; l=lmax*10; */
3880: /* } */
3881: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
3882: /* delts=delt; */
3883: /* } */
3884: /* } /\* End loop k *\/ */
3885: /* } */
3886: /* delti[theta]=delts; */
3887: /* return res; */
3888: /* } */
3889:
3890:
1.126 brouard 3891: /************** Inverse of matrix **************/
3892: void ludcmp(double **a, int n, int *indx, double *d)
3893: {
3894: int i,imax,j,k;
3895: double big,dum,sum,temp;
3896: double *vv;
3897:
3898: vv=vector(1,n);
3899: *d=1.0;
3900: for (i=1;i<=n;i++) {
3901: big=0.0;
3902: for (j=1;j<=n;j++)
3903: if ((temp=fabs(a[i][j])) > big) big=temp;
3904: if (big == 0.0) nrerror("Singular matrix in routine ludcmp");
3905: vv[i]=1.0/big;
3906: }
3907: for (j=1;j<=n;j++) {
3908: for (i=1;i<j;i++) {
3909: sum=a[i][j];
3910: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
3911: a[i][j]=sum;
3912: }
3913: big=0.0;
3914: for (i=j;i<=n;i++) {
3915: sum=a[i][j];
3916: for (k=1;k<j;k++)
3917: sum -= a[i][k]*a[k][j];
3918: a[i][j]=sum;
3919: if ( (dum=vv[i]*fabs(sum)) >= big) {
3920: big=dum;
3921: imax=i;
3922: }
3923: }
3924: if (j != imax) {
3925: for (k=1;k<=n;k++) {
3926: dum=a[imax][k];
3927: a[imax][k]=a[j][k];
3928: a[j][k]=dum;
3929: }
3930: *d = -(*d);
3931: vv[imax]=vv[j];
3932: }
3933: indx[j]=imax;
3934: if (a[j][j] == 0.0) a[j][j]=TINY;
3935: if (j != n) {
3936: dum=1.0/(a[j][j]);
3937: for (i=j+1;i<=n;i++) a[i][j] *= dum;
3938: }
3939: }
3940: free_vector(vv,1,n); /* Doesn't work */
3941: ;
3942: }
3943:
3944: void lubksb(double **a, int n, int *indx, double b[])
3945: {
3946: int i,ii=0,ip,j;
3947: double sum;
3948:
3949: for (i=1;i<=n;i++) {
3950: ip=indx[i];
3951: sum=b[ip];
3952: b[ip]=b[i];
3953: if (ii)
3954: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
3955: else if (sum) ii=i;
3956: b[i]=sum;
3957: }
3958: for (i=n;i>=1;i--) {
3959: sum=b[i];
3960: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
3961: b[i]=sum/a[i][i];
3962: }
3963: }
3964:
3965: void pstamp(FILE *fichier)
3966: {
1.196 brouard 3967: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 3968: }
3969:
3970: /************ Frequencies ********************/
1.226 brouard 3971: void freqsummary(char fileres[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
3972: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
3973: int firstpass, int lastpass, int stepm, int weightopt, char model[])
3974: { /* Some frequencies */
3975:
1.227 brouard 3976: int i, m, jk, j1, bool, z1,j, k, iv;
1.226 brouard 3977: int iind=0, iage=0;
3978: int mi; /* Effective wave */
3979: int first;
3980: double ***freq; /* Frequencies */
3981: double *meanq;
3982: double **meanqt;
3983: double *pp, **prop, *posprop, *pospropt;
3984: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
3985: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
3986: double agebegin, ageend;
3987:
3988: pp=vector(1,nlstate);
3989: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
3990: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
3991: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
3992: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
3993: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
3994: meanqt=matrix(1,lastpass,1,nqtveff);
3995: strcpy(fileresp,"P_");
3996: strcat(fileresp,fileresu);
3997: /*strcat(fileresphtm,fileresu);*/
3998: if((ficresp=fopen(fileresp,"w"))==NULL) {
3999: printf("Problem with prevalence resultfile: %s\n", fileresp);
4000: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4001: exit(0);
4002: }
1.214 brouard 4003:
1.226 brouard 4004: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4005: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4006: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4007: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4008: fflush(ficlog);
4009: exit(70);
4010: }
4011: else{
4012: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.214 brouard 4013: <hr size=\"2\" color=\"#EC5E5E\"> \n\
4014: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4015: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4016: }
4017: 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 4018:
1.226 brouard 4019: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4020: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4021: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4022: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4023: fflush(ficlog);
4024: exit(70);
4025: }
4026: else{
4027: 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 4028: <hr size=\"2\" color=\"#EC5E5E\"> \n\
4029: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4030: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4031: }
4032: 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 4033:
1.226 brouard 4034: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4035: j1=0;
1.126 brouard 4036:
1.227 brouard 4037: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4038: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4039: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.220 brouard 4040:
1.226 brouard 4041: first=1;
1.220 brouard 4042:
1.226 brouard 4043: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4044: reference=low_education V1=0,V2=0
4045: med_educ V1=1 V2=0,
4046: high_educ V1=0 V2=1
4047: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4048: */
1.126 brouard 4049:
1.227 brouard 4050: 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 4051: posproptt=0.;
4052: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4053: scanf("%d", i);*/
4054: for (i=-5; i<=nlstate+ndeath; i++)
4055: for (jk=-5; jk<=nlstate+ndeath; jk++)
1.231 brouard 4056: for(m=iagemin; m <= iagemax+3; m++)
4057: freq[i][jk][m]=0;
4058:
1.226 brouard 4059: for (i=1; i<=nlstate; i++) {
4060: for(m=iagemin; m <= iagemax+3; m++)
1.231 brouard 4061: prop[i][m]=0;
1.226 brouard 4062: posprop[i]=0;
4063: pospropt[i]=0;
4064: }
1.227 brouard 4065: /* for (z1=1; z1<= nqfveff; z1++) { */
4066: /* meanq[z1]+=0.; */
4067: /* for(m=1;m<=lastpass;m++){ */
4068: /* meanqt[m][z1]=0.; */
4069: /* } */
4070: /* } */
1.231 brouard 4071:
1.226 brouard 4072: dateintsum=0;
4073: k2cpt=0;
1.227 brouard 4074: /* For that combination of covariate j1, we count and print the frequencies in one pass */
1.226 brouard 4075: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4076: bool=1;
1.227 brouard 4077: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
1.231 brouard 4078: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
1.227 brouard 4079: /* for (z1=1; z1<= nqfveff; z1++) { */
4080: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4081: /* } */
1.231 brouard 4082: for (z1=1; z1<=cptcoveff; z1++) {
4083: /* if(Tvaraff[z1] ==-20){ */
4084: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4085: /* }else if(Tvaraff[z1] ==-10){ */
4086: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4087: /* }else */
4088: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){
4089: /* Tests if this individual iind responded to j1 (V4=1 V3=0) */
4090: bool=0;
4091: /* 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",
4092: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4093: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4094: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4095: } /* Onlyf fixed */
4096: } /* end z1 */
4097: } /* cptcovn > 0 */
1.227 brouard 4098: } /* end any */
4099: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
1.231 brouard 4100: /* for(m=firstpass; m<=lastpass; m++){ */
4101: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4102: m=mw[mi][iind];
4103: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4104: for (z1=1; z1<=cptcoveff; z1++) {
4105: if( Fixed[Tmodelind[z1]]==1){
4106: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4107: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4108: bool=0;
4109: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4110: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4111: bool=0;
4112: }
4113: }
4114: }
4115: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4116: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4117: if(bool==1){
4118: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4119: and mw[mi+1][iind]. dh depends on stepm. */
4120: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4121: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4122: if(m >=firstpass && m <=lastpass){
4123: k2=anint[m][iind]+(mint[m][iind]/12.);
4124: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4125: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4126: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4127: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4128: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4129: if (m<lastpass) {
4130: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4131: /* 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]); */
4132: if(s[m][iind]==-1)
4133: 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.));
4134: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4135: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4136: 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 */
4137: }
4138: } /* end if between passes */
4139: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99)) {
4140: dateintsum=dateintsum+k2;
4141: k2cpt++;
4142: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
4143: }
4144: } /* end bool 2 */
4145: } /* end m */
1.226 brouard 4146: } /* end bool */
4147: } /* end iind = 1 to imx */
4148: /* prop[s][age] is feeded for any initial and valid live state as well as
4149: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
1.231 brouard 4150:
4151:
1.226 brouard 4152: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4153: pstamp(ficresp);
1.227 brouard 4154: /* if (ncoveff>0) { */
4155: if (cptcoveff>0) {
1.226 brouard 4156: fprintf(ficresp, "\n#********** Variable ");
4157: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4158: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
1.227 brouard 4159: for (z1=1; z1<=cptcoveff; z1++){
1.231 brouard 4160: fprintf(ficresp, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4161: fprintf(ficresphtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4162: fprintf(ficresphtmfr, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.226 brouard 4163: }
4164: fprintf(ficresp, "**********\n#");
4165: fprintf(ficresphtm, "**********</h3>\n");
4166: fprintf(ficresphtmfr, "**********</h3>\n");
4167: fprintf(ficlog, "\n#********** Variable ");
1.227 brouard 4168: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficlog, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.226 brouard 4169: fprintf(ficlog, "**********\n");
4170: }
4171: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4172: for(i=1; i<=nlstate;i++) {
4173: fprintf(ficresp, " Age Prev(%d) N(%d) N",i,i);
4174: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4175: }
4176: fprintf(ficresp, "\n");
4177: fprintf(ficresphtm, "\n");
1.231 brouard 4178:
1.226 brouard 4179: /* Header of frequency table by age */
4180: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4181: fprintf(ficresphtmfr,"<th>Age</th> ");
4182: for(jk=-1; jk <=nlstate+ndeath; jk++){
4183: for(m=-1; m <=nlstate+ndeath; m++){
1.231 brouard 4184: if(jk!=0 && m!=0)
4185: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.226 brouard 4186: }
4187: }
4188: fprintf(ficresphtmfr, "\n");
1.231 brouard 4189:
1.226 brouard 4190: /* For each age */
4191: for(iage=iagemin; iage <= iagemax+3; iage++){
4192: fprintf(ficresphtm,"<tr>");
4193: if(iage==iagemax+1){
1.231 brouard 4194: fprintf(ficlog,"1");
4195: fprintf(ficresphtmfr,"<tr><th>0</th> ");
1.226 brouard 4196: }else if(iage==iagemax+2){
1.231 brouard 4197: fprintf(ficlog,"0");
4198: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
1.226 brouard 4199: }else if(iage==iagemax+3){
1.231 brouard 4200: fprintf(ficlog,"Total");
4201: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
1.226 brouard 4202: }else{
1.231 brouard 4203: if(first==1){
4204: first=0;
4205: printf("See log file for details...\n");
4206: }
4207: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4208: fprintf(ficlog,"Age %d", iage);
1.226 brouard 4209: }
4210: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4211: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4212: pp[jk] += freq[jk][m][iage];
1.226 brouard 4213: }
4214: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4215: for(m=-1, pos=0; m <=0 ; m++)
4216: pos += freq[jk][m][iage];
4217: if(pp[jk]>=1.e-10){
4218: if(first==1){
4219: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4220: }
4221: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4222: }else{
4223: if(first==1)
4224: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4225: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4226: }
1.226 brouard 4227: }
1.231 brouard 4228:
1.226 brouard 4229: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4230: /* posprop[jk]=0; */
4231: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4232: pp[jk] += freq[jk][m][iage];
1.226 brouard 4233: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
1.231 brouard 4234:
1.226 brouard 4235: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
1.231 brouard 4236: pos += pp[jk]; /* pos is the total number of transitions until this age */
4237: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4238: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4239: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
4240: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
1.226 brouard 4241: }
4242: for(jk=1; jk <=nlstate ; jk++){
1.231 brouard 4243: if(pos>=1.e-5){
4244: if(first==1)
4245: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4246: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4247: }else{
4248: if(first==1)
4249: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4250: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4251: }
4252: if( iage <= iagemax){
4253: if(pos>=1.e-5){
4254: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4255: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4256: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4257: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4258: }
4259: else{
4260: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4261: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4262: }
4263: }
4264: pospropt[jk] +=posprop[jk];
1.226 brouard 4265: } /* end loop jk */
4266: /* pospropt=0.; */
4267: for(jk=-1; jk <=nlstate+ndeath; jk++){
1.231 brouard 4268: for(m=-1; m <=nlstate+ndeath; m++){
4269: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4270: if(first==1){
4271: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4272: }
4273: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4274: }
4275: if(jk!=0 && m!=0)
4276: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
4277: }
1.226 brouard 4278: } /* end loop jk */
4279: posproptt=0.;
4280: for(jk=1; jk <=nlstate; jk++){
1.231 brouard 4281: posproptt += pospropt[jk];
1.226 brouard 4282: }
4283: fprintf(ficresphtmfr,"</tr>\n ");
4284: if(iage <= iagemax){
1.231 brouard 4285: fprintf(ficresp,"\n");
4286: fprintf(ficresphtm,"</tr>\n");
1.226 brouard 4287: }
4288: if(first==1)
1.231 brouard 4289: printf("Others in log...\n");
1.226 brouard 4290: fprintf(ficlog,"\n");
4291: } /* end loop age iage */
4292: fprintf(ficresphtm,"<tr><th>Tot</th>");
4293: for(jk=1; jk <=nlstate ; jk++){
4294: if(posproptt < 1.e-5){
1.231 brouard 4295: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
1.226 brouard 4296: }else{
1.231 brouard 4297: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.226 brouard 4298: }
4299: }
4300: fprintf(ficresphtm,"</tr>\n");
4301: fprintf(ficresphtm,"</table>\n");
4302: fprintf(ficresphtmfr,"</table>\n");
4303: if(posproptt < 1.e-5){
4304: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4305: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4306: fprintf(ficres,"\n This combination (%d) is not valid and no result will be produced\n\n",j1);
4307: invalidvarcomb[j1]=1;
4308: }else{
4309: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4310: invalidvarcomb[j1]=0;
4311: }
4312: fprintf(ficresphtmfr,"</table>\n");
4313: } /* end selected combination of covariate j1 */
4314: dateintmean=dateintsum/k2cpt;
1.231 brouard 4315:
1.226 brouard 4316: fclose(ficresp);
4317: fclose(ficresphtm);
4318: fclose(ficresphtmfr);
4319: free_vector(meanq,1,nqfveff);
4320: free_matrix(meanqt,1,lastpass,1,nqtveff);
4321: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4322: free_vector(pospropt,1,nlstate);
4323: free_vector(posprop,1,nlstate);
4324: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4325: free_vector(pp,1,nlstate);
4326: /* End of freqsummary */
4327: }
1.126 brouard 4328:
4329: /************ Prevalence ********************/
1.227 brouard 4330: 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)
4331: {
4332: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4333: in each health status at the date of interview (if between dateprev1 and dateprev2).
4334: We still use firstpass and lastpass as another selection.
4335: */
1.126 brouard 4336:
1.227 brouard 4337: int i, m, jk, j1, bool, z1,j, iv;
4338: int mi; /* Effective wave */
4339: int iage;
4340: double agebegin, ageend;
4341:
4342: double **prop;
4343: double posprop;
4344: double y2; /* in fractional years */
4345: int iagemin, iagemax;
4346: int first; /** to stop verbosity which is redirected to log file */
4347:
4348: iagemin= (int) agemin;
4349: iagemax= (int) agemax;
4350: /*pp=vector(1,nlstate);*/
4351: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4352: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4353: j1=0;
1.222 brouard 4354:
1.227 brouard 4355: /*j=cptcoveff;*/
4356: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4357:
1.227 brouard 4358: first=1;
4359: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4360: for (i=1; i<=nlstate; i++)
4361: for(iage=iagemin-AGEMARGE; iage <= iagemax+3+AGEMARGE; iage++)
4362: prop[i][iage]=0.0;
4363: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4364: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4365: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4366:
4367: for (i=1; i<=imx; i++) { /* Each individual */
4368: bool=1;
4369: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4370: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4371: m=mw[mi][i];
4372: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4373: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4374: for (z1=1; z1<=cptcoveff; z1++){
4375: if( Fixed[Tmodelind[z1]]==1){
4376: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4377: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4378: bool=0;
4379: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4380: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4381: bool=0;
4382: }
4383: }
4384: if(bool==1){ /* Otherwise we skip that wave/person */
4385: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4386: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4387: if(m >=firstpass && m <=lastpass){
4388: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4389: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4390: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4391: if(agev[m][i]==1) agev[m][i]=iagemax+2;
4392: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+3+AGEMARGE){
4393: 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);
4394: exit(1);
4395: }
4396: if (s[m][i]>0 && s[m][i]<=nlstate) {
4397: /*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]]);*/
4398: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4399: prop[s[m][i]][iagemax+3] += weight[i];
4400: } /* end valid statuses */
4401: } /* end selection of dates */
4402: } /* end selection of waves */
4403: } /* end bool */
4404: } /* end wave */
4405: } /* end individual */
4406: for(i=iagemin; i <= iagemax+3; i++){
4407: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4408: posprop += prop[jk][i];
4409: }
4410:
4411: for(jk=1; jk <=nlstate ; jk++){
4412: if( i <= iagemax){
4413: if(posprop>=1.e-5){
4414: probs[i][jk][j1]= prop[jk][i]/posprop;
4415: } else{
4416: if(first==1){
4417: first=0;
4418: 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]);
4419: }
4420: }
4421: }
4422: }/* end jk */
4423: }/* end i */
1.222 brouard 4424: /*} *//* end i1 */
1.227 brouard 4425: } /* end j1 */
1.222 brouard 4426:
1.227 brouard 4427: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4428: /*free_vector(pp,1,nlstate);*/
4429: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4430: } /* End of prevalence */
1.126 brouard 4431:
4432: /************* Waves Concatenation ***************/
4433:
4434: 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)
4435: {
4436: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4437: Death is a valid wave (if date is known).
4438: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4439: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4440: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4441: */
1.126 brouard 4442:
1.224 brouard 4443: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4444: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4445: double sum=0., jmean=0.;*/
1.224 brouard 4446: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4447: int j, k=0,jk, ju, jl;
4448: double sum=0.;
4449: first=0;
1.214 brouard 4450: firstwo=0;
1.217 brouard 4451: firsthree=0;
1.218 brouard 4452: firstfour=0;
1.164 brouard 4453: jmin=100000;
1.126 brouard 4454: jmax=-1;
4455: jmean=0.;
1.224 brouard 4456:
4457: /* Treating live states */
1.214 brouard 4458: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4459: mi=0; /* First valid wave */
1.227 brouard 4460: mli=0; /* Last valid wave */
1.126 brouard 4461: m=firstpass;
1.214 brouard 4462: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4463: 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 */
4464: mli=m-1;/* mw[++mi][i]=m-1; */
4465: }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 */
4466: mw[++mi][i]=m;
4467: mli=m;
1.224 brouard 4468: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4469: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4470: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4471: }
1.227 brouard 4472: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4473: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4474: break;
1.224 brouard 4475: #else
1.227 brouard 4476: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4477: if(firsthree == 0){
4478: 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);
4479: firsthree=1;
4480: }
4481: 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);
4482: mw[++mi][i]=m;
4483: mli=m;
4484: }
4485: if(s[m][i]==-2){ /* Vital status is really unknown */
4486: nbwarn++;
4487: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4488: 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);
4489: 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);
4490: }
4491: break;
4492: }
4493: break;
1.224 brouard 4494: #endif
1.227 brouard 4495: }/* End m >= lastpass */
1.126 brouard 4496: }/* end while */
1.224 brouard 4497:
1.227 brouard 4498: /* 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 4499: /* After last pass */
1.224 brouard 4500: /* Treating death states */
1.214 brouard 4501: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4502: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4503: /* } */
1.126 brouard 4504: mi++; /* Death is another wave */
4505: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4506: /* Only death is a correct wave */
1.126 brouard 4507: mw[mi][i]=m;
1.224 brouard 4508: }
4509: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.227 brouard 4510: 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 4511: /* m++; */
4512: /* mi++; */
4513: /* s[m][i]=nlstate+1; /\* We are setting the status to the last of non live state *\/ */
4514: /* mw[mi][i]=m; */
1.218 brouard 4515: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4516: 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 */
4517: nbwarn++;
4518: if(firstfiv==0){
4519: 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 );
4520: firstfiv=1;
4521: }else{
4522: 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 );
4523: }
4524: }else{ /* Death occured afer last wave potential bias */
4525: nberr++;
4526: if(firstwo==0){
4527: 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 );
4528: firstwo=1;
4529: }
4530: 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 );
4531: }
1.218 brouard 4532: }else{ /* end date of interview is known */
1.227 brouard 4533: /* death is known but not confirmed by death status at any wave */
4534: if(firstfour==0){
4535: 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 );
4536: firstfour=1;
4537: }
4538: 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 4539: }
1.224 brouard 4540: } /* end if date of death is known */
4541: #endif
4542: wav[i]=mi; /* mi should be the last effective wave (or mli) */
4543: /* wav[i]=mw[mi][i]; */
1.126 brouard 4544: if(mi==0){
4545: nbwarn++;
4546: if(first==0){
1.227 brouard 4547: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
4548: first=1;
1.126 brouard 4549: }
4550: if(first==1){
1.227 brouard 4551: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 4552: }
4553: } /* end mi==0 */
4554: } /* End individuals */
1.214 brouard 4555: /* wav and mw are no more changed */
1.223 brouard 4556:
1.214 brouard 4557:
1.126 brouard 4558: for(i=1; i<=imx; i++){
4559: for(mi=1; mi<wav[i];mi++){
4560: if (stepm <=0)
1.227 brouard 4561: dh[mi][i]=1;
1.126 brouard 4562: else{
1.227 brouard 4563: if (s[mw[mi+1][i]][i] > nlstate) { /* A death */
4564: if (agedc[i] < 2*AGESUP) {
4565: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
4566: if(j==0) j=1; /* Survives at least one month after exam */
4567: else if(j<0){
4568: nberr++;
4569: 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]);
4570: j=1; /* Temporary Dangerous patch */
4571: 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);
4572: 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]);
4573: 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);
4574: }
4575: k=k+1;
4576: if (j >= jmax){
4577: jmax=j;
4578: ijmax=i;
4579: }
4580: if (j <= jmin){
4581: jmin=j;
4582: ijmin=i;
4583: }
4584: sum=sum+j;
4585: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
4586: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
4587: }
4588: }
4589: else{
4590: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 4591: /* 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 4592:
1.227 brouard 4593: k=k+1;
4594: if (j >= jmax) {
4595: jmax=j;
4596: ijmax=i;
4597: }
4598: else if (j <= jmin){
4599: jmin=j;
4600: ijmin=i;
4601: }
4602: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
4603: /*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]);*/
4604: if(j<0){
4605: nberr++;
4606: 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]);
4607: 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]);
4608: }
4609: sum=sum+j;
4610: }
4611: jk= j/stepm;
4612: jl= j -jk*stepm;
4613: ju= j -(jk+1)*stepm;
4614: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
4615: if(jl==0){
4616: dh[mi][i]=jk;
4617: bh[mi][i]=0;
4618: }else{ /* We want a negative bias in order to only have interpolation ie
4619: * to avoid the price of an extra matrix product in likelihood */
4620: dh[mi][i]=jk+1;
4621: bh[mi][i]=ju;
4622: }
4623: }else{
4624: if(jl <= -ju){
4625: dh[mi][i]=jk;
4626: bh[mi][i]=jl; /* bias is positive if real duration
4627: * is higher than the multiple of stepm and negative otherwise.
4628: */
4629: }
4630: else{
4631: dh[mi][i]=jk+1;
4632: bh[mi][i]=ju;
4633: }
4634: if(dh[mi][i]==0){
4635: dh[mi][i]=1; /* At least one step */
4636: bh[mi][i]=ju; /* At least one step */
4637: /* 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);*/
4638: }
4639: } /* end if mle */
1.126 brouard 4640: }
4641: } /* end wave */
4642: }
4643: jmean=sum/k;
4644: 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 4645: 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 4646: }
1.126 brouard 4647:
4648: /*********** Tricode ****************************/
1.220 brouard 4649: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.126 brouard 4650: {
1.144 brouard 4651: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
4652: /* 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 4653: * Boring subroutine which should only output nbcode[Tvar[j]][k]
1.224 brouard 4654: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
4655: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
1.144 brouard 4656: */
1.130 brouard 4657:
1.145 brouard 4658: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
1.136 brouard 4659: int modmaxcovj=0; /* Modality max of covariates j */
1.145 brouard 4660: int cptcode=0; /* Modality max of covariates j */
4661: int modmincovj=0; /* Modality min of covariates j */
4662:
4663:
1.220 brouard 4664: /* cptcoveff=0; */
1.224 brouard 4665: /* *cptcov=0; */
1.126 brouard 4666:
1.144 brouard 4667: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 4668:
1.224 brouard 4669: /* Loop on covariates without age and products and no quantitative variable */
4670: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
1.227 brouard 4671: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
4672: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
4673: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
4674: switch(Fixed[k]) {
4675: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
1.231 brouard 4676: 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*/
4677: ij=(int)(covar[Tvar[k]][i]);
4678: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
4679: * If product of Vn*Vm, still boolean *:
4680: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
4681: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
4682: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
4683: modality of the nth covariate of individual i. */
4684: if (ij > modmaxcovj)
4685: modmaxcovj=ij;
4686: else if (ij < modmincovj)
4687: modmincovj=ij;
4688: if ((ij < -1) && (ij > NCOVMAX)){
4689: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
4690: exit(1);
4691: }else
4692: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
4693: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
4694: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
4695: /* getting the maximum value of the modality of the covariate
4696: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
4697: female ies 1, then modmaxcovj=1.
4698: */
4699: } /* end for loop on individuals i */
4700: printf(" Minimal and maximal values of %d th covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4701: fprintf(ficlog," Minimal and maximal values of %d th covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4702: cptcode=modmaxcovj;
4703: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
4704: /*for (i=0; i<=cptcode; i++) {*/
4705: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
4706: printf("Frequencies of covariates %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4707: fprintf(ficlog, "Frequencies of covariates %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4708: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
4709: if( j != -1){
4710: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
4711: covariate for which somebody answered excluding
4712: undefined. Usually 2: 0 and 1. */
4713: }
4714: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
4715: covariate for which somebody answered including
4716: undefined. Usually 3: -1, 0 and 1. */
4717: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
4718: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
4719: } /* Ndum[-1] number of undefined modalities */
4720:
4721: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
4722: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
4723: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
4724: /* modmincovj=3; modmaxcovj = 7; */
4725: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
4726: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
4727: /* defining two dummy variables: variables V1_1 and V1_2.*/
4728: /* nbcode[Tvar[j]][ij]=k; */
4729: /* nbcode[Tvar[j]][1]=0; */
4730: /* nbcode[Tvar[j]][2]=1; */
4731: /* nbcode[Tvar[j]][3]=2; */
4732: /* To be continued (not working yet). */
4733: ij=0; /* ij is similar to i but can jump over null modalities */
4734: 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*/
4735: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
4736: break;
4737: }
4738: ij++;
4739: 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*/
4740: cptcode = ij; /* New max modality for covar j */
4741: } /* end of loop on modality i=-1 to 1 or more */
4742: break;
1.227 brouard 4743: case 1: /* Testing on varying covariate, could be simple and
4744: * should look at waves or product of fixed *
4745: * varying. No time to test -1, assuming 0 and 1 only */
1.231 brouard 4746: ij=0;
4747: for(i=0; i<=1;i++){
4748: nbcode[Tvar[k]][++ij]=i;
4749: }
4750: break;
1.227 brouard 4751: default:
1.231 brouard 4752: break;
1.227 brouard 4753: } /* end switch */
4754: } /* end dummy test */
1.225 brouard 4755:
1.192 brouard 4756: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
4757: /* /\*recode from 0 *\/ */
4758: /* k is a modality. If we have model=V1+V1*sex */
4759: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
4760: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
4761: /* } */
4762: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
4763: /* if (ij > ncodemax[j]) { */
4764: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4765: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4766: /* break; */
4767: /* } */
4768: /* } /\* end of loop on modality k *\/ */
1.137 brouard 4769: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
4770:
1.225 brouard 4771: for (k=-1; k< maxncov; k++) Ndum[k]=0;
1.227 brouard 4772: /* Look at fixed dummy (single or product) covariates to check empty modalities */
1.187 brouard 4773: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
1.225 brouard 4774: /* 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 4775: 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 */
4776: 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 */
4777: /* 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 4778: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
4779:
4780: ij=0;
1.227 brouard 4781: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
4782: 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 4783: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
1.227 brouard 4784: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
4785: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
4786: /* If product not in single variable we don't print results */
1.225 brouard 4787: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
1.230 brouard 4788: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
4789: 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*/
4790: 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 4791: 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 4792: if(Fixed[k]!=0)
4793: anyvaryingduminmodel=1;
1.231 brouard 4794: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
4795: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
4796: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
4797: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
4798: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
4799: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
1.227 brouard 4800: }
1.225 brouard 4801: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
4802: /* ij--; */
4803: /* cptcoveff=ij; /\*Number of total covariates*\/ */
4804: *cptcov=ij; /*Number of total real effective covariates: effective
1.231 brouard 4805: * because they can be excluded from the model and real
4806: * if in the model but excluded because missing values, but how to get k from ij?*/
1.227 brouard 4807: for(j=ij+1; j<= cptcovt; j++){
4808: Tvaraff[j]=0;
4809: Tmodelind[j]=0;
4810: }
1.228 brouard 4811: for(j=ntveff+1; j<= cptcovt; j++){
4812: TmodelInvind[j]=0;
4813: }
1.227 brouard 4814: /* To be sorted */
4815: ;
1.126 brouard 4816: }
4817:
1.145 brouard 4818:
1.126 brouard 4819: /*********** Health Expectancies ****************/
4820:
1.127 brouard 4821: void evsij(double ***eij, double x[], int nlstate, int stepm, int bage, int fage, double **oldm, double **savm, int cij, int estepm,char strstart[] )
1.126 brouard 4822:
4823: {
4824: /* Health expectancies, no variances */
1.164 brouard 4825: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 4826: int nhstepma, nstepma; /* Decreasing with age */
4827: double age, agelim, hf;
4828: double ***p3mat;
4829: double eip;
4830:
4831: pstamp(ficreseij);
4832: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
4833: fprintf(ficreseij,"# Age");
4834: for(i=1; i<=nlstate;i++){
4835: for(j=1; j<=nlstate;j++){
4836: fprintf(ficreseij," e%1d%1d ",i,j);
4837: }
4838: fprintf(ficreseij," e%1d. ",i);
4839: }
4840: fprintf(ficreseij,"\n");
4841:
4842:
4843: if(estepm < stepm){
4844: printf ("Problem %d lower than %d\n",estepm, stepm);
4845: }
4846: else hstepm=estepm;
4847: /* We compute the life expectancy from trapezoids spaced every estepm months
4848: * This is mainly to measure the difference between two models: for example
4849: * if stepm=24 months pijx are given only every 2 years and by summing them
4850: * we are calculating an estimate of the Life Expectancy assuming a linear
4851: * progression in between and thus overestimating or underestimating according
4852: * to the curvature of the survival function. If, for the same date, we
4853: * estimate the model with stepm=1 month, we can keep estepm to 24 months
4854: * to compare the new estimate of Life expectancy with the same linear
4855: * hypothesis. A more precise result, taking into account a more precise
4856: * curvature will be obtained if estepm is as small as stepm. */
4857:
4858: /* For example we decided to compute the life expectancy with the smallest unit */
4859: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
4860: nhstepm is the number of hstepm from age to agelim
4861: nstepm is the number of stepm from age to agelin.
4862: Look at hpijx to understand the reason of that which relies in memory size
4863: and note for a fixed period like estepm months */
4864: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
4865: survival function given by stepm (the optimization length). Unfortunately it
4866: means that if the survival funtion is printed only each two years of age and if
4867: you sum them up and add 1 year (area under the trapezoids) you won't get the same
4868: results. So we changed our mind and took the option of the best precision.
4869: */
4870: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
4871:
4872: agelim=AGESUP;
4873: /* If stepm=6 months */
4874: /* Computed by stepm unit matrices, product of hstepm matrices, stored
4875: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
4876:
4877: /* nhstepm age range expressed in number of stepm */
4878: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
4879: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
4880: /* if (stepm >= YEARM) hstepm=1;*/
4881: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
4882: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
4883:
4884: for (age=bage; age<=fage; age ++){
4885: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
4886: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
4887: /* if (stepm >= YEARM) hstepm=1;*/
4888: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
4889:
4890: /* If stepm=6 months */
4891: /* Computed by stepm unit matrices, product of hstepma matrices, stored
4892: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
4893:
4894: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij);
4895:
4896: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
4897:
4898: printf("%d|",(int)age);fflush(stdout);
4899: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
4900:
4901: /* Computing expectancies */
4902: for(i=1; i<=nlstate;i++)
4903: for(j=1; j<=nlstate;j++)
4904: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
4905: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
4906:
4907: /* 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]);*/
4908:
4909: }
4910:
4911: fprintf(ficreseij,"%3.0f",age );
4912: for(i=1; i<=nlstate;i++){
4913: eip=0;
4914: for(j=1; j<=nlstate;j++){
4915: eip +=eij[i][j][(int)age];
4916: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
4917: }
4918: fprintf(ficreseij,"%9.4f", eip );
4919: }
4920: fprintf(ficreseij,"\n");
4921:
4922: }
4923: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
4924: printf("\n");
4925: fprintf(ficlog,"\n");
4926:
4927: }
4928:
1.127 brouard 4929: 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[] )
1.126 brouard 4930:
4931: {
4932: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 4933: to initial status i, ei. .
1.126 brouard 4934: */
4935: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
4936: int nhstepma, nstepma; /* Decreasing with age */
4937: double age, agelim, hf;
4938: double ***p3matp, ***p3matm, ***varhe;
4939: double **dnewm,**doldm;
4940: double *xp, *xm;
4941: double **gp, **gm;
4942: double ***gradg, ***trgradg;
4943: int theta;
4944:
4945: double eip, vip;
4946:
4947: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
4948: xp=vector(1,npar);
4949: xm=vector(1,npar);
4950: dnewm=matrix(1,nlstate*nlstate,1,npar);
4951: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
4952:
4953: pstamp(ficresstdeij);
4954: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
4955: fprintf(ficresstdeij,"# Age");
4956: for(i=1; i<=nlstate;i++){
4957: for(j=1; j<=nlstate;j++)
4958: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
4959: fprintf(ficresstdeij," e%1d. ",i);
4960: }
4961: fprintf(ficresstdeij,"\n");
4962:
4963: pstamp(ficrescveij);
4964: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
4965: fprintf(ficrescveij,"# Age");
4966: for(i=1; i<=nlstate;i++)
4967: for(j=1; j<=nlstate;j++){
4968: cptj= (j-1)*nlstate+i;
4969: for(i2=1; i2<=nlstate;i2++)
4970: for(j2=1; j2<=nlstate;j2++){
4971: cptj2= (j2-1)*nlstate+i2;
4972: if(cptj2 <= cptj)
4973: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
4974: }
4975: }
4976: fprintf(ficrescveij,"\n");
4977:
4978: if(estepm < stepm){
4979: printf ("Problem %d lower than %d\n",estepm, stepm);
4980: }
4981: else hstepm=estepm;
4982: /* We compute the life expectancy from trapezoids spaced every estepm months
4983: * This is mainly to measure the difference between two models: for example
4984: * if stepm=24 months pijx are given only every 2 years and by summing them
4985: * we are calculating an estimate of the Life Expectancy assuming a linear
4986: * progression in between and thus overestimating or underestimating according
4987: * to the curvature of the survival function. If, for the same date, we
4988: * estimate the model with stepm=1 month, we can keep estepm to 24 months
4989: * to compare the new estimate of Life expectancy with the same linear
4990: * hypothesis. A more precise result, taking into account a more precise
4991: * curvature will be obtained if estepm is as small as stepm. */
4992:
4993: /* For example we decided to compute the life expectancy with the smallest unit */
4994: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
4995: nhstepm is the number of hstepm from age to agelim
4996: nstepm is the number of stepm from age to agelin.
4997: Look at hpijx to understand the reason of that which relies in memory size
4998: and note for a fixed period like estepm months */
4999: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5000: survival function given by stepm (the optimization length). Unfortunately it
5001: means that if the survival funtion is printed only each two years of age and if
5002: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5003: results. So we changed our mind and took the option of the best precision.
5004: */
5005: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5006:
5007: /* If stepm=6 months */
5008: /* nhstepm age range expressed in number of stepm */
5009: agelim=AGESUP;
5010: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5011: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5012: /* if (stepm >= YEARM) hstepm=1;*/
5013: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5014:
5015: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5016: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5017: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5018: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5019: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5020: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5021:
5022: for (age=bage; age<=fage; age ++){
5023: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5024: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5025: /* if (stepm >= YEARM) hstepm=1;*/
5026: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5027:
1.126 brouard 5028: /* If stepm=6 months */
5029: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5030: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5031:
5032: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5033:
1.126 brouard 5034: /* Computing Variances of health expectancies */
5035: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5036: decrease memory allocation */
5037: for(theta=1; theta <=npar; theta++){
5038: for(i=1; i<=npar; i++){
1.222 brouard 5039: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5040: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5041: }
5042: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij);
5043: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij);
1.218 brouard 5044:
1.126 brouard 5045: for(j=1; j<= nlstate; j++){
1.222 brouard 5046: for(i=1; i<=nlstate; i++){
5047: for(h=0; h<=nhstepm-1; h++){
5048: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5049: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5050: }
5051: }
1.126 brouard 5052: }
1.218 brouard 5053:
1.126 brouard 5054: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5055: for(h=0; h<=nhstepm-1; h++){
5056: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5057: }
1.126 brouard 5058: }/* End theta */
5059:
5060:
5061: for(h=0; h<=nhstepm-1; h++)
5062: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5063: for(theta=1; theta <=npar; theta++)
5064: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5065:
1.218 brouard 5066:
1.222 brouard 5067: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5068: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5069: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5070:
1.222 brouard 5071: printf("%d|",(int)age);fflush(stdout);
5072: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5073: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5074: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5075: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5076: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5077: for(ij=1;ij<=nlstate*nlstate;ij++)
5078: for(ji=1;ji<=nlstate*nlstate;ji++)
5079: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5080: }
5081: }
1.218 brouard 5082:
1.126 brouard 5083: /* Computing expectancies */
5084: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij);
5085: for(i=1; i<=nlstate;i++)
5086: for(j=1; j<=nlstate;j++)
1.222 brouard 5087: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5088: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5089:
1.222 brouard 5090: /* 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 5091:
1.222 brouard 5092: }
1.218 brouard 5093:
1.126 brouard 5094: fprintf(ficresstdeij,"%3.0f",age );
5095: for(i=1; i<=nlstate;i++){
5096: eip=0.;
5097: vip=0.;
5098: for(j=1; j<=nlstate;j++){
1.222 brouard 5099: eip += eij[i][j][(int)age];
5100: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5101: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5102: 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 5103: }
5104: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5105: }
5106: fprintf(ficresstdeij,"\n");
1.218 brouard 5107:
1.126 brouard 5108: fprintf(ficrescveij,"%3.0f",age );
5109: for(i=1; i<=nlstate;i++)
5110: for(j=1; j<=nlstate;j++){
1.222 brouard 5111: cptj= (j-1)*nlstate+i;
5112: for(i2=1; i2<=nlstate;i2++)
5113: for(j2=1; j2<=nlstate;j2++){
5114: cptj2= (j2-1)*nlstate+i2;
5115: if(cptj2 <= cptj)
5116: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5117: }
1.126 brouard 5118: }
5119: fprintf(ficrescveij,"\n");
1.218 brouard 5120:
1.126 brouard 5121: }
5122: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5123: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5124: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5125: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5126: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5127: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5128: printf("\n");
5129: fprintf(ficlog,"\n");
1.218 brouard 5130:
1.126 brouard 5131: free_vector(xm,1,npar);
5132: free_vector(xp,1,npar);
5133: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5134: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5135: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5136: }
1.218 brouard 5137:
1.126 brouard 5138: /************ Variance ******************/
1.209 brouard 5139: 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[])
1.218 brouard 5140: {
5141: /* Variance of health expectancies */
5142: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5143: /* double **newm;*/
5144: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5145:
5146: /* int movingaverage(); */
5147: double **dnewm,**doldm;
5148: double **dnewmp,**doldmp;
5149: int i, j, nhstepm, hstepm, h, nstepm ;
5150: int k;
5151: double *xp;
5152: double **gp, **gm; /* for var eij */
5153: double ***gradg, ***trgradg; /*for var eij */
5154: double **gradgp, **trgradgp; /* for var p point j */
5155: double *gpp, *gmp; /* for var p point j */
5156: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5157: double ***p3mat;
5158: double age,agelim, hf;
5159: /* double ***mobaverage; */
5160: int theta;
5161: char digit[4];
5162: char digitp[25];
5163:
5164: char fileresprobmorprev[FILENAMELENGTH];
5165:
5166: if(popbased==1){
5167: if(mobilav!=0)
5168: strcpy(digitp,"-POPULBASED-MOBILAV_");
5169: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5170: }
5171: else
5172: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5173:
1.218 brouard 5174: /* if (mobilav!=0) { */
5175: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5176: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5177: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5178: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5179: /* } */
5180: /* } */
5181:
5182: strcpy(fileresprobmorprev,"PRMORPREV-");
5183: sprintf(digit,"%-d",ij);
5184: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5185: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5186: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5187: strcat(fileresprobmorprev,fileresu);
5188: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5189: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5190: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5191: }
5192: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5193: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5194: pstamp(ficresprobmorprev);
5195: 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);
5196: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5197: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5198: fprintf(ficresprobmorprev," p.%-d SE",j);
5199: for(i=1; i<=nlstate;i++)
5200: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5201: }
5202: fprintf(ficresprobmorprev,"\n");
5203:
5204: fprintf(ficgp,"\n# Routine varevsij");
5205: fprintf(ficgp,"\nunset title \n");
5206: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5207: 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");
5208: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5209: /* } */
5210: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5211: pstamp(ficresvij);
5212: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5213: if(popbased==1)
5214: 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);
5215: else
5216: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5217: fprintf(ficresvij,"# Age");
5218: for(i=1; i<=nlstate;i++)
5219: for(j=1; j<=nlstate;j++)
5220: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5221: fprintf(ficresvij,"\n");
5222:
5223: xp=vector(1,npar);
5224: dnewm=matrix(1,nlstate,1,npar);
5225: doldm=matrix(1,nlstate,1,nlstate);
5226: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5227: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5228:
5229: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5230: gpp=vector(nlstate+1,nlstate+ndeath);
5231: gmp=vector(nlstate+1,nlstate+ndeath);
5232: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5233:
1.218 brouard 5234: if(estepm < stepm){
5235: printf ("Problem %d lower than %d\n",estepm, stepm);
5236: }
5237: else hstepm=estepm;
5238: /* For example we decided to compute the life expectancy with the smallest unit */
5239: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5240: nhstepm is the number of hstepm from age to agelim
5241: nstepm is the number of stepm from age to agelim.
5242: Look at function hpijx to understand why because of memory size limitations,
5243: we decided (b) to get a life expectancy respecting the most precise curvature of the
5244: survival function given by stepm (the optimization length). Unfortunately it
5245: means that if the survival funtion is printed every two years of age and if
5246: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5247: results. So we changed our mind and took the option of the best precision.
5248: */
5249: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5250: agelim = AGESUP;
5251: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5252: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5253: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5254: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5255: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5256: gp=matrix(0,nhstepm,1,nlstate);
5257: gm=matrix(0,nhstepm,1,nlstate);
5258:
5259:
5260: for(theta=1; theta <=npar; theta++){
5261: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5262: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5263: }
5264:
5265: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij);
5266:
5267: if (popbased==1) {
5268: if(mobilav ==0){
5269: for(i=1; i<=nlstate;i++)
5270: prlim[i][i]=probs[(int)age][i][ij];
5271: }else{ /* mobilav */
5272: for(i=1; i<=nlstate;i++)
5273: prlim[i][i]=mobaverage[(int)age][i][ij];
5274: }
5275: }
5276:
5277: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij); /* Returns p3mat[i][j][h] for h=1 to nhstepm */
5278: for(j=1; j<= nlstate; j++){
5279: for(h=0; h<=nhstepm; h++){
5280: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5281: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5282: }
5283: }
5284: /* Next for computing probability of death (h=1 means
5285: computed over hstepm matrices product = hstepm*stepm months)
5286: as a weighted average of prlim.
5287: */
5288: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5289: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5290: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5291: }
5292: /* end probability of death */
5293:
5294: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5295: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5296:
5297: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij);
5298:
5299: if (popbased==1) {
5300: if(mobilav ==0){
5301: for(i=1; i<=nlstate;i++)
5302: prlim[i][i]=probs[(int)age][i][ij];
5303: }else{ /* mobilav */
5304: for(i=1; i<=nlstate;i++)
5305: prlim[i][i]=mobaverage[(int)age][i][ij];
5306: }
5307: }
5308:
5309: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij);
5310:
5311: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5312: for(h=0; h<=nhstepm; h++){
5313: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5314: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5315: }
5316: }
5317: /* This for computing probability of death (h=1 means
5318: computed over hstepm matrices product = hstepm*stepm months)
5319: as a weighted average of prlim.
5320: */
5321: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5322: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5323: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5324: }
5325: /* end probability of death */
5326:
5327: for(j=1; j<= nlstate; j++) /* vareij */
5328: for(h=0; h<=nhstepm; h++){
5329: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5330: }
5331:
5332: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5333: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5334: }
5335:
5336: } /* End theta */
5337:
5338: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5339:
5340: for(h=0; h<=nhstepm; h++) /* veij */
5341: for(j=1; j<=nlstate;j++)
5342: for(theta=1; theta <=npar; theta++)
5343: trgradg[h][j][theta]=gradg[h][theta][j];
5344:
5345: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5346: for(theta=1; theta <=npar; theta++)
5347: trgradgp[j][theta]=gradgp[theta][j];
5348:
5349:
5350: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5351: for(i=1;i<=nlstate;i++)
5352: for(j=1;j<=nlstate;j++)
5353: vareij[i][j][(int)age] =0.;
5354:
5355: for(h=0;h<=nhstepm;h++){
5356: for(k=0;k<=nhstepm;k++){
5357: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5358: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5359: for(i=1;i<=nlstate;i++)
5360: for(j=1;j<=nlstate;j++)
5361: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5362: }
5363: }
5364:
5365: /* pptj */
5366: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5367: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5368: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5369: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5370: varppt[j][i]=doldmp[j][i];
5371: /* end ppptj */
5372: /* x centered again */
5373:
5374: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij);
5375:
5376: if (popbased==1) {
5377: if(mobilav ==0){
5378: for(i=1; i<=nlstate;i++)
5379: prlim[i][i]=probs[(int)age][i][ij];
5380: }else{ /* mobilav */
5381: for(i=1; i<=nlstate;i++)
5382: prlim[i][i]=mobaverage[(int)age][i][ij];
5383: }
5384: }
5385:
5386: /* This for computing probability of death (h=1 means
5387: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5388: as a weighted average of prlim.
5389: */
5390: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij);
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: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5398: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5399: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5400: for(i=1; i<=nlstate;i++){
5401: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5402: }
5403: }
5404: fprintf(ficresprobmorprev,"\n");
5405:
5406: fprintf(ficresvij,"%.0f ",age );
5407: for(i=1; i<=nlstate;i++)
5408: for(j=1; j<=nlstate;j++){
5409: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5410: }
5411: fprintf(ficresvij,"\n");
5412: free_matrix(gp,0,nhstepm,1,nlstate);
5413: free_matrix(gm,0,nhstepm,1,nlstate);
5414: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5415: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5416: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5417: } /* End age */
5418: free_vector(gpp,nlstate+1,nlstate+ndeath);
5419: free_vector(gmp,nlstate+1,nlstate+ndeath);
5420: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5421: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5422: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5423: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5424: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5425: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5426: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5427: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5428: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5429: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5430: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5431: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5432: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5433: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5434: 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);
5435: /* 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 5436: */
1.218 brouard 5437: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5438: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5439:
1.218 brouard 5440: free_vector(xp,1,npar);
5441: free_matrix(doldm,1,nlstate,1,nlstate);
5442: free_matrix(dnewm,1,nlstate,1,npar);
5443: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5444: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5445: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5446: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5447: fclose(ficresprobmorprev);
5448: fflush(ficgp);
5449: fflush(fichtm);
5450: } /* end varevsij */
1.126 brouard 5451:
5452: /************ Variance of prevlim ******************/
1.209 brouard 5453: 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[])
1.126 brouard 5454: {
1.205 brouard 5455: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5456: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5457:
1.126 brouard 5458: double **dnewm,**doldm;
5459: int i, j, nhstepm, hstepm;
5460: double *xp;
5461: double *gp, *gm;
5462: double **gradg, **trgradg;
1.208 brouard 5463: double **mgm, **mgp;
1.126 brouard 5464: double age,agelim;
5465: int theta;
5466:
5467: pstamp(ficresvpl);
5468: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
5469: fprintf(ficresvpl,"# Age");
5470: for(i=1; i<=nlstate;i++)
5471: fprintf(ficresvpl," %1d-%1d",i,i);
5472: fprintf(ficresvpl,"\n");
5473:
5474: xp=vector(1,npar);
5475: dnewm=matrix(1,nlstate,1,npar);
5476: doldm=matrix(1,nlstate,1,nlstate);
5477:
5478: hstepm=1*YEARM; /* Every year of age */
5479: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5480: agelim = AGESUP;
5481: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5482: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5483: if (stepm >= YEARM) hstepm=1;
5484: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5485: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5486: mgp=matrix(1,npar,1,nlstate);
5487: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5488: gp=vector(1,nlstate);
5489: gm=vector(1,nlstate);
5490:
5491: for(theta=1; theta <=npar; theta++){
5492: for(i=1; i<=npar; i++){ /* Computes gradient */
5493: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5494: }
1.209 brouard 5495: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
5496: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij);
5497: else
5498: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij);
1.208 brouard 5499: for(i=1;i<=nlstate;i++){
1.126 brouard 5500: gp[i] = prlim[i][i];
1.208 brouard 5501: mgp[theta][i] = prlim[i][i];
5502: }
1.126 brouard 5503: for(i=1; i<=npar; i++) /* Computes gradient */
5504: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5505: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
5506: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij);
5507: else
5508: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij);
1.208 brouard 5509: for(i=1;i<=nlstate;i++){
1.126 brouard 5510: gm[i] = prlim[i][i];
1.208 brouard 5511: mgm[theta][i] = prlim[i][i];
5512: }
1.126 brouard 5513: for(i=1;i<=nlstate;i++)
5514: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5515: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5516: } /* End theta */
5517:
5518: trgradg =matrix(1,nlstate,1,npar);
5519:
5520: for(j=1; j<=nlstate;j++)
5521: for(theta=1; theta <=npar; theta++)
5522: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 5523: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5524: /* printf("\nmgm mgp %d ",(int)age); */
5525: /* for(j=1; j<=nlstate;j++){ */
5526: /* printf(" %d ",j); */
5527: /* for(theta=1; theta <=npar; theta++) */
5528: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
5529: /* printf("\n "); */
5530: /* } */
5531: /* } */
5532: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5533: /* printf("\n gradg %d ",(int)age); */
5534: /* for(j=1; j<=nlstate;j++){ */
5535: /* printf("%d ",j); */
5536: /* for(theta=1; theta <=npar; theta++) */
5537: /* printf("%d %lf ",theta,gradg[theta][j]); */
5538: /* printf("\n "); */
5539: /* } */
5540: /* } */
1.126 brouard 5541:
5542: for(i=1;i<=nlstate;i++)
5543: varpl[i][(int)age] =0.;
1.209 brouard 5544: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 5545: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5546: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
5547: }else{
1.126 brouard 5548: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5549: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 5550: }
1.126 brouard 5551: for(i=1;i<=nlstate;i++)
5552: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
5553:
5554: fprintf(ficresvpl,"%.0f ",age );
5555: for(i=1; i<=nlstate;i++)
5556: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
5557: fprintf(ficresvpl,"\n");
5558: free_vector(gp,1,nlstate);
5559: free_vector(gm,1,nlstate);
1.208 brouard 5560: free_matrix(mgm,1,npar,1,nlstate);
5561: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 5562: free_matrix(gradg,1,npar,1,nlstate);
5563: free_matrix(trgradg,1,nlstate,1,npar);
5564: } /* End age */
5565:
5566: free_vector(xp,1,npar);
5567: free_matrix(doldm,1,nlstate,1,npar);
5568: free_matrix(dnewm,1,nlstate,1,nlstate);
5569:
5570: }
5571:
5572: /************ Variance of one-step probabilities ******************/
5573: 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 5574: {
5575: int i, j=0, k1, l1, tj;
5576: int k2, l2, j1, z1;
5577: int k=0, l;
5578: int first=1, first1, first2;
5579: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
5580: double **dnewm,**doldm;
5581: double *xp;
5582: double *gp, *gm;
5583: double **gradg, **trgradg;
5584: double **mu;
5585: double age, cov[NCOVMAX+1];
5586: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
5587: int theta;
5588: char fileresprob[FILENAMELENGTH];
5589: char fileresprobcov[FILENAMELENGTH];
5590: char fileresprobcor[FILENAMELENGTH];
5591: double ***varpij;
5592:
5593: strcpy(fileresprob,"PROB_");
5594: strcat(fileresprob,fileres);
5595: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
5596: printf("Problem with resultfile: %s\n", fileresprob);
5597: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
5598: }
5599: strcpy(fileresprobcov,"PROBCOV_");
5600: strcat(fileresprobcov,fileresu);
5601: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
5602: printf("Problem with resultfile: %s\n", fileresprobcov);
5603: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
5604: }
5605: strcpy(fileresprobcor,"PROBCOR_");
5606: strcat(fileresprobcor,fileresu);
5607: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
5608: printf("Problem with resultfile: %s\n", fileresprobcor);
5609: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
5610: }
5611: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5612: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5613: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5614: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5615: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5616: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5617: pstamp(ficresprob);
5618: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
5619: fprintf(ficresprob,"# Age");
5620: pstamp(ficresprobcov);
5621: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
5622: fprintf(ficresprobcov,"# Age");
5623: pstamp(ficresprobcor);
5624: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
5625: fprintf(ficresprobcor,"# Age");
1.126 brouard 5626:
5627:
1.222 brouard 5628: for(i=1; i<=nlstate;i++)
5629: for(j=1; j<=(nlstate+ndeath);j++){
5630: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
5631: fprintf(ficresprobcov," p%1d-%1d ",i,j);
5632: fprintf(ficresprobcor," p%1d-%1d ",i,j);
5633: }
5634: /* fprintf(ficresprob,"\n");
5635: fprintf(ficresprobcov,"\n");
5636: fprintf(ficresprobcor,"\n");
5637: */
5638: xp=vector(1,npar);
5639: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5640: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5641: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
5642: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
5643: first=1;
5644: fprintf(ficgp,"\n# Routine varprob");
5645: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
5646: fprintf(fichtm,"\n");
5647:
5648: 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);
5649: 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);
5650: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 5651: and drawn. It helps understanding how is the covariance between two incidences.\
5652: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 5653: 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 5654: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
5655: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
5656: standard deviations wide on each axis. <br>\
5657: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
5658: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
5659: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
5660:
1.222 brouard 5661: cov[1]=1;
5662: /* tj=cptcoveff; */
1.225 brouard 5663: tj = (int) pow(2,cptcoveff);
1.222 brouard 5664: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
5665: j1=0;
1.224 brouard 5666: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 5667: if (cptcovn>0) {
5668: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 5669: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5670: fprintf(ficresprob, "**********\n#\n");
5671: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 5672: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5673: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 5674:
1.222 brouard 5675: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 5676: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5677: fprintf(ficgp, "**********\n#\n");
1.220 brouard 5678:
5679:
1.222 brouard 5680: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 5681: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5682: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 5683:
1.222 brouard 5684: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 5685: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5686: fprintf(ficresprobcor, "**********\n#");
5687: if(invalidvarcomb[j1]){
5688: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
5689: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
5690: continue;
5691: }
5692: }
5693: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
5694: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5695: gp=vector(1,(nlstate)*(nlstate+ndeath));
5696: gm=vector(1,(nlstate)*(nlstate+ndeath));
5697: for (age=bage; age<=fage; age ++){
5698: cov[2]=age;
5699: if(nagesqr==1)
5700: cov[3]= age*age;
5701: for (k=1; k<=cptcovn;k++) {
5702: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
5703: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
5704: * 1 1 1 1 1
5705: * 2 2 1 1 1
5706: * 3 1 2 1 1
5707: */
5708: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
5709: }
5710: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
5711: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
5712: for (k=1; k<=cptcovprod;k++)
5713: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 5714:
5715:
1.222 brouard 5716: for(theta=1; theta <=npar; theta++){
5717: for(i=1; i<=npar; i++)
5718: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 5719:
1.222 brouard 5720: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 5721:
1.222 brouard 5722: k=0;
5723: for(i=1; i<= (nlstate); i++){
5724: for(j=1; j<=(nlstate+ndeath);j++){
5725: k=k+1;
5726: gp[k]=pmmij[i][j];
5727: }
5728: }
1.220 brouard 5729:
1.222 brouard 5730: for(i=1; i<=npar; i++)
5731: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 5732:
1.222 brouard 5733: pmij(pmmij,cov,ncovmodel,xp,nlstate);
5734: k=0;
5735: for(i=1; i<=(nlstate); i++){
5736: for(j=1; j<=(nlstate+ndeath);j++){
5737: k=k+1;
5738: gm[k]=pmmij[i][j];
5739: }
5740: }
1.220 brouard 5741:
1.222 brouard 5742: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
5743: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
5744: }
1.126 brouard 5745:
1.222 brouard 5746: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
5747: for(theta=1; theta <=npar; theta++)
5748: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 5749:
1.222 brouard 5750: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
5751: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 5752:
1.222 brouard 5753: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 5754:
1.222 brouard 5755: k=0;
5756: for(i=1; i<=(nlstate); i++){
5757: for(j=1; j<=(nlstate+ndeath);j++){
5758: k=k+1;
5759: mu[k][(int) age]=pmmij[i][j];
5760: }
5761: }
5762: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
5763: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
5764: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 5765:
1.222 brouard 5766: /*printf("\n%d ",(int)age);
5767: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
5768: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5769: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5770: }*/
1.220 brouard 5771:
1.222 brouard 5772: fprintf(ficresprob,"\n%d ",(int)age);
5773: fprintf(ficresprobcov,"\n%d ",(int)age);
5774: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 5775:
1.222 brouard 5776: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
5777: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
5778: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
5779: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
5780: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
5781: }
5782: i=0;
5783: for (k=1; k<=(nlstate);k++){
5784: for (l=1; l<=(nlstate+ndeath);l++){
5785: i++;
5786: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
5787: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
5788: for (j=1; j<=i;j++){
5789: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
5790: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
5791: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
5792: }
5793: }
5794: }/* end of loop for state */
5795: } /* end of loop for age */
5796: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
5797: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
5798: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
5799: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
5800:
5801: /* Confidence intervalle of pij */
5802: /*
5803: fprintf(ficgp,"\nunset parametric;unset label");
5804: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
5805: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
5806: 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);
5807: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
5808: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
5809: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
5810: */
5811:
5812: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
5813: first1=1;first2=2;
5814: for (k2=1; k2<=(nlstate);k2++){
5815: for (l2=1; l2<=(nlstate+ndeath);l2++){
5816: if(l2==k2) continue;
5817: j=(k2-1)*(nlstate+ndeath)+l2;
5818: for (k1=1; k1<=(nlstate);k1++){
5819: for (l1=1; l1<=(nlstate+ndeath);l1++){
5820: if(l1==k1) continue;
5821: i=(k1-1)*(nlstate+ndeath)+l1;
5822: if(i<=j) continue;
5823: for (age=bage; age<=fage; age ++){
5824: if ((int)age %5==0){
5825: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
5826: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
5827: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
5828: mu1=mu[i][(int) age]/stepm*YEARM ;
5829: mu2=mu[j][(int) age]/stepm*YEARM;
5830: c12=cv12/sqrt(v1*v2);
5831: /* Computing eigen value of matrix of covariance */
5832: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
5833: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
5834: if ((lc2 <0) || (lc1 <0) ){
5835: if(first2==1){
5836: first1=0;
5837: 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);
5838: }
5839: 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);
5840: /* lc1=fabs(lc1); */ /* If we want to have them positive */
5841: /* lc2=fabs(lc2); */
5842: }
1.220 brouard 5843:
1.222 brouard 5844: /* Eigen vectors */
5845: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
5846: /*v21=sqrt(1.-v11*v11); *//* error */
5847: v21=(lc1-v1)/cv12*v11;
5848: v12=-v21;
5849: v22=v11;
5850: tnalp=v21/v11;
5851: if(first1==1){
5852: first1=0;
5853: 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);
5854: }
5855: 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);
5856: /*printf(fignu*/
5857: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
5858: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
5859: if(first==1){
5860: first=0;
5861: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
5862: fprintf(ficgp,"\nset parametric;unset label");
5863: 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);
5864: fprintf(ficgp,"\nset ter svg size 640, 480");
5865: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 5866: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 5867: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 5868: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
5869: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5870: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5871: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
5872: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5873: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
5874: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
5875: 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", \
5876: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
5877: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
5878: }else{
5879: first=0;
5880: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
5881: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
5882: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
5883: 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", \
5884: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
5885: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
5886: }/* if first */
5887: } /* age mod 5 */
5888: } /* end loop age */
5889: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
5890: first=1;
5891: } /*l12 */
5892: } /* k12 */
5893: } /*l1 */
5894: }/* k1 */
5895: } /* loop on combination of covariates j1 */
5896: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
5897: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
5898: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5899: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
5900: free_vector(xp,1,npar);
5901: fclose(ficresprob);
5902: fclose(ficresprobcov);
5903: fclose(ficresprobcor);
5904: fflush(ficgp);
5905: fflush(fichtmcov);
5906: }
1.126 brouard 5907:
5908:
5909: /******************* Printing html file ***********/
1.201 brouard 5910: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 5911: int lastpass, int stepm, int weightopt, char model[],\
5912: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.217 brouard 5913: int popforecast, int prevfcast, int backcast, int estepm , \
1.213 brouard 5914: double jprev1, double mprev1,double anprev1, double dateprev1, \
5915: double jprev2, double mprev2,double anprev2, double dateprev2){
1.126 brouard 5916: int jj1, k1, i1, cpt;
5917:
5918: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
5919: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
5920: </ul>");
1.214 brouard 5921: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
5922: 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",
5923: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
5924: 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 5925: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
5926: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 5927: fprintf(fichtm,"\
5928: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 5929: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 5930: fprintf(fichtm,"\
1.217 brouard 5931: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
5932: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
5933: fprintf(fichtm,"\
1.126 brouard 5934: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 5935: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 5936: fprintf(fichtm,"\
1.217 brouard 5937: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
5938: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
5939: fprintf(fichtm,"\
1.211 brouard 5940: - (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 5941: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 5942: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 5943: if(prevfcast==1){
5944: fprintf(fichtm,"\
5945: - Prevalence projections by age and states: \
1.201 brouard 5946: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 5947: }
1.126 brouard 5948:
1.222 brouard 5949: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 5950:
1.225 brouard 5951: m=pow(2,cptcoveff);
1.222 brouard 5952: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 5953:
1.222 brouard 5954: jj1=0;
5955: for(k1=1; k1<=m;k1++){
1.220 brouard 5956:
1.222 brouard 5957: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
5958: jj1++;
5959: if (cptcovn > 0) {
5960: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 5961: for (cpt=1; cpt<=cptcoveff;cpt++){
1.222 brouard 5962: fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);
5963: printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout);
5964: }
1.230 brouard 5965: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 5966: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
5967: if(invalidvarcomb[k1]){
5968: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
5969: printf("\nCombination (%d) ignored because no cases \n",k1);
5970: continue;
5971: }
5972: }
5973: /* aij, bij */
5974: 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 5975: <img src=\"%s_%d-1.svg\">",model,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 5976: /* Pij */
5977: 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 5978: <img src=\"%s_%d-2.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 5979: /* Quasi-incidences */
5980: 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 5981: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 5982: incidence (rates) are the limit when h tends to zero of the ratio of the probability <sub>h</sub>P<sub>ij</sub> \
5983: 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 5984: <img src=\"%s_%d-3.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1,subdirf2(optionfilefiname,"PE_"),jj1);
1.222 brouard 5985: /* Survival functions (period) in state j */
5986: for(cpt=1; cpt<=nlstate;cpt++){
5987: 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 5988: <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 5989: }
5990: /* State specific survival functions (period) */
5991: for(cpt=1; cpt<=nlstate;cpt++){
5992: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 5993: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.201 brouard 5994: <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 5995: }
5996: /* Period (stable) prevalence in each health state */
5997: for(cpt=1; cpt<=nlstate;cpt++){
5998: 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 5999: <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 6000: }
6001: if(backcast==1){
6002: /* Period (stable) back prevalence in each health state */
6003: for(cpt=1; cpt<=nlstate;cpt++){
6004: 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 6005: <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 6006: }
1.217 brouard 6007: }
1.222 brouard 6008: if(prevfcast==1){
6009: /* Projection of prevalence up to period (stable) prevalence in each health state */
6010: for(cpt=1; cpt<=nlstate;cpt++){
6011: 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 6012: <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 6013: }
6014: }
1.220 brouard 6015:
1.222 brouard 6016: for(cpt=1; cpt<=nlstate;cpt++) {
6017: 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 6018: <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 6019: }
6020: /* } /\* end i1 *\/ */
6021: }/* End k1 */
6022: fprintf(fichtm,"</ul>");
1.126 brouard 6023:
1.222 brouard 6024: fprintf(fichtm,"\
1.126 brouard 6025: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6026: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6027: - 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 6028: But because parameters are usually highly correlated (a higher incidence of disability \
6029: and a higher incidence of recovery can give very close observed transition) it might \
6030: be very useful to look not only at linear confidence intervals estimated from the \
6031: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6032: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6033: covariance matrix of the one-step probabilities. \
6034: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6035:
1.222 brouard 6036: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6037: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6038: fprintf(fichtm,"\
1.126 brouard 6039: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6040: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6041:
1.222 brouard 6042: fprintf(fichtm,"\
1.126 brouard 6043: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6044: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6045: fprintf(fichtm,"\
1.126 brouard 6046: - 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): \
6047: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6048: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6049: fprintf(fichtm,"\
1.126 brouard 6050: - (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): \
6051: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6052: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6053: fprintf(fichtm,"\
1.128 brouard 6054: - 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 6055: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6056: fprintf(fichtm,"\
1.128 brouard 6057: - 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 6058: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6059: fprintf(fichtm,"\
1.126 brouard 6060: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6061: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6062:
6063: /* if(popforecast==1) fprintf(fichtm,"\n */
6064: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6065: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6066: /* <br>",fileres,fileres,fileres,fileres); */
6067: /* else */
6068: /* 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 6069: fflush(fichtm);
6070: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6071:
1.225 brouard 6072: m=pow(2,cptcoveff);
1.222 brouard 6073: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6074:
1.222 brouard 6075: jj1=0;
6076: for(k1=1; k1<=m;k1++){
6077: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6078: jj1++;
1.126 brouard 6079: if (cptcovn > 0) {
6080: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6081: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.222 brouard 6082: fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);
1.126 brouard 6083: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6084:
1.222 brouard 6085: if(invalidvarcomb[k1]){
6086: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6087: continue;
6088: }
1.126 brouard 6089: }
6090: for(cpt=1; cpt<=nlstate;cpt++) {
1.218 brouard 6091: fprintf(fichtm,"\n<br>- Observed (cross-sectional) and period (incidence based) \
6092: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d.svg\"> %s_%d-%d.svg</a>\n <br>\
1.205 brouard 6093: <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 6094: }
6095: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6096: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6097: true period expectancies (those weighted with period prevalences are also\
6098: drawn in addition to the population based expectancies computed using\
1.218 brouard 6099: observed and cahotic prevalences: <a href=\"%s_%d.svg\">%s_%d.svg</a>\n<br>\
1.205 brouard 6100: <img src=\"%s_%d.svg\">",subdirf2(optionfilefiname,"E_"),jj1,subdirf2(optionfilefiname,"E_"),jj1,subdirf2(optionfilefiname,"E_"),jj1);
1.222 brouard 6101: /* } /\* end i1 *\/ */
6102: }/* End k1 */
6103: fprintf(fichtm,"</ul>");
6104: fflush(fichtm);
1.126 brouard 6105: }
6106:
6107: /******************* Gnuplot file **************/
1.223 brouard 6108: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6109:
6110: char dirfileres[132],optfileres[132];
1.223 brouard 6111: char gplotcondition[132];
1.164 brouard 6112: int cpt=0,k1=0,i=0,k=0,j=0,jk=0,k2=0,k3=0,ij=0,l=0;
1.211 brouard 6113: int lv=0, vlv=0, kl=0;
1.130 brouard 6114: int ng=0;
1.201 brouard 6115: int vpopbased;
1.223 brouard 6116: int ioffset; /* variable offset for columns */
1.219 brouard 6117:
1.126 brouard 6118: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6119: /* printf("Problem with file %s",optionfilegnuplot); */
6120: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6121: /* } */
6122:
6123: /*#ifdef windows */
6124: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6125: /*#endif */
1.225 brouard 6126: m=pow(2,cptcoveff);
1.126 brouard 6127:
1.202 brouard 6128: /* Contribution to likelihood */
6129: /* Plot the probability implied in the likelihood */
1.223 brouard 6130: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6131: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6132: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6133: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6134: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6135: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6136: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6137: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6138: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6139: 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));
6140: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6141: 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));
6142: for (i=1; i<= nlstate ; i ++) {
6143: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6144: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6145: 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);
6146: for (j=2; j<= nlstate+ndeath ; j ++) {
6147: 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);
6148: }
6149: fprintf(ficgp,";\nset out; unset ylabel;\n");
6150: }
6151: /* 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 */
6152: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6153: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6154: fprintf(ficgp,"\nset out;unset log\n");
6155: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6156:
1.126 brouard 6157: strcpy(dirfileres,optionfilefiname);
6158: strcpy(optfileres,"vpl");
1.223 brouard 6159: /* 1eme*/
1.211 brouard 6160: for (cpt=1; cpt<= nlstate ; cpt ++) { /* For each live state */
1.230 brouard 6161: for (k1=1; k1<= m && selected(k1) ; k1 ++) { /* For each valid combination of covariate */
1.211 brouard 6162: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
6163: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files ");
1.225 brouard 6164: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6165: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
1.223 brouard 6166: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6167: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6168: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6169: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6170: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
6171: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6172: }
6173: fprintf(ficgp,"\n#\n");
1.223 brouard 6174: if(invalidvarcomb[k1]){
6175: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6176: continue;
6177: }
1.211 brouard 6178:
1.223 brouard 6179: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1);
6180: fprintf(ficgp,"\n#set out \"V_%s_%d-%d.svg\" \n",optionfilefiname,cpt,k1);
6181: fprintf(ficgp,"set xlabel \"Age\" \n\
1.219 brouard 6182: set ylabel \"Probability\" \n \
6183: set ter svg size 640, 480\n \
1.201 brouard 6184: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:2 \"%%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),k1-1,k1-1);
1.219 brouard 6185:
1.223 brouard 6186: for (i=1; i<= nlstate ; i ++) {
6187: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6188: else fprintf(ficgp," %%*lf (%%*lf)");
6189: }
6190: 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);
6191: for (i=1; i<= nlstate ; i ++) {
6192: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6193: else fprintf(ficgp," %%*lf (%%*lf)");
6194: }
6195: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2-1.96*$3) \"%%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1);
6196: for (i=1; i<= nlstate ; i ++) {
6197: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6198: else fprintf(ficgp," %%*lf (%%*lf)");
6199: }
6200: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6201: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6202: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
6203: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1 */
1.225 brouard 6204: if(cptcoveff ==0){
1.223 brouard 6205: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line ", 2+(cpt-1), cpt );
6206: }else{
6207: kl=0;
1.225 brouard 6208: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6209: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
1.223 brouard 6210: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6211: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6212: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6213: vlv= nbcode[Tvaraff[k]][lv];
6214: kl++;
6215: /* 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 *\/ */
6216: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6217: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6218: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
1.225 brouard 6219: if(k==cptcoveff){
1.227 brouard 6220: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
6221: 4+(cpt-1), cpt ); /* 4 or 6 ?*/
1.223 brouard 6222: }else{
6223: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6224: kl++;
6225: }
6226: } /* end covariate */
6227: } /* end if no covariate */
6228: } /* end if backcast */
6229: fprintf(ficgp,"\nset out \n");
1.201 brouard 6230: } /* k1 */
6231: } /* cpt */
1.126 brouard 6232: /*2 eme*/
6233: for (k1=1; k1<= m ; k1 ++) {
1.220 brouard 6234:
1.223 brouard 6235: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
1.225 brouard 6236: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6237: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6238: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6239: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6240: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6241: vlv= nbcode[Tvaraff[k]][lv];
6242: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6243: }
6244: fprintf(ficgp,"\n#\n");
6245: if(invalidvarcomb[k1]){
6246: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6247: continue;
6248: }
1.219 brouard 6249:
1.223 brouard 6250: fprintf(ficgp,"\nset out \"%s_%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1);
6251: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6252: if(vpopbased==0)
6253: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6254: else
6255: fprintf(ficgp,"\nreplot ");
6256: for (i=1; i<= nlstate+1 ; i ++) {
6257: k=2*i;
6258: 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);
6259: for (j=1; j<= nlstate+1 ; j ++) {
6260: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6261: else fprintf(ficgp," %%*lf (%%*lf)");
6262: }
6263: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6264: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
6265: 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);
6266: for (j=1; j<= nlstate+1 ; j ++) {
6267: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6268: else fprintf(ficgp," %%*lf (%%*lf)");
6269: }
6270: fprintf(ficgp,"\" t\"\" w l lt 0,");
6271: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4+$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1,vpopbased);
6272: for (j=1; j<= nlstate+1 ; j ++) {
6273: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6274: else fprintf(ficgp," %%*lf (%%*lf)");
6275: }
6276: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6277: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6278: } /* state */
6279: } /* vpopbased */
6280: fprintf(ficgp,"\nset out;set out \"%s_%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1); /* Buggy gnuplot */
1.201 brouard 6281: } /* k1 */
1.219 brouard 6282:
6283:
1.126 brouard 6284: /*3eme*/
6285: for (k1=1; k1<= m ; k1 ++) {
1.220 brouard 6286:
1.126 brouard 6287: for (cpt=1; cpt<= nlstate ; cpt ++) {
1.211 brouard 6288: fprintf(ficgp,"\n# 3d: Life expectancy with EXP_ files: cov=%d state=%d",k1, cpt);
1.225 brouard 6289: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6290: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6291: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6292: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6293: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6294: vlv= nbcode[Tvaraff[k]][lv];
6295: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6296: }
6297: fprintf(ficgp,"\n#\n");
1.223 brouard 6298: if(invalidvarcomb[k1]){
6299: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6300: continue;
6301: }
1.219 brouard 6302:
1.126 brouard 6303: /* k=2+nlstate*(2*cpt-2); */
6304: k=2+(nlstate+1)*(cpt-1);
1.201 brouard 6305: fprintf(ficgp,"\nset out \"%s_%d%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1);
1.199 brouard 6306: fprintf(ficgp,"set ter svg size 640, 480\n\
1.201 brouard 6307: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:%d t \"e%d1\" w l",ageminpar,fage,subdirf2(fileresu,"E_"),k1-1,k1-1,k,cpt);
1.126 brouard 6308: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
1.223 brouard 6309: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6310: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6311: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6312: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6313: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6314:
1.126 brouard 6315: */
6316: for (i=1; i< nlstate ; i ++) {
1.223 brouard 6317: 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);
6318: /* 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 6319:
1.126 brouard 6320: }
1.201 brouard 6321: 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 6322: }
6323: }
6324:
1.223 brouard 6325: /* 4eme */
1.201 brouard 6326: /* Survival functions (period) from state i in state j by initial state i */
6327: for (k1=1; k1<= m ; k1 ++) { /* For each multivariate if any */
1.220 brouard 6328:
1.201 brouard 6329: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.211 brouard 6330: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
1.225 brouard 6331: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6332: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6333: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6334: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6335: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6336: vlv= nbcode[Tvaraff[k]][lv];
6337: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6338: }
6339: fprintf(ficgp,"\n#\n");
1.223 brouard 6340: if(invalidvarcomb[k1]){
6341: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6342: continue;
6343: }
1.220 brouard 6344:
1.201 brouard 6345: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1);
6346: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
1.220 brouard 6347: set ter svg size 640, 480\n \
6348: unset log y\n \
1.201 brouard 6349: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6350: k=3;
1.201 brouard 6351: for (i=1; i<= nlstate ; i ++){
1.223 brouard 6352: if(i==1){
6353: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6354: }else{
6355: fprintf(ficgp,", '' ");
6356: }
6357: l=(nlstate+ndeath)*(i-1)+1;
6358: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6359: for (j=2; j<= nlstate+ndeath ; j ++)
6360: fprintf(ficgp,"+$%d",k+l+j-1);
6361: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
1.201 brouard 6362: } /* nlstate */
6363: fprintf(ficgp,"\nset out\n");
6364: } /* end cpt state*/
6365: } /* end covariate */
1.220 brouard 6366:
6367: /* 5eme */
1.201 brouard 6368: /* Survival functions (period) from state i in state j by final state j */
1.202 brouard 6369: for (k1=1; k1<= m ; k1 ++) { /* For each covariate if any */
1.201 brouard 6370: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
1.223 brouard 6371:
1.201 brouard 6372: 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 6373: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6374: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6375: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6376: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6377: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6378: vlv= nbcode[Tvaraff[k]][lv];
6379: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6380: }
6381: fprintf(ficgp,"\n#\n");
1.223 brouard 6382: if(invalidvarcomb[k1]){
1.227 brouard 6383: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6384: continue;
1.223 brouard 6385: }
1.227 brouard 6386:
1.201 brouard 6387: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1);
6388: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
1.227 brouard 6389: set ter svg size 640, 480\n \
6390: unset log y\n \
1.201 brouard 6391: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6392: k=3;
1.201 brouard 6393: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
1.227 brouard 6394: if(j==1)
6395: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6396: else
6397: fprintf(ficgp,", '' ");
6398: l=(nlstate+ndeath)*(cpt-1) +j;
6399: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6400: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6401: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6402: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
1.201 brouard 6403: } /* nlstate */
6404: fprintf(ficgp,", '' ");
6405: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6406: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
1.227 brouard 6407: l=(nlstate+ndeath)*(cpt-1) +j;
6408: if(j < nlstate)
6409: fprintf(ficgp,"$%d +",k+l);
6410: else
6411: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
1.201 brouard 6412: }
6413: fprintf(ficgp,"\nset out\n");
6414: } /* end cpt state*/
6415: } /* end covariate */
1.227 brouard 6416:
1.220 brouard 6417: /* 6eme */
1.202 brouard 6418: /* CV preval stable (period) for each covariate */
1.211 brouard 6419: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
1.153 brouard 6420: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6421:
1.211 brouard 6422: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6423: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6424: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6425: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6426: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6427: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6428: vlv= nbcode[Tvaraff[k]][lv];
6429: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6430: }
6431: fprintf(ficgp,"\n#\n");
1.223 brouard 6432: if(invalidvarcomb[k1]){
1.227 brouard 6433: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6434: continue;
1.223 brouard 6435: }
1.227 brouard 6436:
1.201 brouard 6437: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1);
1.126 brouard 6438: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.227 brouard 6439: set ter svg size 640, 480\n \
6440: unset log y\n \
1.153 brouard 6441: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6442: k=3; /* Offset */
1.153 brouard 6443: for (i=1; i<= nlstate ; i ++){
1.227 brouard 6444: if(i==1)
6445: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6446: else
6447: fprintf(ficgp,", '' ");
6448: l=(nlstate+ndeath)*(i-1)+1;
6449: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6450: for (j=2; j<= nlstate ; j ++)
6451: fprintf(ficgp,"+$%d",k+l+j-1);
6452: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6453: } /* nlstate */
1.201 brouard 6454: fprintf(ficgp,"\nset out\n");
1.153 brouard 6455: } /* end cpt state*/
6456: } /* end covariate */
1.227 brouard 6457:
6458:
1.220 brouard 6459: /* 7eme */
1.218 brouard 6460: if(backcast == 1){
1.217 brouard 6461: /* CV back preval stable (period) for each covariate */
1.218 brouard 6462: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
6463: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6464: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
6465: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6466: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6467: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6468: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 6469: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 6470: vlv= nbcode[Tvaraff[k]][lv];
6471: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6472: }
6473: fprintf(ficgp,"\n#\n");
6474: if(invalidvarcomb[k1]){
6475: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6476: continue;
6477: }
6478:
6479: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1);
6480: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
6481: set ter svg size 640, 480\n \
6482: unset log y\n \
1.218 brouard 6483: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6484: k=3; /* Offset */
6485: for (i=1; i<= nlstate ; i ++){
6486: if(i==1)
6487: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
6488: else
6489: fprintf(ficgp,", '' ");
6490: /* l=(nlstate+ndeath)*(i-1)+1; */
6491: l=(nlstate+ndeath)*(cpt-1)+1;
6492: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
6493: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
6494: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+(cpt-1)+i-1); /* a vérifier */
6495: /* for (j=2; j<= nlstate ; j ++) */
6496: /* fprintf(ficgp,"+$%d",k+l+j-1); */
6497: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
6498: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
6499: } /* nlstate */
6500: fprintf(ficgp,"\nset out\n");
1.218 brouard 6501: } /* end cpt state*/
6502: } /* end covariate */
6503: } /* End if backcast */
6504:
1.223 brouard 6505: /* 8eme */
1.218 brouard 6506: if(prevfcast==1){
6507: /* Projection from cross-sectional to stable (period) for each covariate */
6508:
6509: for (k1=1; k1<= m ; k1 ++) { /* For each covariate combination (1 to m=2**k), if any covariate is present */
1.211 brouard 6510: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6511: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
6512: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
6513: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6514: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6515: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6516: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6517: vlv= nbcode[Tvaraff[k]][lv];
6518: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6519: }
6520: fprintf(ficgp,"\n#\n");
6521: if(invalidvarcomb[k1]){
6522: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6523: continue;
6524: }
6525:
6526: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
6527: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1);
6528: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
6529: set ter svg size 640, 480\n \
6530: unset log y\n \
1.219 brouard 6531: plot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6532: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
6533: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6534: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6535: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6536: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6537: if(i==1){
6538: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
6539: }else{
6540: fprintf(ficgp,",\\\n '' ");
6541: }
6542: if(cptcoveff ==0){ /* No covariate */
6543: ioffset=2; /* Age is in 2 */
6544: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6545: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6546: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6547: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6548: fprintf(ficgp," u %d:(", ioffset);
6549: if(i==nlstate+1)
6550: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
6551: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6552: else
6553: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
6554: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6555: }else{ /* more than 2 covariates */
6556: if(cptcoveff ==1){
6557: ioffset=4; /* Age is in 4 */
6558: }else{
6559: ioffset=6; /* Age is in 6 */
6560: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6561: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6562: }
6563: fprintf(ficgp," u %d:(",ioffset);
6564: kl=0;
6565: strcpy(gplotcondition,"(");
6566: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
6567: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
6568: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6569: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6570: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6571: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
6572: kl++;
6573: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
6574: kl++;
6575: if(k <cptcoveff && cptcoveff>1)
6576: sprintf(gplotcondition+strlen(gplotcondition)," && ");
6577: }
6578: strcpy(gplotcondition+strlen(gplotcondition),")");
6579: /* 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 *\/ */
6580: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6581: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6582: /* '' 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*/
6583: if(i==nlstate+1){
6584: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
6585: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6586: }else{
6587: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
6588: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6589: }
6590: } /* end if covariate */
6591: } /* nlstate */
6592: fprintf(ficgp,"\nset out\n");
1.223 brouard 6593: } /* end cpt state*/
6594: } /* end covariate */
6595: } /* End if prevfcast */
1.227 brouard 6596:
6597:
1.223 brouard 6598: /* proba elementaires */
6599: fprintf(ficgp,"\n##############\n#MLE estimated parameters\n#############\n");
1.126 brouard 6600: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 6601: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 6602: for(k=1; k <=(nlstate+ndeath); k++){
6603: if (k != i) {
1.227 brouard 6604: fprintf(ficgp,"# current state %d\n",k);
6605: for(j=1; j <=ncovmodel; j++){
6606: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
6607: jk++;
6608: }
6609: fprintf(ficgp,"\n");
1.126 brouard 6610: }
6611: }
1.223 brouard 6612: }
1.187 brouard 6613: fprintf(ficgp,"##############\n#\n");
1.227 brouard 6614:
1.145 brouard 6615: /*goto avoid;*/
1.200 brouard 6616: fprintf(ficgp,"\n##############\n#Graphics of probabilities or incidences\n#############\n");
1.187 brouard 6617: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
6618: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
6619: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
6620: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
6621: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6622: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6623: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6624: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6625: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
6626: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6627: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
6628: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
6629: fprintf(ficgp,"#\n");
1.223 brouard 6630: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
6631: fprintf(ficgp,"# ng=%d\n",ng);
1.225 brouard 6632: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);
1.223 brouard 6633: for(jk=1; jk <=m; jk++) {
6634: fprintf(ficgp,"# jk=%d\n",jk);
6635: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng);
6636: fprintf(ficgp,"\nset ter svg size 640, 480 ");
6637: if (ng==1){
6638: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
6639: fprintf(ficgp,"\nunset log y");
6640: }else if (ng==2){
6641: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
6642: fprintf(ficgp,"\nset log y");
6643: }else if (ng==3){
6644: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
6645: fprintf(ficgp,"\nset log y");
6646: }else
6647: fprintf(ficgp,"\nunset title ");
6648: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
6649: i=1;
6650: for(k2=1; k2<=nlstate; k2++) {
6651: k3=i;
6652: for(k=1; k<=(nlstate+ndeath); k++) {
6653: if (k != k2){
6654: switch( ng) {
6655: case 1:
6656: if(nagesqr==0)
6657: fprintf(ficgp," p%d+p%d*x",i,i+1);
6658: else /* nagesqr =1 */
6659: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6660: break;
6661: case 2: /* ng=2 */
6662: if(nagesqr==0)
6663: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
6664: else /* nagesqr =1 */
6665: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6666: break;
6667: case 3:
6668: if(nagesqr==0)
6669: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
6670: else /* nagesqr =1 */
6671: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
6672: break;
6673: }
6674: ij=1;/* To be checked else nbcode[0][0] wrong */
6675: for(j=3; j <=ncovmodel-nagesqr; j++) {
6676: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
6677: if(ij <=cptcovage) { /* Bug valgrind */
6678: if((j-2)==Tage[ij]) { /* Bug valgrind */
6679: fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
6680: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
6681: ij++;
6682: }
6683: }
6684: else
1.227 brouard 6685: fprintf(ficgp,"+p%d*%d",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,j-2)]); /* Valgrind bug nbcode */
1.223 brouard 6686: }
6687: }else{
6688: i=i-ncovmodel;
6689: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
6690: fprintf(ficgp," (1.");
6691: }
1.227 brouard 6692:
1.223 brouard 6693: if(ng != 1){
6694: fprintf(ficgp,")/(1");
1.227 brouard 6695:
1.223 brouard 6696: for(k1=1; k1 <=nlstate; k1++){
6697: if(nagesqr==0)
6698: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
6699: else /* nagesqr =1 */
6700: 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 6701:
1.223 brouard 6702: ij=1;
6703: for(j=3; j <=ncovmodel-nagesqr; j++){
6704: if(ij <=cptcovage) { /* Bug valgrind */
6705: if((j-2)==Tage[ij]) { /* Bug valgrind */
6706: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
6707: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
6708: ij++;
6709: }
6710: }
6711: else
1.225 brouard 6712: 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 6713: }
6714: fprintf(ficgp,")");
6715: }
6716: fprintf(ficgp,")");
6717: if(ng ==2)
6718: fprintf(ficgp," t \"p%d%d\" ", k2,k);
6719: else /* ng= 3 */
6720: fprintf(ficgp," t \"i%d%d\" ", k2,k);
6721: }else{ /* end ng <> 1 */
6722: if( k !=k2) /* logit p11 is hard to draw */
6723: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
6724: }
6725: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
6726: fprintf(ficgp,",");
6727: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
6728: fprintf(ficgp,",");
6729: i=i+ncovmodel;
6730: } /* end k */
6731: } /* end k2 */
6732: fprintf(ficgp,"\n set out\n");
6733: } /* end jk */
6734: } /* end ng */
6735: /* avoid: */
6736: fflush(ficgp);
1.126 brouard 6737: } /* end gnuplot */
6738:
6739:
6740: /*************** Moving average **************/
1.219 brouard 6741: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 6742: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 6743:
1.222 brouard 6744: int i, cpt, cptcod;
6745: int modcovmax =1;
6746: int mobilavrange, mob;
6747: int iage=0;
6748:
6749: double sum=0.;
6750: double age;
6751: double *sumnewp, *sumnewm;
6752: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
6753:
6754:
1.225 brouard 6755: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 6756: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
6757:
6758: sumnewp = vector(1,ncovcombmax);
6759: sumnewm = vector(1,ncovcombmax);
6760: agemingood = vector(1,ncovcombmax);
6761: agemaxgood = vector(1,ncovcombmax);
6762:
6763: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6764: sumnewm[cptcod]=0.;
6765: sumnewp[cptcod]=0.;
6766: agemingood[cptcod]=0;
6767: agemaxgood[cptcod]=0;
6768: }
6769: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
6770:
6771: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
6772: if(mobilav==1) mobilavrange=5; /* default */
6773: else mobilavrange=mobilav;
6774: for (age=bage; age<=fage; age++)
6775: for (i=1; i<=nlstate;i++)
6776: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
6777: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
6778: /* We keep the original values on the extreme ages bage, fage and for
6779: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
6780: we use a 5 terms etc. until the borders are no more concerned.
6781: */
6782: for (mob=3;mob <=mobilavrange;mob=mob+2){
6783: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
6784: for (i=1; i<=nlstate;i++){
6785: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6786: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
6787: for (cpt=1;cpt<=(mob-1)/2;cpt++){
6788: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
6789: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
6790: }
6791: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
6792: }
6793: }
6794: }/* end age */
6795: }/* end mob */
6796: }else
6797: return -1;
6798: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
6799: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
6800: if(invalidvarcomb[cptcod]){
6801: printf("\nCombination (%d) ignored because no cases \n",cptcod);
6802: continue;
6803: }
1.219 brouard 6804:
1.222 brouard 6805: agemingood[cptcod]=fage-(mob-1)/2;
6806: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
6807: sumnewm[cptcod]=0.;
6808: for (i=1; i<=nlstate;i++){
6809: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6810: }
6811: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
6812: agemingood[cptcod]=age;
6813: }else{ /* bad */
6814: for (i=1; i<=nlstate;i++){
6815: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6816: } /* i */
6817: } /* end bad */
6818: }/* age */
6819: sum=0.;
6820: for (i=1; i<=nlstate;i++){
6821: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6822: }
6823: if(fabs(sum - 1.) > 1.e-3) { /* bad */
6824: 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);
6825: /* for (i=1; i<=nlstate;i++){ */
6826: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
6827: /* } /\* i *\/ */
6828: } /* end bad */
6829: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
6830: /* From youngest, finding the oldest wrong */
6831: agemaxgood[cptcod]=bage+(mob-1)/2;
6832: for (age=bage+(mob-1)/2; age<=fage; age++){
6833: sumnewm[cptcod]=0.;
6834: for (i=1; i<=nlstate;i++){
6835: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6836: }
6837: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
6838: agemaxgood[cptcod]=age;
6839: }else{ /* bad */
6840: for (i=1; i<=nlstate;i++){
6841: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6842: } /* i */
6843: } /* end bad */
6844: }/* age */
6845: sum=0.;
6846: for (i=1; i<=nlstate;i++){
6847: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6848: }
6849: if(fabs(sum - 1.) > 1.e-3) { /* bad */
6850: 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);
6851: /* for (i=1; i<=nlstate;i++){ */
6852: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
6853: /* } /\* i *\/ */
6854: } /* end bad */
6855:
6856: for (age=bage; age<=fage; age++){
6857: printf("%d %d ", cptcod, (int)age);
6858: sumnewp[cptcod]=0.;
6859: sumnewm[cptcod]=0.;
6860: for (i=1; i<=nlstate;i++){
6861: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
6862: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
6863: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
6864: }
6865: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
6866: }
6867: /* printf("\n"); */
6868: /* } */
6869: /* brutal averaging */
6870: for (i=1; i<=nlstate;i++){
6871: for (age=1; age<=bage; age++){
6872: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
6873: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
6874: }
6875: for (age=fage; age<=AGESUP; age++){
6876: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
6877: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
6878: }
6879: } /* end i status */
6880: for (i=nlstate+1; i<=nlstate+ndeath;i++){
6881: for (age=1; age<=AGESUP; age++){
6882: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
6883: mobaverage[(int)age][i][cptcod]=0.;
6884: }
6885: }
6886: }/* end cptcod */
6887: free_vector(sumnewm,1, ncovcombmax);
6888: free_vector(sumnewp,1, ncovcombmax);
6889: free_vector(agemaxgood,1, ncovcombmax);
6890: free_vector(agemingood,1, ncovcombmax);
6891: return 0;
6892: }/* End movingaverage */
1.218 brouard 6893:
1.126 brouard 6894:
6895: /************** Forecasting ******************/
1.225 brouard 6896: 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 6897: /* proj1, year, month, day of starting projection
6898: agemin, agemax range of age
6899: dateprev1 dateprev2 range of dates during which prevalence is computed
6900: anproj2 year of en of projection (same day and month as proj1).
6901: */
1.164 brouard 6902: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1;
1.126 brouard 6903: double agec; /* generic age */
6904: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
6905: double *popeffectif,*popcount;
6906: double ***p3mat;
1.218 brouard 6907: /* double ***mobaverage; */
1.126 brouard 6908: char fileresf[FILENAMELENGTH];
6909:
6910: agelim=AGESUP;
1.211 brouard 6911: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
6912: in each health status at the date of interview (if between dateprev1 and dateprev2).
6913: We still use firstpass and lastpass as another selection.
6914: */
1.214 brouard 6915: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
6916: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 6917:
1.201 brouard 6918: strcpy(fileresf,"F_");
6919: strcat(fileresf,fileresu);
1.126 brouard 6920: if((ficresf=fopen(fileresf,"w"))==NULL) {
6921: printf("Problem with forecast resultfile: %s\n", fileresf);
6922: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
6923: }
1.215 brouard 6924: printf("Computing forecasting: result on file '%s', please wait... \n", fileresf);
6925: fprintf(ficlog,"Computing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 6926:
1.225 brouard 6927: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 6928:
6929:
6930: stepsize=(int) (stepm+YEARM-1)/YEARM;
6931: if (stepm<=12) stepsize=1;
6932: if(estepm < stepm){
6933: printf ("Problem %d lower than %d\n",estepm, stepm);
6934: }
6935: else hstepm=estepm;
6936:
6937: hstepm=hstepm/stepm;
6938: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
6939: fractional in yp1 */
6940: anprojmean=yp;
6941: yp2=modf((yp1*12),&yp);
6942: mprojmean=yp;
6943: yp1=modf((yp2*30.5),&yp);
6944: jprojmean=yp;
6945: if(jprojmean==0) jprojmean=1;
6946: if(mprojmean==0) jprojmean=1;
6947:
1.227 brouard 6948: i1=pow(2,cptcoveff);
1.126 brouard 6949: if (cptcovn < 1){i1=1;}
6950:
6951: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
6952:
6953: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 6954:
1.126 brouard 6955: /* if (h==(int)(YEARM*yearp)){ */
1.227 brouard 6956: for(k=1;k<=i1;k++){
6957: if(invalidvarcomb[k]){
6958: printf("\nCombination (%d) projection ignored because no cases \n",k);
6959: continue;
6960: }
6961: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
6962: for(j=1;j<=cptcoveff;j++) {
6963: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
6964: }
6965: fprintf(ficresf," yearproj age");
6966: for(j=1; j<=nlstate+ndeath;j++){
6967: for(i=1; i<=nlstate;i++)
6968: fprintf(ficresf," p%d%d",i,j);
6969: fprintf(ficresf," wp.%d",j);
6970: }
6971: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
6972: fprintf(ficresf,"\n");
6973: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
6974: for (agec=fage; agec>=(ageminpar-1); agec--){
6975: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
6976: nhstepm = nhstepm/hstepm;
6977: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
6978: oldm=oldms;savm=savms;
6979: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k);
6980:
6981: for (h=0; h<=nhstepm; h++){
6982: if (h*hstepm/YEARM*stepm ==yearp) {
6983: fprintf(ficresf,"\n");
6984: for(j=1;j<=cptcoveff;j++)
6985: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
6986: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
6987: }
6988: for(j=1; j<=nlstate+ndeath;j++) {
6989: ppij=0.;
6990: for(i=1; i<=nlstate;i++) {
6991: if (mobilav==1)
6992: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
6993: else {
6994: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
6995: }
6996: if (h*hstepm/YEARM*stepm== yearp) {
6997: fprintf(ficresf," %.3f", p3mat[i][j][h]);
6998: }
6999: } /* end i */
7000: if (h*hstepm/YEARM*stepm==yearp) {
7001: fprintf(ficresf," %.3f", ppij);
7002: }
7003: }/* end j */
7004: } /* end h */
7005: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7006: } /* end agec */
7007: } /* end yearp */
7008: } /* end k */
1.219 brouard 7009:
1.126 brouard 7010: fclose(ficresf);
1.215 brouard 7011: printf("End of Computing forecasting \n");
7012: fprintf(ficlog,"End of Computing forecasting\n");
7013:
1.126 brouard 7014: }
7015:
1.218 brouard 7016: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7017: /* 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 7018: /* /\* back1, year, month, day of starting backection */
7019: /* agemin, agemax range of age */
7020: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7021: /* anback2 year of en of backection (same day and month as back1). */
7022: /* *\/ */
7023: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7024: /* double agec; /\* generic age *\/ */
7025: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7026: /* double *popeffectif,*popcount; */
7027: /* double ***p3mat; */
7028: /* /\* double ***mobaverage; *\/ */
7029: /* char fileresfb[FILENAMELENGTH]; */
7030:
7031: /* agelim=AGESUP; */
7032: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7033: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7034: /* We still use firstpass and lastpass as another selection. */
7035: /* *\/ */
7036: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7037: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7038: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7039:
7040: /* strcpy(fileresfb,"FB_"); */
7041: /* strcat(fileresfb,fileresu); */
7042: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7043: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7044: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7045: /* } */
7046: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7047: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7048:
1.225 brouard 7049: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7050:
7051: /* /\* if (mobilav!=0) { *\/ */
7052: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7053: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7054: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7055: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7056: /* /\* } *\/ */
7057: /* /\* } *\/ */
7058:
7059: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7060: /* if (stepm<=12) stepsize=1; */
7061: /* if(estepm < stepm){ */
7062: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7063: /* } */
7064: /* else hstepm=estepm; */
7065:
7066: /* hstepm=hstepm/stepm; */
7067: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7068: /* fractional in yp1 *\/ */
7069: /* anprojmean=yp; */
7070: /* yp2=modf((yp1*12),&yp); */
7071: /* mprojmean=yp; */
7072: /* yp1=modf((yp2*30.5),&yp); */
7073: /* jprojmean=yp; */
7074: /* if(jprojmean==0) jprojmean=1; */
7075: /* if(mprojmean==0) jprojmean=1; */
7076:
1.225 brouard 7077: /* i1=cptcoveff; */
1.218 brouard 7078: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7079:
1.218 brouard 7080: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7081:
1.218 brouard 7082: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7083:
7084: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7085: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7086: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7087: /* k=k+1; */
7088: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7089: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7090: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7091: /* } */
7092: /* fprintf(ficresfb," yearbproj age"); */
7093: /* for(j=1; j<=nlstate+ndeath;j++){ */
7094: /* for(i=1; i<=nlstate;i++) */
7095: /* fprintf(ficresfb," p%d%d",i,j); */
7096: /* fprintf(ficresfb," p.%d",j); */
7097: /* } */
7098: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7099: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7100: /* fprintf(ficresfb,"\n"); */
7101: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7102: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7103: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7104: /* nhstepm = nhstepm/hstepm; */
7105: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7106: /* oldm=oldms;savm=savms; */
7107: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7108: /* for (h=0; h<=nhstepm; h++){ */
7109: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7110: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7111: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7112: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7113: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7114: /* } */
7115: /* for(j=1; j<=nlstate+ndeath;j++) { */
7116: /* ppij=0.; */
7117: /* for(i=1; i<=nlstate;i++) { */
7118: /* if (mobilav==1) */
7119: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7120: /* else { */
7121: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7122: /* } */
7123: /* if (h*hstepm/YEARM*stepm== yearp) { */
7124: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7125: /* } */
7126: /* } /\* end i *\/ */
7127: /* if (h*hstepm/YEARM*stepm==yearp) { */
7128: /* fprintf(ficresfb," %.3f", ppij); */
7129: /* } */
7130: /* }/\* end j *\/ */
7131: /* } /\* end h *\/ */
7132: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7133: /* } /\* end agec *\/ */
7134: /* } /\* end yearp *\/ */
7135: /* } /\* end cptcod *\/ */
7136: /* } /\* end cptcov *\/ */
7137:
7138: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7139:
7140: /* fclose(ficresfb); */
7141: /* printf("End of Computing Back forecasting \n"); */
7142: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7143:
1.218 brouard 7144: /* } */
1.217 brouard 7145:
1.126 brouard 7146: /************** Forecasting *****not tested NB*************/
1.227 brouard 7147: /* 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 7148:
1.227 brouard 7149: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7150: /* int *popage; */
7151: /* double calagedatem, agelim, kk1, kk2; */
7152: /* double *popeffectif,*popcount; */
7153: /* double ***p3mat,***tabpop,***tabpopprev; */
7154: /* /\* double ***mobaverage; *\/ */
7155: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7156:
1.227 brouard 7157: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7158: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7159: /* agelim=AGESUP; */
7160: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7161:
1.227 brouard 7162: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7163:
7164:
1.227 brouard 7165: /* strcpy(filerespop,"POP_"); */
7166: /* strcat(filerespop,fileresu); */
7167: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7168: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7169: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7170: /* } */
7171: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7172: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7173:
1.227 brouard 7174: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7175:
1.227 brouard 7176: /* /\* if (mobilav!=0) { *\/ */
7177: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7178: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7179: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7180: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7181: /* /\* } *\/ */
7182: /* /\* } *\/ */
1.126 brouard 7183:
1.227 brouard 7184: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7185: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7186:
1.227 brouard 7187: /* agelim=AGESUP; */
1.126 brouard 7188:
1.227 brouard 7189: /* hstepm=1; */
7190: /* hstepm=hstepm/stepm; */
1.218 brouard 7191:
1.227 brouard 7192: /* if (popforecast==1) { */
7193: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7194: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7195: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7196: /* } */
7197: /* popage=ivector(0,AGESUP); */
7198: /* popeffectif=vector(0,AGESUP); */
7199: /* popcount=vector(0,AGESUP); */
1.126 brouard 7200:
1.227 brouard 7201: /* i=1; */
7202: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7203:
1.227 brouard 7204: /* imx=i; */
7205: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7206: /* } */
1.218 brouard 7207:
1.227 brouard 7208: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7209: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7210: /* k=k+1; */
7211: /* fprintf(ficrespop,"\n#******"); */
7212: /* for(j=1;j<=cptcoveff;j++) { */
7213: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7214: /* } */
7215: /* fprintf(ficrespop,"******\n"); */
7216: /* fprintf(ficrespop,"# Age"); */
7217: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7218: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7219:
1.227 brouard 7220: /* for (cpt=0; cpt<=0;cpt++) { */
7221: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7222:
1.227 brouard 7223: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7224: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7225: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7226:
1.227 brouard 7227: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7228: /* oldm=oldms;savm=savms; */
7229: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7230:
1.227 brouard 7231: /* for (h=0; h<=nhstepm; h++){ */
7232: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7233: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7234: /* } */
7235: /* for(j=1; j<=nlstate+ndeath;j++) { */
7236: /* kk1=0.;kk2=0; */
7237: /* for(i=1; i<=nlstate;i++) { */
7238: /* if (mobilav==1) */
7239: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7240: /* else { */
7241: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7242: /* } */
7243: /* } */
7244: /* if (h==(int)(calagedatem+12*cpt)){ */
7245: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7246: /* /\*fprintf(ficrespop," %.3f", kk1); */
7247: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7248: /* } */
7249: /* } */
7250: /* for(i=1; i<=nlstate;i++){ */
7251: /* kk1=0.; */
7252: /* for(j=1; j<=nlstate;j++){ */
7253: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7254: /* } */
7255: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7256: /* } */
1.218 brouard 7257:
1.227 brouard 7258: /* if (h==(int)(calagedatem+12*cpt)) */
7259: /* for(j=1; j<=nlstate;j++) */
7260: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7261: /* } */
7262: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7263: /* } */
7264: /* } */
1.218 brouard 7265:
1.227 brouard 7266: /* /\******\/ */
1.218 brouard 7267:
1.227 brouard 7268: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7269: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7270: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7271: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7272: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7273:
1.227 brouard 7274: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7275: /* oldm=oldms;savm=savms; */
7276: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7277: /* for (h=0; h<=nhstepm; h++){ */
7278: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7279: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7280: /* } */
7281: /* for(j=1; j<=nlstate+ndeath;j++) { */
7282: /* kk1=0.;kk2=0; */
7283: /* for(i=1; i<=nlstate;i++) { */
7284: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7285: /* } */
7286: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7287: /* } */
7288: /* } */
7289: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7290: /* } */
7291: /* } */
7292: /* } */
7293: /* } */
1.218 brouard 7294:
1.227 brouard 7295: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7296:
1.227 brouard 7297: /* if (popforecast==1) { */
7298: /* free_ivector(popage,0,AGESUP); */
7299: /* free_vector(popeffectif,0,AGESUP); */
7300: /* free_vector(popcount,0,AGESUP); */
7301: /* } */
7302: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7303: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7304: /* fclose(ficrespop); */
7305: /* } /\* End of popforecast *\/ */
1.218 brouard 7306:
1.126 brouard 7307: int fileappend(FILE *fichier, char *optionfich)
7308: {
7309: if((fichier=fopen(optionfich,"a"))==NULL) {
7310: printf("Problem with file: %s\n", optionfich);
7311: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7312: return (0);
7313: }
7314: fflush(fichier);
7315: return (1);
7316: }
7317:
7318:
7319: /**************** function prwizard **********************/
7320: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7321: {
7322:
7323: /* Wizard to print covariance matrix template */
7324:
1.164 brouard 7325: char ca[32], cb[32];
7326: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7327: int numlinepar;
7328:
7329: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7330: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7331: for(i=1; i <=nlstate; i++){
7332: jj=0;
7333: for(j=1; j <=nlstate+ndeath; j++){
7334: if(j==i) continue;
7335: jj++;
7336: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7337: printf("%1d%1d",i,j);
7338: fprintf(ficparo,"%1d%1d",i,j);
7339: for(k=1; k<=ncovmodel;k++){
7340: /* printf(" %lf",param[i][j][k]); */
7341: /* fprintf(ficparo," %lf",param[i][j][k]); */
7342: printf(" 0.");
7343: fprintf(ficparo," 0.");
7344: }
7345: printf("\n");
7346: fprintf(ficparo,"\n");
7347: }
7348: }
7349: printf("# Scales (for hessian or gradient estimation)\n");
7350: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7351: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
7352: for(i=1; i <=nlstate; i++){
7353: jj=0;
7354: for(j=1; j <=nlstate+ndeath; j++){
7355: if(j==i) continue;
7356: jj++;
7357: fprintf(ficparo,"%1d%1d",i,j);
7358: printf("%1d%1d",i,j);
7359: fflush(stdout);
7360: for(k=1; k<=ncovmodel;k++){
7361: /* printf(" %le",delti3[i][j][k]); */
7362: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7363: printf(" 0.");
7364: fprintf(ficparo," 0.");
7365: }
7366: numlinepar++;
7367: printf("\n");
7368: fprintf(ficparo,"\n");
7369: }
7370: }
7371: printf("# Covariance matrix\n");
7372: /* # 121 Var(a12)\n\ */
7373: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7374: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7375: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7376: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7377: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7378: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7379: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7380: fflush(stdout);
7381: fprintf(ficparo,"# Covariance matrix\n");
7382: /* # 121 Var(a12)\n\ */
7383: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7384: /* # ...\n\ */
7385: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7386:
7387: for(itimes=1;itimes<=2;itimes++){
7388: jj=0;
7389: for(i=1; i <=nlstate; i++){
7390: for(j=1; j <=nlstate+ndeath; j++){
7391: if(j==i) continue;
7392: for(k=1; k<=ncovmodel;k++){
7393: jj++;
7394: ca[0]= k+'a'-1;ca[1]='\0';
7395: if(itimes==1){
7396: printf("#%1d%1d%d",i,j,k);
7397: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7398: }else{
7399: printf("%1d%1d%d",i,j,k);
7400: fprintf(ficparo,"%1d%1d%d",i,j,k);
7401: /* printf(" %.5le",matcov[i][j]); */
7402: }
7403: ll=0;
7404: for(li=1;li <=nlstate; li++){
7405: for(lj=1;lj <=nlstate+ndeath; lj++){
7406: if(lj==li) continue;
7407: for(lk=1;lk<=ncovmodel;lk++){
7408: ll++;
7409: if(ll<=jj){
7410: cb[0]= lk +'a'-1;cb[1]='\0';
7411: if(ll<jj){
7412: if(itimes==1){
7413: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7414: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7415: }else{
7416: printf(" 0.");
7417: fprintf(ficparo," 0.");
7418: }
7419: }else{
7420: if(itimes==1){
7421: printf(" Var(%s%1d%1d)",ca,i,j);
7422: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
7423: }else{
7424: printf(" 0.");
7425: fprintf(ficparo," 0.");
7426: }
7427: }
7428: }
7429: } /* end lk */
7430: } /* end lj */
7431: } /* end li */
7432: printf("\n");
7433: fprintf(ficparo,"\n");
7434: numlinepar++;
7435: } /* end k*/
7436: } /*end j */
7437: } /* end i */
7438: } /* end itimes */
7439:
7440: } /* end of prwizard */
7441: /******************* Gompertz Likelihood ******************************/
7442: double gompertz(double x[])
7443: {
7444: double A,B,L=0.0,sump=0.,num=0.;
7445: int i,n=0; /* n is the size of the sample */
7446:
1.220 brouard 7447: for (i=1;i<=imx ; i++) {
1.126 brouard 7448: sump=sump+weight[i];
7449: /* sump=sump+1;*/
7450: num=num+1;
7451: }
7452:
7453:
7454: /* for (i=0; i<=imx; i++)
7455: 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]);*/
7456:
7457: for (i=1;i<=imx ; i++)
7458: {
7459: if (cens[i] == 1 && wav[i]>1)
7460: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
7461:
7462: if (cens[i] == 0 && wav[i]>1)
7463: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
7464: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
7465:
7466: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7467: if (wav[i] > 1 ) { /* ??? */
7468: L=L+A*weight[i];
7469: /* 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]);*/
7470: }
7471: }
7472:
7473: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7474:
7475: return -2*L*num/sump;
7476: }
7477:
1.136 brouard 7478: #ifdef GSL
7479: /******************* Gompertz_f Likelihood ******************************/
7480: double gompertz_f(const gsl_vector *v, void *params)
7481: {
7482: double A,B,LL=0.0,sump=0.,num=0.;
7483: double *x= (double *) v->data;
7484: int i,n=0; /* n is the size of the sample */
7485:
7486: for (i=0;i<=imx-1 ; i++) {
7487: sump=sump+weight[i];
7488: /* sump=sump+1;*/
7489: num=num+1;
7490: }
7491:
7492:
7493: /* for (i=0; i<=imx; i++)
7494: 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]);*/
7495: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
7496: for (i=1;i<=imx ; i++)
7497: {
7498: if (cens[i] == 1 && wav[i]>1)
7499: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
7500:
7501: if (cens[i] == 0 && wav[i]>1)
7502: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
7503: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
7504:
7505: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7506: if (wav[i] > 1 ) { /* ??? */
7507: LL=LL+A*weight[i];
7508: /* 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]);*/
7509: }
7510: }
7511:
7512: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7513: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
7514:
7515: return -2*LL*num/sump;
7516: }
7517: #endif
7518:
1.126 brouard 7519: /******************* Printing html file ***********/
1.201 brouard 7520: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 7521: int lastpass, int stepm, int weightopt, char model[],\
7522: int imx, double p[],double **matcov,double agemortsup){
7523: int i,k;
7524:
7525: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
7526: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
7527: for (i=1;i<=2;i++)
7528: 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 7529: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 7530: fprintf(fichtm,"</ul>");
7531:
7532: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
7533:
7534: 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>");
7535:
7536: for (k=agegomp;k<(agemortsup-2);k++)
7537: 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]);
7538:
7539:
7540: fflush(fichtm);
7541: }
7542:
7543: /******************* Gnuplot file **************/
1.201 brouard 7544: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 7545:
7546: char dirfileres[132],optfileres[132];
1.164 brouard 7547:
1.126 brouard 7548: int ng;
7549:
7550:
7551: /*#ifdef windows */
7552: fprintf(ficgp,"cd \"%s\" \n",pathc);
7553: /*#endif */
7554:
7555:
7556: strcpy(dirfileres,optionfilefiname);
7557: strcpy(optfileres,"vpl");
1.199 brouard 7558: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 7559: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 7560: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 7561: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 7562: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
7563:
7564: }
7565:
1.136 brouard 7566: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
7567: {
1.126 brouard 7568:
1.136 brouard 7569: /*-------- data file ----------*/
7570: FILE *fic;
7571: char dummy[]=" ";
1.223 brouard 7572: int i=0, j=0, n=0, iv=0;
7573: int lstra;
1.136 brouard 7574: int linei, month, year,iout;
7575: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 7576: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 7577: char *stratrunc;
1.223 brouard 7578:
1.126 brouard 7579:
7580:
1.136 brouard 7581: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 7582: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
7583: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 7584: }
1.126 brouard 7585:
1.136 brouard 7586: i=1;
7587: linei=0;
7588: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
7589: linei=linei+1;
7590: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
7591: if(line[j] == '\t')
7592: line[j] = ' ';
7593: }
7594: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
7595: ;
7596: };
7597: line[j+1]=0; /* Trims blanks at end of line */
7598: if(line[0]=='#'){
7599: fprintf(ficlog,"Comment line\n%s\n",line);
7600: printf("Comment line\n%s\n",line);
7601: continue;
7602: }
7603: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 7604: strcpy(line, linetmp);
1.223 brouard 7605:
7606: /* Loops on waves */
7607: for (j=maxwav;j>=1;j--){
7608: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.232 brouard 7609: cutv(stra, strb, line, ' ');
7610: if(strb[0]=='.') { /* Missing value */
7611: lval=-1;
7612: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
7613: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
7614: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
7615: 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);
7616: 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);
7617: return 1;
7618: }
7619: }else{
7620: errno=0;
7621: /* what_kind_of_number(strb); */
7622: dval=strtod(strb,&endptr);
7623: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
7624: /* if(strb != endptr && *endptr == '\0') */
7625: /* dval=dlval; */
7626: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
7627: if( strb[0]=='\0' || (*endptr != '\0')){
7628: 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);
7629: 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);
7630: return 1;
7631: }
7632: cotqvar[j][iv][i]=dval;
7633: cotvar[j][ntv+iv][i]=dval;
7634: }
7635: strcpy(line,stra);
1.223 brouard 7636: }/* end loop ntqv */
1.225 brouard 7637:
1.223 brouard 7638: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.232 brouard 7639: cutv(stra, strb, line, ' ');
7640: if(strb[0]=='.') { /* Missing value */
7641: lval=-1;
7642: }else{
7643: errno=0;
7644: lval=strtol(strb,&endptr,10);
7645: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
7646: if( strb[0]=='\0' || (*endptr != '\0')){
7647: 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);
7648: 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);
7649: return 1;
7650: }
7651: }
7652: if(lval <-1 || lval >1){
7653: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 7654: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7655: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.232 brouard 7656: For example, for multinomial values like 1, 2 and 3,\n \
7657: build V1=0 V2=0 for the reference value (1),\n \
7658: V1=1 V2=0 for (2) \n \
1.223 brouard 7659: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.232 brouard 7660: output of IMaCh is often meaningless.\n \
1.223 brouard 7661: Exiting.\n",lval,linei, i,line,j);
1.232 brouard 7662: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 7663: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7664: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.232 brouard 7665: For example, for multinomial values like 1, 2 and 3,\n \
7666: build V1=0 V2=0 for the reference value (1),\n \
7667: V1=1 V2=0 for (2) \n \
1.223 brouard 7668: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.232 brouard 7669: output of IMaCh is often meaningless.\n \
1.223 brouard 7670: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.232 brouard 7671: return 1;
7672: }
7673: cotvar[j][iv][i]=(double)(lval);
7674: strcpy(line,stra);
1.223 brouard 7675: }/* end loop ntv */
1.225 brouard 7676:
1.223 brouard 7677: /* Statuses at wave */
1.137 brouard 7678: cutv(stra, strb, line, ' ');
1.223 brouard 7679: if(strb[0]=='.') { /* Missing value */
1.232 brouard 7680: lval=-1;
1.136 brouard 7681: }else{
1.232 brouard 7682: errno=0;
7683: lval=strtol(strb,&endptr,10);
7684: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
7685: if( strb[0]=='\0' || (*endptr != '\0')){
7686: 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);
7687: 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);
7688: return 1;
7689: }
1.136 brouard 7690: }
1.225 brouard 7691:
1.136 brouard 7692: s[j][i]=lval;
1.225 brouard 7693:
1.223 brouard 7694: /* Date of Interview */
1.136 brouard 7695: strcpy(line,stra);
7696: cutv(stra, strb,line,' ');
1.169 brouard 7697: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7698: }
1.169 brouard 7699: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 7700: month=99;
7701: year=9999;
1.136 brouard 7702: }else{
1.225 brouard 7703: 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);
7704: 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);
7705: return 1;
1.136 brouard 7706: }
7707: anint[j][i]= (double) year;
7708: mint[j][i]= (double)month;
7709: strcpy(line,stra);
1.223 brouard 7710: } /* End loop on waves */
1.225 brouard 7711:
1.223 brouard 7712: /* Date of death */
1.136 brouard 7713: cutv(stra, strb,line,' ');
1.169 brouard 7714: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7715: }
1.169 brouard 7716: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 7717: month=99;
7718: year=9999;
7719: }else{
1.141 brouard 7720: 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 7721: 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);
7722: return 1;
1.136 brouard 7723: }
7724: andc[i]=(double) year;
7725: moisdc[i]=(double) month;
7726: strcpy(line,stra);
7727:
1.223 brouard 7728: /* Date of birth */
1.136 brouard 7729: cutv(stra, strb,line,' ');
1.169 brouard 7730: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 7731: }
1.169 brouard 7732: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 7733: month=99;
7734: year=9999;
7735: }else{
1.141 brouard 7736: 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);
7737: 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 7738: return 1;
1.136 brouard 7739: }
7740: if (year==9999) {
1.141 brouard 7741: 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);
7742: 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 7743: return 1;
7744:
1.136 brouard 7745: }
7746: annais[i]=(double)(year);
7747: moisnais[i]=(double)(month);
7748: strcpy(line,stra);
1.225 brouard 7749:
1.223 brouard 7750: /* Sample weight */
1.136 brouard 7751: cutv(stra, strb,line,' ');
7752: errno=0;
7753: dval=strtod(strb,&endptr);
7754: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 7755: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
7756: 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 7757: fflush(ficlog);
7758: return 1;
7759: }
7760: weight[i]=dval;
7761: strcpy(line,stra);
1.225 brouard 7762:
1.223 brouard 7763: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
7764: cutv(stra, strb, line, ' ');
7765: if(strb[0]=='.') { /* Missing value */
1.225 brouard 7766: lval=-1;
1.223 brouard 7767: }else{
1.225 brouard 7768: errno=0;
7769: /* what_kind_of_number(strb); */
7770: dval=strtod(strb,&endptr);
7771: /* if(strb != endptr && *endptr == '\0') */
7772: /* dval=dlval; */
7773: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
7774: if( strb[0]=='\0' || (*endptr != '\0')){
7775: 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);
7776: 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);
7777: return 1;
7778: }
7779: coqvar[iv][i]=dval;
1.226 brouard 7780: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 7781: }
7782: strcpy(line,stra);
7783: }/* end loop nqv */
1.136 brouard 7784:
1.223 brouard 7785: /* Covariate values */
1.136 brouard 7786: for (j=ncovcol;j>=1;j--){
7787: cutv(stra, strb,line,' ');
1.223 brouard 7788: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 7789: lval=-1;
1.136 brouard 7790: }else{
1.225 brouard 7791: errno=0;
7792: lval=strtol(strb,&endptr,10);
7793: if( strb[0]=='\0' || (*endptr != '\0')){
7794: 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);
7795: 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);
7796: return 1;
7797: }
1.136 brouard 7798: }
7799: if(lval <-1 || lval >1){
1.225 brouard 7800: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 7801: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7802: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 7803: For example, for multinomial values like 1, 2 and 3,\n \
7804: build V1=0 V2=0 for the reference value (1),\n \
7805: V1=1 V2=0 for (2) \n \
1.136 brouard 7806: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 7807: output of IMaCh is often meaningless.\n \
1.136 brouard 7808: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 7809: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 7810: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
7811: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 7812: For example, for multinomial values like 1, 2 and 3,\n \
7813: build V1=0 V2=0 for the reference value (1),\n \
7814: V1=1 V2=0 for (2) \n \
1.136 brouard 7815: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 7816: output of IMaCh is often meaningless.\n \
1.136 brouard 7817: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 7818: return 1;
1.136 brouard 7819: }
7820: covar[j][i]=(double)(lval);
7821: strcpy(line,stra);
7822: }
7823: lstra=strlen(stra);
1.225 brouard 7824:
1.136 brouard 7825: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
7826: stratrunc = &(stra[lstra-9]);
7827: num[i]=atol(stratrunc);
7828: }
7829: else
7830: num[i]=atol(stra);
7831: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
7832: 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;}*/
7833:
7834: i=i+1;
7835: } /* End loop reading data */
1.225 brouard 7836:
1.136 brouard 7837: *imax=i-1; /* Number of individuals */
7838: fclose(fic);
1.225 brouard 7839:
1.136 brouard 7840: return (0);
1.164 brouard 7841: /* endread: */
1.225 brouard 7842: printf("Exiting readdata: ");
7843: fclose(fic);
7844: return (1);
1.223 brouard 7845: }
1.126 brouard 7846:
1.230 brouard 7847: void removespace(char **stri){/*, char stro[]) {*/
7848: char *p1 = *stri, *p2 = *stri;
1.145 brouard 7849: do
7850: while (*p2 == ' ')
7851: p2++;
1.169 brouard 7852: while (*p1++ == *p2++);
1.230 brouard 7853: *stri=p1;
1.145 brouard 7854: }
7855:
1.230 brouard 7856: int decoderesult ( char resultline[])
7857: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
7858: {
7859: int j=0, k=0;
7860: char resultsav[MAXLINE];
7861: char stra[80], strb[80], strc[80], strd[80],stre[80];
7862:
7863: removespace(&resultline);
1.233 ! brouard 7864: printf("decoderesult:%s\n",resultline);
1.230 brouard 7865:
7866: if (strstr(resultline,"v") !=0){
7867: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
7868: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
7869: return 1;
7870: }
7871: trimbb(resultsav, resultline);
7872: if (strlen(resultsav) >1){
7873: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
7874: }
7875:
7876: for(k=1; k<=j;k++){ /* Loop on total covariates of the model */
7877: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
7878: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
7879: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
7880: Tvalsel[k]=atof(strc); /* 1 */
7881:
7882: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
7883: Tvarsel[k]=atoi(strc);
7884: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
7885: /* cptcovsel++; */
7886: if (nbocc(stra,'=') >0)
7887: strcpy(resultsav,stra); /* and analyzes it */
7888: }
7889: return (0);
7890: }
7891: int selected( int kvar){ /* Selected combination of covariates */
7892: if(Tvarsel[kvar])
7893: return (0);
7894: else
7895: return(1);
7896: }
7897: int decodemodel( char model[], int lastobs)
7898: /**< This routine decodes the model and returns:
1.224 brouard 7899: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
7900: * - nagesqr = 1 if age*age in the model, otherwise 0.
7901: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
7902: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
7903: * - cptcovage number of covariates with age*products =2
7904: * - cptcovs number of simple covariates
7905: * - 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
7906: * which is a new column after the 9 (ncovcol) variables.
7907: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
7908: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
7909: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
7910: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
7911: */
1.136 brouard 7912: {
1.145 brouard 7913: int i, j, k, ks;
1.227 brouard 7914: int j1, k1, k2, k3, k4;
1.136 brouard 7915: char modelsav[80];
1.145 brouard 7916: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 7917: char *strpt;
1.136 brouard 7918:
1.145 brouard 7919: /*removespace(model);*/
1.136 brouard 7920: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 7921: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 7922: if (strstr(model,"AGE") !=0){
1.192 brouard 7923: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
7924: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 7925: return 1;
7926: }
1.141 brouard 7927: if (strstr(model,"v") !=0){
7928: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
7929: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
7930: return 1;
7931: }
1.187 brouard 7932: strcpy(modelsav,model);
7933: if ((strpt=strstr(model,"age*age")) !=0){
7934: printf(" strpt=%s, model=%s\n",strpt, model);
7935: if(strpt != model){
1.231 brouard 7936: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 7937: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 7938: corresponding column of parameters.\n",model);
1.231 brouard 7939: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 7940: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 7941: corresponding column of parameters.\n",model); fflush(ficlog);
1.231 brouard 7942: return 1;
1.225 brouard 7943: }
1.187 brouard 7944: nagesqr=1;
7945: if (strstr(model,"+age*age") !=0)
1.231 brouard 7946: substrchaine(modelsav, model, "+age*age");
1.187 brouard 7947: else if (strstr(model,"age*age+") !=0)
1.231 brouard 7948: substrchaine(modelsav, model, "age*age+");
1.187 brouard 7949: else
1.231 brouard 7950: substrchaine(modelsav, model, "age*age");
1.187 brouard 7951: }else
7952: nagesqr=0;
7953: if (strlen(modelsav) >1){
7954: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
7955: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 7956: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 7957: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 7958: * cst, age and age*age
7959: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
7960: /* including age products which are counted in cptcovage.
7961: * but the covariates which are products must be treated
7962: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 7963: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
7964: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 7965:
7966:
1.187 brouard 7967: /* Design
7968: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
7969: * < ncovcol=8 >
7970: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
7971: * k= 1 2 3 4 5 6 7 8
7972: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
7973: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 7974: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
7975: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 7976: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
7977: * Tage[++cptcovage]=k
7978: * if products, new covar are created after ncovcol with k1
7979: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
7980: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
7981: * 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
7982: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
7983: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
7984: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
7985: * < ncovcol=8 >
7986: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
7987: * k= 1 2 3 4 5 6 7 8 9 10 11 12
7988: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
7989: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
7990: * p Tprod[1]@2={ 6, 5}
7991: *p Tvard[1][1]@4= {7, 8, 5, 6}
7992: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
7993: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
7994: *How to reorganize?
7995: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
7996: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
7997: * {2, 1, 4, 8, 5, 6, 3, 7}
7998: * Struct []
7999: */
1.225 brouard 8000:
1.187 brouard 8001: /* This loop fills the array Tvar from the string 'model'.*/
8002: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8003: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8004: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8005: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8006: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8007: /* k=1 Tvar[1]=2 (from V2) */
8008: /* k=5 Tvar[5] */
8009: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8010: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8011: /* } */
1.198 brouard 8012: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8013: /*
8014: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8015: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8016: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8017: }
1.187 brouard 8018: cptcovage=0;
8019: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.232 brouard 8020: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8021: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.232 brouard 8022: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8023: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8024: /*scanf("%d",i);*/
8025: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8026: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8027: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8028: /* covar is not filled and then is empty */
8029: cptcovprod--;
8030: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8031: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8032: Typevar[k]=1; /* 1 for age product */
8033: cptcovage++; /* Sums the number of covariates which include age as a product */
8034: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8035: /*printf("stre=%s ", stre);*/
8036: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8037: cptcovprod--;
8038: cutl(stre,strb,strc,'V');
8039: Tvar[k]=atoi(stre);
8040: Typevar[k]=1; /* 1 for age product */
8041: cptcovage++;
8042: Tage[cptcovage]=k;
8043: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8044: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8045: cptcovn++;
8046: cptcovprodnoage++;k1++;
8047: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8048: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8049: because this model-covariate is a construction we invent a new column
8050: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8051: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8052: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8053: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8054: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8055: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8056: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8057: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8058: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8059: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8060: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8061: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8062: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.232 brouard 8063: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8064: for (i=1; i<=lastobs;i++){
8065: /* Computes the new covariate which is a product of
8066: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8067: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8068: }
8069: } /* End age is not in the model */
8070: } /* End if model includes a product */
8071: else { /* no more sum */
8072: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8073: /* scanf("%d",i);*/
8074: cutl(strd,strc,strb,'V');
8075: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8076: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8077: Tvar[k]=atoi(strd);
8078: Typevar[k]=0; /* 0 for simple covariates */
8079: }
8080: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8081: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8082: scanf("%d",i);*/
1.187 brouard 8083: } /* end of loop + on total covariates */
8084: } /* end if strlen(modelsave == 0) age*age might exist */
8085: } /* end if strlen(model == 0) */
1.136 brouard 8086:
8087: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8088: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8089:
1.136 brouard 8090: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8091: printf("cptcovprod=%d ", cptcovprod);
8092: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8093: scanf("%d ",i);*/
8094:
8095:
1.230 brouard 8096: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8097: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8098: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8099: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8100: k = 1 2 3 4 5 6 7 8 9
8101: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8102: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8103: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8104: Dummy[k] 1 0 0 0 3 1 1 2 3
8105: Tmodelind[combination of covar]=k;
1.225 brouard 8106: */
8107: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8108: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8109: /* 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 8110: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8111: printf("Model=%s\n\
8112: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8113: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8114: 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);
8115: fprintf(ficlog,"Model=%s\n\
8116: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8117: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8118: 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);
8119:
1.232 brouard 8120: for(k=1, ncovf=0, ncovv=0, ncova=0, ncoveff=0, nqfveff=0, ntveff=0, nqtveff=0;k<=cptcovt; k++){ /* or cptocvt */
1.231 brouard 8121: if (Tvar[k] <=ncovcol && (Typevar[k]==0 || Typevar[k]==2)){ /* Simple or product fixed dummy (<=ncovcol) covariates */
1.227 brouard 8122: Fixed[k]= 0;
8123: Dummy[k]= 0;
1.225 brouard 8124: ncoveff++;
1.232 brouard 8125: ncovf++;
1.231 brouard 8126: modell[k].maintype= FTYPE;
1.232 brouard 8127: TvarF[ncovf]=Tvar[k];
8128: TvarFind[ncovf]=k;
1.230 brouard 8129: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8130: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8131: }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 8132: Fixed[k]= 0;
8133: Dummy[k]= 1;
1.230 brouard 8134: nqfveff++;
1.231 brouard 8135: modell[k].maintype= FTYPE;
8136: modell[k].subtype= FQ;
1.232 brouard 8137: ncovf++;
8138: TvarF[ncovf]=Tvar[k];
8139: TvarFind[ncovf]=k;
1.231 brouard 8140: 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 8141: 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 */
1.225 brouard 8142: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){
1.227 brouard 8143: Fixed[k]= 1;
8144: Dummy[k]= 0;
1.225 brouard 8145: ntveff++; /* Only simple time varying dummy variable */
1.231 brouard 8146: modell[k].maintype= VTYPE;
8147: modell[k].subtype= VD;
1.232 brouard 8148: ncovv++; /* Only simple time varying variables */
8149: TvarV[ncovv]=Tvar[k];
8150: TvarVind[ncovv]=k;
1.231 brouard 8151: 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 */
8152: 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 8153: 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);
8154: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8155: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
8156: Fixed[k]= 1;
8157: Dummy[k]= 1;
8158: nqtveff++;
8159: modell[k].maintype= VTYPE;
8160: modell[k].subtype= VQ;
1.232 brouard 8161: ncovv++; /* Only simple time varying variables */
8162: TvarV[ncovv]=Tvar[k];
8163: TvarVind[ncovv]=k;
1.231 brouard 8164: 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 */
8165: 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 */
8166: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8167: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8168: 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 8169: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8170: }else if (Typevar[k] == 1) { /* product with age */
1.232 brouard 8171: ncova++;
8172: TvarA[ncova]=Tvar[k];
8173: TvarAind[ncova]=k;
1.231 brouard 8174: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
8175: Fixed[k]= 2;
8176: Dummy[k]= 2;
8177: modell[k].maintype= ATYPE;
8178: modell[k].subtype= APFD;
8179: /* ncoveff++; */
1.227 brouard 8180: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.231 brouard 8181: Fixed[k]= 2;
8182: Dummy[k]= 3;
8183: modell[k].maintype= ATYPE;
8184: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8185: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8186: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.231 brouard 8187: Fixed[k]= 3;
8188: Dummy[k]= 2;
8189: modell[k].maintype= ATYPE;
8190: modell[k].subtype= APVD; /* Product age * varying dummy */
8191: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8192: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.231 brouard 8193: Fixed[k]= 3;
8194: Dummy[k]= 3;
8195: modell[k].maintype= ATYPE;
8196: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8197: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8198: }
8199: }else if (Typevar[k] == 2) { /* product without age */
8200: k1=Tposprod[k];
8201: if(Tvard[k1][1] <=ncovcol){
1.231 brouard 8202: if(Tvard[k1][2] <=ncovcol){
8203: Fixed[k]= 1;
8204: Dummy[k]= 0;
8205: modell[k].maintype= FTYPE;
8206: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
1.233 ! brouard 8207: ncovf++; /* Fixed variables without age */
! 8208: TvarF[ncovf]=Tvar[k];
! 8209: TvarFind[ncovf]=k;
1.231 brouard 8210: }else if(Tvard[k1][2] <=ncovcol+nqv){
8211: Fixed[k]= 0; /* or 2 ?*/
8212: Dummy[k]= 1;
8213: modell[k].maintype= FTYPE;
8214: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
1.233 ! brouard 8215: ncovf++; /* Varying variables without age */
! 8216: TvarF[ncovf]=Tvar[k];
! 8217: TvarFind[ncovf]=k;
1.231 brouard 8218: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8219: Fixed[k]= 1;
8220: Dummy[k]= 0;
8221: modell[k].maintype= VTYPE;
8222: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
1.233 ! brouard 8223: ncovv++; /* Varying variables without age */
! 8224: TvarV[ncovv]=Tvar[k];
! 8225: TvarVind[ncovv]=k;
1.231 brouard 8226: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8227: Fixed[k]= 1;
8228: Dummy[k]= 1;
8229: modell[k].maintype= VTYPE;
8230: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
1.233 ! brouard 8231: ncovv++; /* Varying variables without age */
! 8232: TvarV[ncovv]=Tvar[k];
! 8233: TvarVind[ncovv]=k;
1.231 brouard 8234: }
1.227 brouard 8235: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.231 brouard 8236: if(Tvard[k1][2] <=ncovcol){
8237: Fixed[k]= 0; /* or 2 ?*/
8238: Dummy[k]= 1;
8239: modell[k].maintype= FTYPE;
8240: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
1.233 ! brouard 8241: ncovf++; /* Fixed variables without age */
! 8242: TvarF[ncovf]=Tvar[k];
! 8243: TvarFind[ncovf]=k;
1.231 brouard 8244: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8245: Fixed[k]= 1;
8246: Dummy[k]= 1;
8247: modell[k].maintype= VTYPE;
8248: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
1.233 ! brouard 8249: ncovv++; /* Varying variables without age */
! 8250: TvarV[ncovv]=Tvar[k];
! 8251: TvarVind[ncovv]=k;
1.231 brouard 8252: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8253: Fixed[k]= 1;
8254: Dummy[k]= 1;
8255: modell[k].maintype= VTYPE;
8256: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
1.233 ! brouard 8257: ncovv++; /* Varying variables without age */
! 8258: TvarV[ncovv]=Tvar[k];
! 8259: TvarVind[ncovv]=k;
! 8260: ncovv++; /* Varying variables without age */
! 8261: TvarV[ncovv]=Tvar[k];
! 8262: TvarVind[ncovv]=k;
1.231 brouard 8263: }
1.227 brouard 8264: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.231 brouard 8265: if(Tvard[k1][2] <=ncovcol){
8266: Fixed[k]= 1;
8267: Dummy[k]= 1;
8268: modell[k].maintype= VTYPE;
8269: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
1.233 ! brouard 8270: ncovv++; /* Varying variables without age */
! 8271: TvarV[ncovv]=Tvar[k];
! 8272: TvarVind[ncovv]=k;
1.231 brouard 8273: }else if(Tvard[k1][2] <=ncovcol+nqv){
8274: Fixed[k]= 1;
8275: Dummy[k]= 1;
8276: modell[k].maintype= VTYPE;
8277: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
1.233 ! brouard 8278: ncovv++; /* Varying variables without age */
! 8279: TvarV[ncovv]=Tvar[k];
! 8280: TvarVind[ncovv]=k;
1.231 brouard 8281: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8282: Fixed[k]= 1;
8283: Dummy[k]= 0;
8284: modell[k].maintype= VTYPE;
8285: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
1.233 ! brouard 8286: ncovv++; /* Varying variables without age */
! 8287: TvarV[ncovv]=Tvar[k];
! 8288: TvarVind[ncovv]=k;
1.231 brouard 8289: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8290: Fixed[k]= 1;
8291: Dummy[k]= 1;
8292: modell[k].maintype= VTYPE;
8293: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
1.233 ! brouard 8294: ncovv++; /* Varying variables without age */
! 8295: TvarV[ncovv]=Tvar[k];
! 8296: TvarVind[ncovv]=k;
1.231 brouard 8297: }
1.227 brouard 8298: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.231 brouard 8299: if(Tvard[k1][2] <=ncovcol){
8300: Fixed[k]= 1;
8301: Dummy[k]= 1;
8302: modell[k].maintype= VTYPE;
8303: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
1.233 ! brouard 8304: ncovv++; /* Varying variables without age */
! 8305: TvarV[ncovv]=Tvar[k];
! 8306: TvarVind[ncovv]=k;
1.231 brouard 8307: }else if(Tvard[k1][2] <=ncovcol+nqv){
8308: Fixed[k]= 1;
8309: Dummy[k]= 1;
8310: modell[k].maintype= VTYPE;
8311: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
1.233 ! brouard 8312: ncovv++; /* Varying variables without age */
! 8313: TvarV[ncovv]=Tvar[k];
! 8314: TvarVind[ncovv]=k;
1.231 brouard 8315: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8316: Fixed[k]= 1;
8317: Dummy[k]= 1;
8318: modell[k].maintype= VTYPE;
8319: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
1.233 ! brouard 8320: ncovv++; /* Varying variables without age */
! 8321: TvarV[ncovv]=Tvar[k];
! 8322: TvarVind[ncovv]=k;
1.231 brouard 8323: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8324: Fixed[k]= 1;
8325: Dummy[k]= 1;
8326: modell[k].maintype= VTYPE;
8327: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
1.233 ! brouard 8328: ncovv++; /* Varying variables without age */
! 8329: TvarV[ncovv]=Tvar[k];
! 8330: TvarVind[ncovv]=k;
1.231 brouard 8331: }
1.227 brouard 8332: }else{
1.231 brouard 8333: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8334: 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 8335: } /* end k1 */
1.225 brouard 8336: }else{
1.226 brouard 8337: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
8338: 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 8339: }
1.227 brouard 8340: 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 8341: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 8342: 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]);
8343: }
8344: /* Searching for doublons in the model */
8345: for(k1=1; k1<= cptcovt;k1++){
8346: for(k2=1; k2 <k1;k2++){
8347: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.231 brouard 8348: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
8349: if(Tvar[k1]==Tvar[k2]){
8350: 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]]);
8351: 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);
8352: return(1);
8353: }
8354: }else if (Typevar[k1] ==2){
8355: k3=Tposprod[k1];
8356: k4=Tposprod[k2];
8357: 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])) ){
8358: 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]]);
8359: 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);
8360: return(1);
8361: }
8362: }
1.227 brouard 8363: }
8364: }
1.225 brouard 8365: }
8366: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
8367: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.232 brouard 8368: printf("ncovf=%d, ncovv=%d, ncova=%d\n",ncovf,ncovv,ncova);
8369: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d\n",ncovf,ncovv,ncova);
1.137 brouard 8370: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 8371: /*endread:*/
1.225 brouard 8372: printf("Exiting decodemodel: ");
8373: return (1);
1.136 brouard 8374: }
8375:
1.169 brouard 8376: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.136 brouard 8377: {
8378: int i, m;
1.218 brouard 8379: int firstone=0;
8380:
1.136 brouard 8381: for (i=1; i<=imx; i++) {
8382: for(m=2; (m<= maxwav); m++) {
8383: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
8384: anint[m][i]=9999;
1.216 brouard 8385: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
8386: s[m][i]=-1;
1.136 brouard 8387: }
8388: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.169 brouard 8389: *nberr = *nberr + 1;
1.218 brouard 8390: if(firstone == 0){
8391: firstone=1;
8392: 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);
8393: }
8394: 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 8395: s[m][i]=-1;
8396: }
8397: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 8398: (*nberr)++;
1.136 brouard 8399: 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]);
8400: 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]);
8401: s[m][i]=-1; /* We prefer to skip it (and to skip it in version 0.8a1 too */
8402: }
8403: }
8404: }
8405:
8406: for (i=1; i<=imx; i++) {
8407: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
8408: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 8409: 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 8410: if (s[m][i] >= nlstate+1) {
1.169 brouard 8411: if(agedc[i]>0){
8412: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 8413: agev[m][i]=agedc[i];
1.214 brouard 8414: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 8415: }else {
1.136 brouard 8416: if ((int)andc[i]!=9999){
8417: nbwarn++;
8418: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
8419: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
8420: agev[m][i]=-1;
8421: }
8422: }
1.169 brouard 8423: } /* agedc > 0 */
1.214 brouard 8424: } /* end if */
1.136 brouard 8425: else if(s[m][i] !=9){ /* Standard case, age in fractional
8426: years but with the precision of a month */
8427: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
8428: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
8429: agev[m][i]=1;
8430: else if(agev[m][i] < *agemin){
8431: *agemin=agev[m][i];
8432: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
8433: }
8434: else if(agev[m][i] >*agemax){
8435: *agemax=agev[m][i];
1.156 brouard 8436: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 8437: }
8438: /*agev[m][i]=anint[m][i]-annais[i];*/
8439: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 8440: } /* en if 9*/
1.136 brouard 8441: else { /* =9 */
1.214 brouard 8442: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 8443: agev[m][i]=1;
8444: s[m][i]=-1;
8445: }
8446: }
1.214 brouard 8447: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 8448: agev[m][i]=1;
1.214 brouard 8449: else{
8450: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8451: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8452: agev[m][i]=0;
8453: }
8454: } /* End for lastpass */
8455: }
1.136 brouard 8456:
8457: for (i=1; i<=imx; i++) {
8458: for(m=firstpass; (m<=lastpass); m++){
8459: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 8460: (*nberr)++;
1.136 brouard 8461: 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);
8462: 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);
8463: return 1;
8464: }
8465: }
8466: }
8467:
8468: /*for (i=1; i<=imx; i++){
8469: for (m=firstpass; (m<lastpass); m++){
8470: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
8471: }
8472:
8473: }*/
8474:
8475:
1.139 brouard 8476: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
8477: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 8478:
8479: return (0);
1.164 brouard 8480: /* endread:*/
1.136 brouard 8481: printf("Exiting calandcheckages: ");
8482: return (1);
8483: }
8484:
1.172 brouard 8485: #if defined(_MSC_VER)
8486: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8487: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8488: //#include "stdafx.h"
8489: //#include <stdio.h>
8490: //#include <tchar.h>
8491: //#include <windows.h>
8492: //#include <iostream>
8493: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
8494:
8495: LPFN_ISWOW64PROCESS fnIsWow64Process;
8496:
8497: BOOL IsWow64()
8498: {
8499: BOOL bIsWow64 = FALSE;
8500:
8501: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
8502: // (HANDLE, PBOOL);
8503:
8504: //LPFN_ISWOW64PROCESS fnIsWow64Process;
8505:
8506: HMODULE module = GetModuleHandle(_T("kernel32"));
8507: const char funcName[] = "IsWow64Process";
8508: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
8509: GetProcAddress(module, funcName);
8510:
8511: if (NULL != fnIsWow64Process)
8512: {
8513: if (!fnIsWow64Process(GetCurrentProcess(),
8514: &bIsWow64))
8515: //throw std::exception("Unknown error");
8516: printf("Unknown error\n");
8517: }
8518: return bIsWow64 != FALSE;
8519: }
8520: #endif
1.177 brouard 8521:
1.191 brouard 8522: void syscompilerinfo(int logged)
1.167 brouard 8523: {
8524: /* #include "syscompilerinfo.h"*/
1.185 brouard 8525: /* command line Intel compiler 32bit windows, XP compatible:*/
8526: /* /GS /W3 /Gy
8527: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
8528: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
8529: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 8530: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
8531: */
8532: /* 64 bits */
1.185 brouard 8533: /*
8534: /GS /W3 /Gy
8535: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
8536: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
8537: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
8538: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
8539: /* Optimization are useless and O3 is slower than O2 */
8540: /*
8541: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
8542: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
8543: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
8544: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
8545: */
1.186 brouard 8546: /* Link is */ /* /OUT:"visual studio
1.185 brouard 8547: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
8548: /PDB:"visual studio
8549: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
8550: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
8551: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
8552: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
8553: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
8554: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
8555: uiAccess='false'"
8556: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
8557: /NOLOGO /TLBID:1
8558: */
1.177 brouard 8559: #if defined __INTEL_COMPILER
1.178 brouard 8560: #if defined(__GNUC__)
8561: struct utsname sysInfo; /* For Intel on Linux and OS/X */
8562: #endif
1.177 brouard 8563: #elif defined(__GNUC__)
1.179 brouard 8564: #ifndef __APPLE__
1.174 brouard 8565: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 8566: #endif
1.177 brouard 8567: struct utsname sysInfo;
1.178 brouard 8568: int cross = CROSS;
8569: if (cross){
8570: printf("Cross-");
1.191 brouard 8571: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 8572: }
1.174 brouard 8573: #endif
8574:
1.171 brouard 8575: #include <stdint.h>
1.178 brouard 8576:
1.191 brouard 8577: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 8578: #if defined(__clang__)
1.191 brouard 8579: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 8580: #endif
8581: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 8582: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 8583: #endif
8584: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 8585: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 8586: #endif
8587: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 8588: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 8589: #endif
8590: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 8591: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 8592: #endif
8593: #if defined(_MSC_VER)
1.191 brouard 8594: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 8595: #endif
8596: #if defined(__PGI)
1.191 brouard 8597: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 8598: #endif
8599: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 8600: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 8601: #endif
1.191 brouard 8602: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 8603:
1.167 brouard 8604: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
8605: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
8606: // Windows (x64 and x86)
1.191 brouard 8607: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 8608: #elif __unix__ // all unices, not all compilers
8609: // Unix
1.191 brouard 8610: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 8611: #elif __linux__
8612: // linux
1.191 brouard 8613: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 8614: #elif __APPLE__
1.174 brouard 8615: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 8616: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 8617: #endif
8618:
8619: /* __MINGW32__ */
8620: /* __CYGWIN__ */
8621: /* __MINGW64__ */
8622: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
8623: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
8624: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
8625: /* _WIN64 // Defined for applications for Win64. */
8626: /* _M_X64 // Defined for compilations that target x64 processors. */
8627: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 8628:
1.167 brouard 8629: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 8630: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 8631: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 8632: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 8633: #else
1.191 brouard 8634: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 8635: #endif
8636:
1.169 brouard 8637: #if defined(__GNUC__)
8638: # if defined(__GNUC_PATCHLEVEL__)
8639: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
8640: + __GNUC_MINOR__ * 100 \
8641: + __GNUC_PATCHLEVEL__)
8642: # else
8643: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
8644: + __GNUC_MINOR__ * 100)
8645: # endif
1.174 brouard 8646: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 8647: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 8648:
8649: if (uname(&sysInfo) != -1) {
8650: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 8651: 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 8652: }
8653: else
8654: perror("uname() error");
1.179 brouard 8655: //#ifndef __INTEL_COMPILER
8656: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 8657: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 8658: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 8659: #endif
1.169 brouard 8660: #endif
1.172 brouard 8661:
8662: // void main()
8663: // {
1.169 brouard 8664: #if defined(_MSC_VER)
1.174 brouard 8665: if (IsWow64()){
1.191 brouard 8666: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
8667: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 8668: }
8669: else{
1.191 brouard 8670: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
8671: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 8672: }
1.172 brouard 8673: // printf("\nPress Enter to continue...");
8674: // getchar();
8675: // }
8676:
1.169 brouard 8677: #endif
8678:
1.167 brouard 8679:
1.219 brouard 8680: }
1.136 brouard 8681:
1.219 brouard 8682: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 8683: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
8684: int i, j, k, i1 ;
1.202 brouard 8685: /* double ftolpl = 1.e-10; */
1.180 brouard 8686: double age, agebase, agelim;
1.203 brouard 8687: double tot;
1.180 brouard 8688:
1.202 brouard 8689: strcpy(filerespl,"PL_");
8690: strcat(filerespl,fileresu);
8691: if((ficrespl=fopen(filerespl,"w"))==NULL) {
8692: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
8693: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
8694: }
1.227 brouard 8695: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
8696: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 8697: pstamp(ficrespl);
1.203 brouard 8698: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 8699: fprintf(ficrespl,"#Age ");
8700: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
8701: fprintf(ficrespl,"\n");
1.180 brouard 8702:
1.219 brouard 8703: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 8704:
1.219 brouard 8705: agebase=ageminpar;
8706: agelim=agemaxpar;
1.180 brouard 8707:
1.227 brouard 8708: /* i1=pow(2,ncoveff); */
8709: i1=pow(2,cptcoveff); /* Number of dummy covariates */
1.219 brouard 8710: if (cptcovn < 1){i1=1;}
1.180 brouard 8711:
1.220 brouard 8712: for(k=1; k<=i1;k++){
8713: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
1.180 brouard 8714: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
1.219 brouard 8715: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
1.220 brouard 8716: /* k=k+1; */
1.219 brouard 8717: /* to clean */
8718: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
8719: fprintf(ficrespl,"#******");
8720: printf("#******");
8721: fprintf(ficlog,"#******");
1.227 brouard 8722: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
8723: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
1.219 brouard 8724: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8725: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8726: }
8727: fprintf(ficrespl,"******\n");
8728: printf("******\n");
8729: fprintf(ficlog,"******\n");
1.227 brouard 8730: if(invalidvarcomb[k]){
8731: printf("\nCombination (%d) ignored because no case \n",k);
8732: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
8733: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
1.220 brouard 8734: continue;
1.227 brouard 8735: }
1.219 brouard 8736:
8737: fprintf(ficrespl,"#Age ");
1.227 brouard 8738: for(j=1;j<=cptcoveff;j++) {
1.219 brouard 8739: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8740: }
8741: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
8742: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 8743:
1.219 brouard 8744: for (age=agebase; age<=agelim; age++){
8745: /* for (age=agebase; age<=agebase; age++){ */
8746: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k);
8747: fprintf(ficrespl,"%.0f ",age );
1.227 brouard 8748: for(j=1;j<=cptcoveff;j++)
8749: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.219 brouard 8750: tot=0.;
8751: for(i=1; i<=nlstate;i++){
1.227 brouard 8752: tot += prlim[i][i];
8753: fprintf(ficrespl," %.5f", prlim[i][i]);
1.219 brouard 8754: }
8755: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
8756: } /* Age */
8757: /* was end of cptcod */
8758: } /* cptcov */
8759: return 0;
1.180 brouard 8760: }
8761:
1.218 brouard 8762: 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){
8763: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
8764:
8765: /* Computes the back prevalence limit for any combination of covariate values
8766: * at any age between ageminpar and agemaxpar
8767: */
1.217 brouard 8768: int i, j, k, i1 ;
8769: /* double ftolpl = 1.e-10; */
8770: double age, agebase, agelim;
8771: double tot;
1.218 brouard 8772: /* double ***mobaverage; */
8773: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 8774:
8775: strcpy(fileresplb,"PLB_");
8776: strcat(fileresplb,fileresu);
8777: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
8778: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
8779: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
8780: }
8781: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
8782: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
8783: pstamp(ficresplb);
8784: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
8785: fprintf(ficresplb,"#Age ");
8786: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
8787: fprintf(ficresplb,"\n");
8788:
1.218 brouard 8789:
8790: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
8791:
8792: agebase=ageminpar;
8793: agelim=agemaxpar;
8794:
8795:
1.227 brouard 8796: i1=pow(2,cptcoveff);
1.218 brouard 8797: if (cptcovn < 1){i1=1;}
1.227 brouard 8798:
8799: for(k=1; k<=i1;k++){
1.218 brouard 8800: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
8801: fprintf(ficresplb,"#******");
8802: printf("#******");
8803: fprintf(ficlog,"#******");
1.227 brouard 8804: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
1.218 brouard 8805: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8806: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8807: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8808: }
8809: fprintf(ficresplb,"******\n");
8810: printf("******\n");
8811: fprintf(ficlog,"******\n");
1.227 brouard 8812: if(invalidvarcomb[k]){
8813: printf("\nCombination (%d) ignored because no cases \n",k);
8814: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
8815: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
8816: continue;
8817: }
1.218 brouard 8818:
8819: fprintf(ficresplb,"#Age ");
1.227 brouard 8820: for(j=1;j<=cptcoveff;j++) {
1.218 brouard 8821: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8822: }
8823: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
8824: fprintf(ficresplb,"Total Years_to_converge\n");
8825:
8826:
8827: for (age=agebase; age<=agelim; age++){
8828: /* for (age=agebase; age<=agebase; age++){ */
8829: if(mobilavproj > 0){
8830: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
8831: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.227 brouard 8832: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k);
1.218 brouard 8833: }else if (mobilavproj == 0){
1.227 brouard 8834: 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);
8835: 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);
8836: exit(1);
1.218 brouard 8837: }else{
1.227 brouard 8838: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
8839: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k);
1.218 brouard 8840: }
8841: fprintf(ficresplb,"%.0f ",age );
1.227 brouard 8842: for(j=1;j<=cptcoveff;j++)
8843: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.218 brouard 8844: tot=0.;
8845: for(i=1; i<=nlstate;i++){
1.227 brouard 8846: tot += bprlim[i][i];
8847: fprintf(ficresplb," %.5f", bprlim[i][i]);
1.218 brouard 8848: }
8849: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
8850: } /* Age */
8851: /* was end of cptcod */
8852: } /* cptcov */
8853:
8854: /* hBijx(p, bage, fage); */
8855: /* fclose(ficrespijb); */
8856:
8857: return 0;
1.217 brouard 8858: }
1.218 brouard 8859:
1.180 brouard 8860: int hPijx(double *p, int bage, int fage){
8861: /*------------- h Pij x at various ages ------------*/
8862:
8863: int stepsize;
8864: int agelim;
8865: int hstepm;
8866: int nhstepm;
8867: int h, i, i1, j, k;
8868:
8869: double agedeb;
8870: double ***p3mat;
8871:
1.201 brouard 8872: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 8873: if((ficrespij=fopen(filerespij,"w"))==NULL) {
8874: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
8875: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
8876: }
8877: printf("Computing pij: result on file '%s' \n", filerespij);
8878: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
8879:
8880: stepsize=(int) (stepm+YEARM-1)/YEARM;
8881: /*if (stepm<=24) stepsize=2;*/
8882:
8883: agelim=AGESUP;
8884: hstepm=stepsize*YEARM; /* Every year of age */
8885: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 8886:
1.180 brouard 8887: /* hstepm=1; aff par mois*/
8888: pstamp(ficrespij);
8889: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 8890: i1= pow(2,cptcoveff);
1.218 brouard 8891: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
8892: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
8893: /* k=k+1; */
1.227 brouard 8894: for (k=1; k <= (int) pow(2,cptcoveff); k++){
1.183 brouard 8895: fprintf(ficrespij,"\n#****** ");
1.227 brouard 8896: for(j=1;j<=cptcoveff;j++)
1.198 brouard 8897: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.183 brouard 8898: fprintf(ficrespij,"******\n");
8899:
8900: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
8901: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
8902: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
8903:
8904: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 8905:
1.183 brouard 8906: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
8907: oldm=oldms;savm=savms;
8908: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
8909: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
8910: for(i=1; i<=nlstate;i++)
8911: for(j=1; j<=nlstate+ndeath;j++)
8912: fprintf(ficrespij," %1d-%1d",i,j);
8913: fprintf(ficrespij,"\n");
8914: for (h=0; h<=nhstepm; h++){
8915: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
8916: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 8917: for(i=1; i<=nlstate;i++)
8918: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 8919: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 8920: fprintf(ficrespij,"\n");
8921: }
1.183 brouard 8922: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
8923: fprintf(ficrespij,"\n");
8924: }
1.180 brouard 8925: /*}*/
8926: }
1.218 brouard 8927: return 0;
1.180 brouard 8928: }
1.218 brouard 8929:
8930: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 8931: /*------------- h Bij x at various ages ------------*/
8932:
8933: int stepsize;
1.218 brouard 8934: /* int agelim; */
8935: int ageminl;
1.217 brouard 8936: int hstepm;
8937: int nhstepm;
8938: int h, i, i1, j, k;
1.218 brouard 8939:
1.217 brouard 8940: double agedeb;
8941: double ***p3mat;
1.218 brouard 8942:
8943: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
8944: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
8945: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
8946: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
8947: }
8948: printf("Computing pij back: result on file '%s' \n", filerespijb);
8949: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
8950:
8951: stepsize=(int) (stepm+YEARM-1)/YEARM;
8952: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 8953:
1.218 brouard 8954: /* agelim=AGESUP; */
8955: ageminl=30;
8956: hstepm=stepsize*YEARM; /* Every year of age */
8957: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
8958:
8959: /* hstepm=1; aff par mois*/
8960: pstamp(ficrespijb);
8961: fprintf(ficrespijb,"#****** h Pij x Back Probability to be in state i at age x-h being in j at x ");
1.227 brouard 8962: i1= pow(2,cptcoveff);
1.218 brouard 8963: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
8964: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
8965: /* k=k+1; */
1.227 brouard 8966: for (k=1; k <= (int) pow(2,cptcoveff); k++){
1.218 brouard 8967: fprintf(ficrespijb,"\n#****** ");
1.227 brouard 8968: for(j=1;j<=cptcoveff;j++)
1.218 brouard 8969: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
8970: fprintf(ficrespijb,"******\n");
1.222 brouard 8971: if(invalidvarcomb[k]){
8972: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
8973: continue;
8974: }
1.218 brouard 8975:
8976: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
8977: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
8978: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
8979: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
8980: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
8981:
8982: /* nhstepm=nhstepm*YEARM; aff par mois*/
8983:
8984: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
8985: /* oldm=oldms;savm=savms; */
8986: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
8987: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
8988: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
8989: fprintf(ficrespijb,"# Cov Agex agex-h hpijx with i,j=");
8990: for(i=1; i<=nlstate;i++)
8991: for(j=1; j<=nlstate+ndeath;j++)
8992: fprintf(ficrespijb," %1d-%1d",i,j);
8993: fprintf(ficrespijb,"\n");
8994: for (h=0; h<=nhstepm; h++){
8995: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
8996: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
8997: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
1.217 brouard 8998: for(i=1; i<=nlstate;i++)
8999: for(j=1; j<=nlstate+ndeath;j++)
1.218 brouard 9000: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
1.217 brouard 9001: fprintf(ficrespijb,"\n");
9002: }
1.218 brouard 9003: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9004: fprintf(ficrespijb,"\n");
1.217 brouard 9005: }
1.218 brouard 9006: /*}*/
9007: }
9008: return 0;
9009: } /* hBijx */
1.217 brouard 9010:
1.180 brouard 9011:
1.136 brouard 9012: /***********************************************/
9013: /**************** Main Program *****************/
9014: /***********************************************/
9015:
9016: int main(int argc, char *argv[])
9017: {
9018: #ifdef GSL
9019: const gsl_multimin_fminimizer_type *T;
9020: size_t iteri = 0, it;
9021: int rval = GSL_CONTINUE;
9022: int status = GSL_SUCCESS;
9023: double ssval;
9024: #endif
9025: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9026: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9027: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9028: int jj, ll, li, lj, lk;
1.136 brouard 9029: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9030: int num_filled;
1.136 brouard 9031: int itimes;
9032: int NDIM=2;
9033: int vpopbased=0;
9034:
1.164 brouard 9035: char ca[32], cb[32];
1.136 brouard 9036: /* FILE *fichtm; *//* Html File */
9037: /* FILE *ficgp;*/ /*Gnuplot File */
9038: struct stat info;
1.191 brouard 9039: double agedeb=0.;
1.194 brouard 9040:
9041: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9042: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9043:
1.165 brouard 9044: double fret;
1.191 brouard 9045: double dum=0.; /* Dummy variable */
1.136 brouard 9046: double ***p3mat;
1.218 brouard 9047: /* double ***mobaverage; */
1.164 brouard 9048:
9049: char line[MAXLINE];
1.197 brouard 9050: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9051:
9052: char model[MAXLINE], modeltemp[MAXLINE];
1.230 brouard 9053: char resultline[MAXLINE];
9054:
1.136 brouard 9055: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9056: char *tok, *val; /* pathtot */
1.136 brouard 9057: int firstobs=1, lastobs=10;
1.195 brouard 9058: int c, h , cpt, c2;
1.191 brouard 9059: int jl=0;
9060: int i1, j1, jk, stepsize=0;
1.194 brouard 9061: int count=0;
9062:
1.164 brouard 9063: int *tab;
1.136 brouard 9064: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9065: int backcast=0;
1.136 brouard 9066: int mobilav=0,popforecast=0;
1.191 brouard 9067: int hstepm=0, nhstepm=0;
1.136 brouard 9068: int agemortsup;
9069: float sumlpop=0.;
9070: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9071: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9072:
1.191 brouard 9073: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9074: double ftolpl=FTOL;
9075: double **prlim;
1.217 brouard 9076: double **bprlim;
1.136 brouard 9077: double ***param; /* Matrix of parameters */
9078: double *p;
9079: double **matcov; /* Matrix of covariance */
1.203 brouard 9080: double **hess; /* Hessian matrix */
1.136 brouard 9081: double ***delti3; /* Scale */
9082: double *delti; /* Scale */
9083: double ***eij, ***vareij;
9084: double **varpl; /* Variances of prevalence limits by age */
9085: double *epj, vepp;
1.164 brouard 9086:
1.136 brouard 9087: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9088: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9089:
1.136 brouard 9090: double **ximort;
1.145 brouard 9091: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9092: int *dcwave;
9093:
1.164 brouard 9094: char z[1]="c";
1.136 brouard 9095:
9096: /*char *strt;*/
9097: char strtend[80];
1.126 brouard 9098:
1.164 brouard 9099:
1.126 brouard 9100: /* setlocale (LC_ALL, ""); */
9101: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9102: /* textdomain (PACKAGE); */
9103: /* setlocale (LC_CTYPE, ""); */
9104: /* setlocale (LC_MESSAGES, ""); */
9105:
9106: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9107: rstart_time = time(NULL);
9108: /* (void) gettimeofday(&start_time,&tzp);*/
9109: start_time = *localtime(&rstart_time);
1.126 brouard 9110: curr_time=start_time;
1.157 brouard 9111: /*tml = *localtime(&start_time.tm_sec);*/
9112: /* strcpy(strstart,asctime(&tml)); */
9113: strcpy(strstart,asctime(&start_time));
1.126 brouard 9114:
9115: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9116: /* tp.tm_sec = tp.tm_sec +86400; */
9117: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9118: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9119: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9120: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9121: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9122: /* strt=asctime(&tmg); */
9123: /* printf("Time(after) =%s",strstart); */
9124: /* (void) time (&time_value);
9125: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9126: * tm = *localtime(&time_value);
9127: * strstart=asctime(&tm);
9128: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9129: */
9130:
9131: nberr=0; /* Number of errors and warnings */
9132: nbwarn=0;
1.184 brouard 9133: #ifdef WIN32
9134: _getcwd(pathcd, size);
9135: #else
1.126 brouard 9136: getcwd(pathcd, size);
1.184 brouard 9137: #endif
1.191 brouard 9138: syscompilerinfo(0);
1.196 brouard 9139: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9140: if(argc <=1){
9141: printf("\nEnter the parameter file name: ");
1.205 brouard 9142: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9143: printf("ERROR Empty parameter file name\n");
9144: goto end;
9145: }
1.126 brouard 9146: i=strlen(pathr);
9147: if(pathr[i-1]=='\n')
9148: pathr[i-1]='\0';
1.156 brouard 9149: i=strlen(pathr);
1.205 brouard 9150: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9151: pathr[i-1]='\0';
1.205 brouard 9152: }
9153: i=strlen(pathr);
9154: if( i==0 ){
9155: printf("ERROR Empty parameter file name\n");
9156: goto end;
9157: }
9158: for (tok = pathr; tok != NULL; ){
1.126 brouard 9159: printf("Pathr |%s|\n",pathr);
9160: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9161: printf("val= |%s| pathr=%s\n",val,pathr);
9162: strcpy (pathtot, val);
9163: if(pathr[0] == '\0') break; /* Dirty */
9164: }
9165: }
9166: else{
9167: strcpy(pathtot,argv[1]);
9168: }
9169: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9170: /*cygwin_split_path(pathtot,path,optionfile);
9171: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9172: /* cutv(path,optionfile,pathtot,'\\');*/
9173:
9174: /* Split argv[0], imach program to get pathimach */
9175: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9176: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9177: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9178: /* strcpy(pathimach,argv[0]); */
9179: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9180: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9181: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9182: #ifdef WIN32
9183: _chdir(path); /* Can be a relative path */
9184: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9185: #else
1.126 brouard 9186: chdir(path); /* Can be a relative path */
1.184 brouard 9187: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9188: #endif
9189: printf("Current directory %s!\n",pathcd);
1.126 brouard 9190: strcpy(command,"mkdir ");
9191: strcat(command,optionfilefiname);
9192: if((outcmd=system(command)) != 0){
1.169 brouard 9193: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9194: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9195: /* fclose(ficlog); */
9196: /* exit(1); */
9197: }
9198: /* if((imk=mkdir(optionfilefiname))<0){ */
9199: /* perror("mkdir"); */
9200: /* } */
9201:
9202: /*-------- arguments in the command line --------*/
9203:
1.186 brouard 9204: /* Main Log file */
1.126 brouard 9205: strcat(filelog, optionfilefiname);
9206: strcat(filelog,".log"); /* */
9207: if((ficlog=fopen(filelog,"w"))==NULL) {
9208: printf("Problem with logfile %s\n",filelog);
9209: goto end;
9210: }
9211: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9212: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9213: fprintf(ficlog,"\nEnter the parameter file name: \n");
9214: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9215: path=%s \n\
9216: optionfile=%s\n\
9217: optionfilext=%s\n\
1.156 brouard 9218: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9219:
1.197 brouard 9220: syscompilerinfo(1);
1.167 brouard 9221:
1.126 brouard 9222: printf("Local time (at start):%s",strstart);
9223: fprintf(ficlog,"Local time (at start): %s",strstart);
9224: fflush(ficlog);
9225: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9226: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9227:
9228: /* */
9229: strcpy(fileres,"r");
9230: strcat(fileres, optionfilefiname);
1.201 brouard 9231: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9232: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9233: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 9234:
1.186 brouard 9235: /* Main ---------arguments file --------*/
1.126 brouard 9236:
9237: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 9238: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
9239: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 9240: fflush(ficlog);
1.149 brouard 9241: /* goto end; */
9242: exit(70);
1.126 brouard 9243: }
9244:
9245:
9246:
9247: strcpy(filereso,"o");
1.201 brouard 9248: strcat(filereso,fileresu);
1.126 brouard 9249: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
9250: printf("Problem with Output resultfile: %s\n", filereso);
9251: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
9252: fflush(ficlog);
9253: goto end;
9254: }
9255:
9256: /* Reads comments: lines beginning with '#' */
9257: numlinepar=0;
1.197 brouard 9258:
9259: /* First parameter line */
9260: while(fgets(line, MAXLINE, ficpar)) {
9261: /* If line starts with a # it is a comment */
9262: if (line[0] == '#') {
9263: numlinepar++;
9264: fputs(line,stdout);
9265: fputs(line,ficparo);
9266: fputs(line,ficlog);
9267: continue;
9268: }else
9269: break;
9270: }
9271: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
9272: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
9273: if (num_filled != 5) {
9274: printf("Should be 5 parameters\n");
9275: }
1.126 brouard 9276: numlinepar++;
1.197 brouard 9277: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
9278: }
9279: /* Second parameter line */
9280: while(fgets(line, MAXLINE, ficpar)) {
9281: /* If line starts with a # it is a comment */
9282: if (line[0] == '#') {
9283: numlinepar++;
9284: fputs(line,stdout);
9285: fputs(line,ficparo);
9286: fputs(line,ficlog);
9287: continue;
9288: }else
9289: break;
9290: }
1.223 brouard 9291: 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", \
9292: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
9293: if (num_filled != 11) {
9294: 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 9295: printf("but line=%s\n",line);
1.197 brouard 9296: }
1.223 brouard 9297: 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 9298: }
1.203 brouard 9299: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 9300: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 9301: /* Third parameter line */
9302: while(fgets(line, MAXLINE, ficpar)) {
9303: /* If line starts with a # it is a comment */
9304: if (line[0] == '#') {
9305: numlinepar++;
9306: fputs(line,stdout);
9307: fputs(line,ficparo);
9308: fputs(line,ficlog);
9309: continue;
9310: }else
9311: break;
9312: }
1.201 brouard 9313: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
9314: if (num_filled == 0)
9315: model[0]='\0';
9316: else if (num_filled != 1){
1.197 brouard 9317: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9318: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9319: model[0]='\0';
9320: goto end;
9321: }
9322: else{
9323: if (model[0]=='+'){
9324: for(i=1; i<=strlen(model);i++)
9325: modeltemp[i-1]=model[i];
1.201 brouard 9326: strcpy(model,modeltemp);
1.197 brouard 9327: }
9328: }
1.199 brouard 9329: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 9330: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 9331: }
9332: /* 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); */
9333: /* numlinepar=numlinepar+3; /\* In general *\/ */
9334: /* 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 9335: 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);
9336: 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 9337: fflush(ficlog);
1.190 brouard 9338: /* if(model[0]=='#'|| model[0]== '\0'){ */
9339: if(model[0]=='#'){
1.187 brouard 9340: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
9341: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
9342: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
9343: if(mle != -1){
9344: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
9345: exit(1);
9346: }
9347: }
1.126 brouard 9348: while((c=getc(ficpar))=='#' && c!= EOF){
9349: ungetc(c,ficpar);
9350: fgets(line, MAXLINE, ficpar);
9351: numlinepar++;
1.195 brouard 9352: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
9353: z[0]=line[1];
9354: }
9355: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 9356: fputs(line, stdout);
9357: //puts(line);
1.126 brouard 9358: fputs(line,ficparo);
9359: fputs(line,ficlog);
9360: }
9361: ungetc(c,ficpar);
9362:
9363:
1.145 brouard 9364: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 9365: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 ! brouard 9366: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 9367: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 9368: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
9369: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
9370: v1+v2*age+v2*v3 makes cptcovn = 3
9371: */
9372: if (strlen(model)>1)
1.187 brouard 9373: 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 9374: else
1.187 brouard 9375: ncovmodel=2; /* Constant and age */
1.133 brouard 9376: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
9377: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 9378: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
9379: 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);
9380: 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);
9381: fflush(stdout);
9382: fclose (ficlog);
9383: goto end;
9384: }
1.126 brouard 9385: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9386: delti=delti3[1][1];
9387: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
9388: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
9389: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 9390: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
9391: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9392: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
9393: fclose (ficparo);
9394: fclose (ficlog);
9395: goto end;
9396: exit(0);
1.220 brouard 9397: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 9398: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 9399: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
9400: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9401: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9402: matcov=matrix(1,npar,1,npar);
1.203 brouard 9403: hess=matrix(1,npar,1,npar);
1.220 brouard 9404: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 9405: /* Read guessed parameters */
1.126 brouard 9406: /* Reads comments: lines beginning with '#' */
9407: while((c=getc(ficpar))=='#' && c!= EOF){
9408: ungetc(c,ficpar);
9409: fgets(line, MAXLINE, ficpar);
9410: numlinepar++;
1.141 brouard 9411: fputs(line,stdout);
1.126 brouard 9412: fputs(line,ficparo);
9413: fputs(line,ficlog);
9414: }
9415: ungetc(c,ficpar);
9416:
9417: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9418: for(i=1; i <=nlstate; i++){
1.220 brouard 9419: j=0;
1.126 brouard 9420: for(jj=1; jj <=nlstate+ndeath; jj++){
1.220 brouard 9421: if(jj==i) continue;
9422: j++;
9423: fscanf(ficpar,"%1d%1d",&i1,&j1);
9424: if ((i1 != i) || (j1 != jj)){
9425: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 9426: It might be a problem of design; if ncovcol and the model are correct\n \
9427: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.220 brouard 9428: exit(1);
9429: }
9430: fprintf(ficparo,"%1d%1d",i1,j1);
9431: if(mle==1)
9432: printf("%1d%1d",i,jj);
9433: fprintf(ficlog,"%1d%1d",i,jj);
9434: for(k=1; k<=ncovmodel;k++){
9435: fscanf(ficpar," %lf",¶m[i][j][k]);
9436: if(mle==1){
9437: printf(" %lf",param[i][j][k]);
9438: fprintf(ficlog," %lf",param[i][j][k]);
9439: }
9440: else
9441: fprintf(ficlog," %lf",param[i][j][k]);
9442: fprintf(ficparo," %lf",param[i][j][k]);
9443: }
9444: fscanf(ficpar,"\n");
9445: numlinepar++;
9446: if(mle==1)
9447: printf("\n");
9448: fprintf(ficlog,"\n");
9449: fprintf(ficparo,"\n");
1.126 brouard 9450: }
9451: }
9452: fflush(ficlog);
9453:
1.145 brouard 9454: /* Reads scales values */
1.126 brouard 9455: p=param[1][1];
9456:
9457: /* Reads comments: lines beginning with '#' */
9458: while((c=getc(ficpar))=='#' && c!= EOF){
9459: ungetc(c,ficpar);
9460: fgets(line, MAXLINE, ficpar);
9461: numlinepar++;
1.141 brouard 9462: fputs(line,stdout);
1.126 brouard 9463: fputs(line,ficparo);
9464: fputs(line,ficlog);
9465: }
9466: ungetc(c,ficpar);
9467:
9468: for(i=1; i <=nlstate; i++){
9469: for(j=1; j <=nlstate+ndeath-1; j++){
1.220 brouard 9470: fscanf(ficpar,"%1d%1d",&i1,&j1);
9471: if ( (i1-i) * (j1-j) != 0){
9472: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
9473: exit(1);
9474: }
9475: printf("%1d%1d",i,j);
9476: fprintf(ficparo,"%1d%1d",i1,j1);
9477: fprintf(ficlog,"%1d%1d",i1,j1);
9478: for(k=1; k<=ncovmodel;k++){
9479: fscanf(ficpar,"%le",&delti3[i][j][k]);
9480: printf(" %le",delti3[i][j][k]);
9481: fprintf(ficparo," %le",delti3[i][j][k]);
9482: fprintf(ficlog," %le",delti3[i][j][k]);
9483: }
9484: fscanf(ficpar,"\n");
9485: numlinepar++;
9486: printf("\n");
9487: fprintf(ficparo,"\n");
9488: fprintf(ficlog,"\n");
1.126 brouard 9489: }
9490: }
9491: fflush(ficlog);
1.220 brouard 9492:
1.145 brouard 9493: /* Reads covariance matrix */
1.126 brouard 9494: delti=delti3[1][1];
1.220 brouard 9495:
9496:
1.126 brouard 9497: /* 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 9498:
1.126 brouard 9499: /* Reads comments: lines beginning with '#' */
9500: while((c=getc(ficpar))=='#' && c!= EOF){
9501: ungetc(c,ficpar);
9502: fgets(line, MAXLINE, ficpar);
9503: numlinepar++;
1.141 brouard 9504: fputs(line,stdout);
1.126 brouard 9505: fputs(line,ficparo);
9506: fputs(line,ficlog);
9507: }
9508: ungetc(c,ficpar);
1.220 brouard 9509:
1.126 brouard 9510: matcov=matrix(1,npar,1,npar);
1.203 brouard 9511: hess=matrix(1,npar,1,npar);
1.131 brouard 9512: for(i=1; i <=npar; i++)
9513: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 9514:
1.194 brouard 9515: /* Scans npar lines */
1.126 brouard 9516: for(i=1; i <=npar; i++){
1.226 brouard 9517: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 9518: if(count != 3){
1.226 brouard 9519: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 9520: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
9521: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 9522: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 9523: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
9524: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 9525: exit(1);
1.220 brouard 9526: }else{
1.226 brouard 9527: if(mle==1)
9528: printf("%1d%1d%d",i1,j1,jk);
9529: }
9530: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
9531: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 9532: for(j=1; j <=i; j++){
1.226 brouard 9533: fscanf(ficpar," %le",&matcov[i][j]);
9534: if(mle==1){
9535: printf(" %.5le",matcov[i][j]);
9536: }
9537: fprintf(ficlog," %.5le",matcov[i][j]);
9538: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 9539: }
9540: fscanf(ficpar,"\n");
9541: numlinepar++;
9542: if(mle==1)
1.220 brouard 9543: printf("\n");
1.126 brouard 9544: fprintf(ficlog,"\n");
9545: fprintf(ficparo,"\n");
9546: }
1.194 brouard 9547: /* End of read covariance matrix npar lines */
1.126 brouard 9548: for(i=1; i <=npar; i++)
9549: for(j=i+1;j<=npar;j++)
1.226 brouard 9550: matcov[i][j]=matcov[j][i];
1.126 brouard 9551:
9552: if(mle==1)
9553: printf("\n");
9554: fprintf(ficlog,"\n");
9555:
9556: fflush(ficlog);
9557:
9558: /*-------- Rewriting parameter file ----------*/
9559: strcpy(rfileres,"r"); /* "Rparameterfile */
9560: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
9561: strcat(rfileres,"."); /* */
9562: strcat(rfileres,optionfilext); /* Other files have txt extension */
9563: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 9564: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
9565: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 9566: }
9567: fprintf(ficres,"#%s\n",version);
9568: } /* End of mle != -3 */
1.218 brouard 9569:
1.186 brouard 9570: /* Main data
9571: */
1.126 brouard 9572: n= lastobs;
9573: num=lvector(1,n);
9574: moisnais=vector(1,n);
9575: annais=vector(1,n);
9576: moisdc=vector(1,n);
9577: andc=vector(1,n);
1.220 brouard 9578: weight=vector(1,n);
1.126 brouard 9579: agedc=vector(1,n);
9580: cod=ivector(1,n);
1.220 brouard 9581: for(i=1;i<=n;i++){
9582: num[i]=0;
9583: moisnais[i]=0;
9584: annais[i]=0;
9585: moisdc[i]=0;
9586: andc[i]=0;
9587: agedc[i]=0;
9588: cod[i]=0;
9589: weight[i]=1.0; /* Equal weights, 1 by default */
9590: }
1.126 brouard 9591: mint=matrix(1,maxwav,1,n);
9592: anint=matrix(1,maxwav,1,n);
1.131 brouard 9593: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 9594: tab=ivector(1,NCOVMAX);
1.144 brouard 9595: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 9596: 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 9597:
1.136 brouard 9598: /* Reads data from file datafile */
9599: if (readdata(datafile, firstobs, lastobs, &imx)==1)
9600: goto end;
9601:
9602: /* Calculation of the number of parameters from char model */
1.137 brouard 9603: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
9604: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
9605: k=3 V4 Tvar[k=3]= 4 (from V4)
9606: k=2 V1 Tvar[k=2]= 1 (from V1)
9607: k=1 Tvar[1]=2 (from V2)
9608: */
1.231 brouard 9609:
9610: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
1.232 brouard 9611: TvarF=ivector(1,NCOVMAX); /* */
9612: TvarFind=ivector(1,NCOVMAX); /* */
9613: TvarV=ivector(1,NCOVMAX); /* */
9614: TvarVind=ivector(1,NCOVMAX); /* */
9615: TvarA=ivector(1,NCOVMAX); /* */
9616: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 9617: TvarFD=ivector(1,NCOVMAX); /* */
9618: TvarFDind=ivector(1,NCOVMAX); /* */
9619: TvarFQ=ivector(1,NCOVMAX); /* */
9620: TvarFQind=ivector(1,NCOVMAX); /* */
9621: TvarVD=ivector(1,NCOVMAX); /* */
9622: TvarVDind=ivector(1,NCOVMAX); /* */
9623: TvarVQ=ivector(1,NCOVMAX); /* */
9624: TvarVQind=ivector(1,NCOVMAX); /* */
9625:
1.230 brouard 9626: Tvalsel=vector(1,NCOVMAX); /* */
1.233 ! brouard 9627: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 9628: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
9629: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
9630: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 9631: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
9632: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
9633: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
9634: */
9635: /* For model-covariate k tells which data-covariate to use but
9636: because this model-covariate is a construction we invent a new column
9637: ncovcol + k1
9638: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
9639: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 9640: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
9641: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 9642: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
9643: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 9644: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 9645: */
1.145 brouard 9646: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
9647: 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 9648: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
9649: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 9650: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 9651: 4 covariates (3 plus signs)
9652: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
9653: */
1.230 brouard 9654: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 9655: * individual dummy, fixed or varying:
9656: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
9657: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 9658: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
9659: * V1 df, V2 qf, V3 & V4 dv, V5 qv
9660: * Tmodelind[1]@9={9,0,3,2,}*/
9661: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
9662: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 9663: * individual quantitative, fixed or varying:
9664: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
9665: * 3, 1, 0, 0, 0, 0, 0, 0},
9666: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 9667: /* Main decodemodel */
9668:
1.187 brouard 9669:
1.223 brouard 9670: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 9671: goto end;
9672:
1.137 brouard 9673: if((double)(lastobs-imx)/(double)imx > 1.10){
9674: nbwarn++;
9675: 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);
9676: 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);
9677: }
1.136 brouard 9678: /* if(mle==1){*/
1.137 brouard 9679: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
9680: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 9681: }
9682:
9683: /*-calculation of age at interview from date of interview and age at death -*/
9684: agev=matrix(1,maxwav,1,imx);
9685:
9686: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
9687: goto end;
9688:
1.126 brouard 9689:
1.136 brouard 9690: agegomp=(int)agemin;
9691: free_vector(moisnais,1,n);
9692: free_vector(annais,1,n);
1.126 brouard 9693: /* free_matrix(mint,1,maxwav,1,n);
9694: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 9695: /* free_vector(moisdc,1,n); */
9696: /* free_vector(andc,1,n); */
1.145 brouard 9697: /* */
9698:
1.126 brouard 9699: wav=ivector(1,imx);
1.214 brouard 9700: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
9701: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
9702: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
9703: 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.*/
9704: bh=imatrix(1,lastpass-firstpass+2,1,imx);
9705: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 9706:
9707: /* Concatenates waves */
1.214 brouard 9708: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
9709: Death is a valid wave (if date is known).
9710: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
9711: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
9712: and mw[mi+1][i]. dh depends on stepm.
9713: */
9714:
1.126 brouard 9715: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.145 brouard 9716: /* */
9717:
1.215 brouard 9718: free_vector(moisdc,1,n);
9719: free_vector(andc,1,n);
9720:
1.126 brouard 9721: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
9722: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
9723: ncodemax[1]=1;
1.145 brouard 9724: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 9725: cptcoveff=0;
1.220 brouard 9726: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
9727: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 9728: }
9729:
9730: ncovcombmax=pow(2,cptcoveff);
9731: invalidvarcomb=ivector(1, ncovcombmax);
9732: for(i=1;i<ncovcombmax;i++)
9733: invalidvarcomb[i]=0;
9734:
1.211 brouard 9735: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 9736: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 9737: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 9738:
1.200 brouard 9739: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 9740: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 9741: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 9742: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
9743: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
9744: * (currently 0 or 1) in the data.
9745: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
9746: * corresponding modality (h,j).
9747: */
9748:
1.145 brouard 9749: h=0;
9750: /*if (cptcovn > 0) */
1.126 brouard 9751: m=pow(2,cptcoveff);
9752:
1.144 brouard 9753: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 9754: * For k=4 covariates, h goes from 1 to m=2**k
9755: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
9756: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 9757: * h\k 1 2 3 4
1.143 brouard 9758: *______________________________
9759: * 1 i=1 1 i=1 1 i=1 1 i=1 1
9760: * 2 2 1 1 1
9761: * 3 i=2 1 2 1 1
9762: * 4 2 2 1 1
9763: * 5 i=3 1 i=2 1 2 1
9764: * 6 2 1 2 1
9765: * 7 i=4 1 2 2 1
9766: * 8 2 2 2 1
1.197 brouard 9767: * 9 i=5 1 i=3 1 i=2 1 2
9768: * 10 2 1 1 2
9769: * 11 i=6 1 2 1 2
9770: * 12 2 2 1 2
9771: * 13 i=7 1 i=4 1 2 2
9772: * 14 2 1 2 2
9773: * 15 i=8 1 2 2 2
9774: * 16 2 2 2 2
1.143 brouard 9775: */
1.212 brouard 9776: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 9777: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
9778: * and the value of each covariate?
9779: * V1=1, V2=1, V3=2, V4=1 ?
9780: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
9781: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
9782: * In order to get the real value in the data, we use nbcode
9783: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
9784: * We are keeping this crazy system in order to be able (in the future?)
9785: * to have more than 2 values (0 or 1) for a covariate.
9786: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
9787: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
9788: * bbbbbbbb
9789: * 76543210
9790: * h-1 00000101 (6-1=5)
1.219 brouard 9791: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 9792: * &
9793: * 1 00000001 (1)
1.219 brouard 9794: * 00000000 = 1 & ((h-1) >> (k-1))
9795: * +1= 00000001 =1
1.211 brouard 9796: *
9797: * h=14, k=3 => h'=h-1=13, k'=k-1=2
9798: * h' 1101 =2^3+2^2+0x2^1+2^0
9799: * >>k' 11
9800: * & 00000001
9801: * = 00000001
9802: * +1 = 00000010=2 = codtabm(14,3)
9803: * Reverse h=6 and m=16?
9804: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
9805: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
9806: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
9807: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
9808: * V3=decodtabm(14,3,2**4)=2
9809: * h'=13 1101 =2^3+2^2+0x2^1+2^0
9810: *(h-1) >> (j-1) 0011 =13 >> 2
9811: * &1 000000001
9812: * = 000000001
9813: * +1= 000000010 =2
9814: * 2211
9815: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
9816: * V3=2
1.220 brouard 9817: * codtabm and decodtabm are identical
1.211 brouard 9818: */
9819:
1.145 brouard 9820:
9821: free_ivector(Ndum,-1,NCOVMAX);
9822:
9823:
1.126 brouard 9824:
1.186 brouard 9825: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 9826: strcpy(optionfilegnuplot,optionfilefiname);
9827: if(mle==-3)
1.201 brouard 9828: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 9829: strcat(optionfilegnuplot,".gp");
9830:
9831: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
9832: printf("Problem with file %s",optionfilegnuplot);
9833: }
9834: else{
1.204 brouard 9835: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 9836: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 9837: //fprintf(ficgp,"set missing 'NaNq'\n");
9838: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 9839: }
9840: /* fclose(ficgp);*/
1.186 brouard 9841:
9842:
9843: /* Initialisation of --------- index.htm --------*/
1.126 brouard 9844:
9845: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
9846: if(mle==-3)
1.201 brouard 9847: strcat(optionfilehtm,"-MORT_");
1.126 brouard 9848: strcat(optionfilehtm,".htm");
9849: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 9850: printf("Problem with %s \n",optionfilehtm);
9851: exit(0);
1.126 brouard 9852: }
9853:
9854: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
9855: strcat(optionfilehtmcov,"-cov.htm");
9856: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
9857: printf("Problem with %s \n",optionfilehtmcov), exit(0);
9858: }
9859: else{
9860: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
9861: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 9862: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 9863: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
9864: }
9865:
1.213 brouard 9866: 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 9867: <hr size=\"2\" color=\"#EC5E5E\"> \n\
9868: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 9869: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 9870: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 9871: \n\
9872: <hr size=\"2\" color=\"#EC5E5E\">\
9873: <ul><li><h4>Parameter files</h4>\n\
9874: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
9875: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
9876: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
9877: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
9878: - Date and time at start: %s</ul>\n",\
9879: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
9880: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
9881: fileres,fileres,\
9882: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
9883: fflush(fichtm);
9884:
9885: strcpy(pathr,path);
9886: strcat(pathr,optionfilefiname);
1.184 brouard 9887: #ifdef WIN32
9888: _chdir(optionfilefiname); /* Move to directory named optionfile */
9889: #else
1.126 brouard 9890: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 9891: #endif
9892:
1.126 brouard 9893:
1.220 brouard 9894: /* Calculates basic frequencies. Computes observed prevalence at single age
9895: and for any valid combination of covariates
1.126 brouard 9896: and prints on file fileres'p'. */
1.227 brouard 9897: freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
9898: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 9899:
9900: fprintf(fichtm,"\n");
9901: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
9902: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
9903: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
9904: imx,agemin,agemax,jmin,jmax,jmean);
9905: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 9906: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
9907: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
9908: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
9909: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 9910:
1.126 brouard 9911: /* For Powell, parameters are in a vector p[] starting at p[1]
9912: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
9913: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
9914:
9915: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 9916: /* For mortality only */
1.126 brouard 9917: if (mle==-3){
1.136 brouard 9918: ximort=matrix(1,NDIM,1,NDIM);
1.220 brouard 9919: for(i=1;i<=NDIM;i++)
9920: for(j=1;j<=NDIM;j++)
9921: ximort[i][j]=0.;
1.186 brouard 9922: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 9923: cens=ivector(1,n);
9924: ageexmed=vector(1,n);
9925: agecens=vector(1,n);
9926: dcwave=ivector(1,n);
1.223 brouard 9927:
1.126 brouard 9928: for (i=1; i<=imx; i++){
9929: dcwave[i]=-1;
9930: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 9931: if (s[m][i]>nlstate) {
9932: dcwave[i]=m;
9933: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
9934: break;
9935: }
1.126 brouard 9936: }
1.226 brouard 9937:
1.126 brouard 9938: for (i=1; i<=imx; i++) {
9939: if (wav[i]>0){
1.226 brouard 9940: ageexmed[i]=agev[mw[1][i]][i];
9941: j=wav[i];
9942: agecens[i]=1.;
9943:
9944: if (ageexmed[i]> 1 && wav[i] > 0){
9945: agecens[i]=agev[mw[j][i]][i];
9946: cens[i]= 1;
9947: }else if (ageexmed[i]< 1)
9948: cens[i]= -1;
9949: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
9950: cens[i]=0 ;
1.126 brouard 9951: }
9952: else cens[i]=-1;
9953: }
9954:
9955: for (i=1;i<=NDIM;i++) {
9956: for (j=1;j<=NDIM;j++)
1.226 brouard 9957: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 9958: }
9959:
1.145 brouard 9960: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 9961: /*printf("%lf %lf", p[1], p[2]);*/
9962:
9963:
1.136 brouard 9964: #ifdef GSL
9965: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 9966: #else
1.126 brouard 9967: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 9968: #endif
1.201 brouard 9969: strcpy(filerespow,"POW-MORT_");
9970: strcat(filerespow,fileresu);
1.126 brouard 9971: if((ficrespow=fopen(filerespow,"w"))==NULL) {
9972: printf("Problem with resultfile: %s\n", filerespow);
9973: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
9974: }
1.136 brouard 9975: #ifdef GSL
9976: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 9977: #else
1.126 brouard 9978: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 9979: #endif
1.126 brouard 9980: /* for (i=1;i<=nlstate;i++)
9981: for(j=1;j<=nlstate+ndeath;j++)
9982: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
9983: */
9984: fprintf(ficrespow,"\n");
1.136 brouard 9985: #ifdef GSL
9986: /* gsl starts here */
9987: T = gsl_multimin_fminimizer_nmsimplex;
9988: gsl_multimin_fminimizer *sfm = NULL;
9989: gsl_vector *ss, *x;
9990: gsl_multimin_function minex_func;
9991:
9992: /* Initial vertex size vector */
9993: ss = gsl_vector_alloc (NDIM);
9994:
9995: if (ss == NULL){
9996: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
9997: }
9998: /* Set all step sizes to 1 */
9999: gsl_vector_set_all (ss, 0.001);
10000:
10001: /* Starting point */
1.126 brouard 10002:
1.136 brouard 10003: x = gsl_vector_alloc (NDIM);
10004:
10005: if (x == NULL){
10006: gsl_vector_free(ss);
10007: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10008: }
10009:
10010: /* Initialize method and iterate */
10011: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10012: /* gsl_vector_set(x, 0, 0.0268); */
10013: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10014: gsl_vector_set(x, 0, p[1]);
10015: gsl_vector_set(x, 1, p[2]);
10016:
10017: minex_func.f = &gompertz_f;
10018: minex_func.n = NDIM;
10019: minex_func.params = (void *)&p; /* ??? */
10020:
10021: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10022: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10023:
10024: printf("Iterations beginning .....\n\n");
10025: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10026:
10027: iteri=0;
10028: while (rval == GSL_CONTINUE){
10029: iteri++;
10030: status = gsl_multimin_fminimizer_iterate(sfm);
10031:
10032: if (status) printf("error: %s\n", gsl_strerror (status));
10033: fflush(0);
10034:
10035: if (status)
10036: break;
10037:
10038: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10039: ssval = gsl_multimin_fminimizer_size (sfm);
10040:
10041: if (rval == GSL_SUCCESS)
10042: printf ("converged to a local maximum at\n");
10043:
10044: printf("%5d ", iteri);
10045: for (it = 0; it < NDIM; it++){
10046: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10047: }
10048: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10049: }
10050:
10051: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10052:
10053: gsl_vector_free(x); /* initial values */
10054: gsl_vector_free(ss); /* inital step size */
10055: for (it=0; it<NDIM; it++){
10056: p[it+1]=gsl_vector_get(sfm->x,it);
10057: fprintf(ficrespow," %.12lf", p[it]);
10058: }
10059: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10060: #endif
10061: #ifdef POWELL
10062: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10063: #endif
1.126 brouard 10064: fclose(ficrespow);
10065:
1.203 brouard 10066: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10067:
10068: for(i=1; i <=NDIM; i++)
10069: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10070: matcov[i][j]=matcov[j][i];
1.126 brouard 10071:
10072: printf("\nCovariance matrix\n ");
1.203 brouard 10073: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10074: for(i=1; i <=NDIM; i++) {
10075: for(j=1;j<=NDIM;j++){
1.220 brouard 10076: printf("%f ",matcov[i][j]);
10077: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10078: }
1.203 brouard 10079: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10080: }
10081:
10082: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10083: for (i=1;i<=NDIM;i++) {
1.126 brouard 10084: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10085: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10086: }
1.126 brouard 10087: lsurv=vector(1,AGESUP);
10088: lpop=vector(1,AGESUP);
10089: tpop=vector(1,AGESUP);
10090: lsurv[agegomp]=100000;
10091:
10092: for (k=agegomp;k<=AGESUP;k++) {
10093: agemortsup=k;
10094: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10095: }
10096:
10097: for (k=agegomp;k<agemortsup;k++)
10098: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10099:
10100: for (k=agegomp;k<agemortsup;k++){
10101: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10102: sumlpop=sumlpop+lpop[k];
10103: }
10104:
10105: tpop[agegomp]=sumlpop;
10106: for (k=agegomp;k<(agemortsup-3);k++){
10107: /* tpop[k+1]=2;*/
10108: tpop[k+1]=tpop[k]-lpop[k];
10109: }
10110:
10111:
10112: printf("\nAge lx qx dx Lx Tx e(x)\n");
10113: for (k=agegomp;k<(agemortsup-2);k++)
10114: 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]);
10115:
10116:
10117: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10118: ageminpar=50;
10119: agemaxpar=100;
1.194 brouard 10120: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10121: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10122: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10123: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10124: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10125: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10126: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10127: }else{
10128: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10129: 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 10130: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10131: }
1.201 brouard 10132: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10133: stepm, weightopt,\
10134: model,imx,p,matcov,agemortsup);
10135:
10136: free_vector(lsurv,1,AGESUP);
10137: free_vector(lpop,1,AGESUP);
10138: free_vector(tpop,1,AGESUP);
1.220 brouard 10139: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10140: free_ivector(cens,1,n);
10141: free_vector(agecens,1,n);
10142: free_ivector(dcwave,1,n);
1.220 brouard 10143: #ifdef GSL
1.136 brouard 10144: #endif
1.186 brouard 10145: } /* Endof if mle==-3 mortality only */
1.205 brouard 10146: /* Standard */
10147: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10148: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10149: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10150: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10151: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10152: for (k=1; k<=npar;k++)
10153: printf(" %d %8.5f",k,p[k]);
10154: printf("\n");
1.205 brouard 10155: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10156: /* mlikeli uses func not funcone */
10157: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10158: }
10159: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10160: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10161: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10162: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10163: }
10164: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10165: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10166: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10167: for (k=1; k<=npar;k++)
10168: printf(" %d %8.5f",k,p[k]);
10169: printf("\n");
10170:
10171: /*--------- results files --------------*/
1.224 brouard 10172: 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 10173:
10174:
10175: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10176: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10177: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10178: for(i=1,jk=1; i <=nlstate; i++){
10179: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10180: if (k != i) {
10181: printf("%d%d ",i,k);
10182: fprintf(ficlog,"%d%d ",i,k);
10183: fprintf(ficres,"%1d%1d ",i,k);
10184: for(j=1; j <=ncovmodel; j++){
10185: printf("%12.7f ",p[jk]);
10186: fprintf(ficlog,"%12.7f ",p[jk]);
10187: fprintf(ficres,"%12.7f ",p[jk]);
10188: jk++;
10189: }
10190: printf("\n");
10191: fprintf(ficlog,"\n");
10192: fprintf(ficres,"\n");
10193: }
1.126 brouard 10194: }
10195: }
1.203 brouard 10196: if(mle != 0){
10197: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10198: ftolhess=ftol; /* Usually correct */
1.203 brouard 10199: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10200: 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");
10201: 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");
10202: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10203: for(k=1; k <=(nlstate+ndeath); k++){
10204: if (k != i) {
10205: printf("%d%d ",i,k);
10206: fprintf(ficlog,"%d%d ",i,k);
10207: for(j=1; j <=ncovmodel; j++){
10208: 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]));
10209: 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]));
10210: jk++;
10211: }
10212: printf("\n");
10213: fprintf(ficlog,"\n");
10214: }
10215: }
1.193 brouard 10216: }
1.203 brouard 10217: } /* end of hesscov and Wald tests */
1.225 brouard 10218:
1.203 brouard 10219: /* */
1.126 brouard 10220: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
10221: printf("# Scales (for hessian or gradient estimation)\n");
10222: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
10223: for(i=1,jk=1; i <=nlstate; i++){
10224: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 10225: if (j!=i) {
10226: fprintf(ficres,"%1d%1d",i,j);
10227: printf("%1d%1d",i,j);
10228: fprintf(ficlog,"%1d%1d",i,j);
10229: for(k=1; k<=ncovmodel;k++){
10230: printf(" %.5e",delti[jk]);
10231: fprintf(ficlog," %.5e",delti[jk]);
10232: fprintf(ficres," %.5e",delti[jk]);
10233: jk++;
10234: }
10235: printf("\n");
10236: fprintf(ficlog,"\n");
10237: fprintf(ficres,"\n");
10238: }
1.126 brouard 10239: }
10240: }
10241:
10242: 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 10243: if(mle >= 1) /* To big for the screen */
1.126 brouard 10244: 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");
10245: 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");
10246: /* # 121 Var(a12)\n\ */
10247: /* # 122 Cov(b12,a12) Var(b12)\n\ */
10248: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
10249: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
10250: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
10251: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
10252: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
10253: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
10254:
10255:
10256: /* Just to have a covariance matrix which will be more understandable
10257: even is we still don't want to manage dictionary of variables
10258: */
10259: for(itimes=1;itimes<=2;itimes++){
10260: jj=0;
10261: for(i=1; i <=nlstate; i++){
1.225 brouard 10262: for(j=1; j <=nlstate+ndeath; j++){
10263: if(j==i) continue;
10264: for(k=1; k<=ncovmodel;k++){
10265: jj++;
10266: ca[0]= k+'a'-1;ca[1]='\0';
10267: if(itimes==1){
10268: if(mle>=1)
10269: printf("#%1d%1d%d",i,j,k);
10270: fprintf(ficlog,"#%1d%1d%d",i,j,k);
10271: fprintf(ficres,"#%1d%1d%d",i,j,k);
10272: }else{
10273: if(mle>=1)
10274: printf("%1d%1d%d",i,j,k);
10275: fprintf(ficlog,"%1d%1d%d",i,j,k);
10276: fprintf(ficres,"%1d%1d%d",i,j,k);
10277: }
10278: ll=0;
10279: for(li=1;li <=nlstate; li++){
10280: for(lj=1;lj <=nlstate+ndeath; lj++){
10281: if(lj==li) continue;
10282: for(lk=1;lk<=ncovmodel;lk++){
10283: ll++;
10284: if(ll<=jj){
10285: cb[0]= lk +'a'-1;cb[1]='\0';
10286: if(ll<jj){
10287: if(itimes==1){
10288: if(mle>=1)
10289: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10290: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10291: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10292: }else{
10293: if(mle>=1)
10294: printf(" %.5e",matcov[jj][ll]);
10295: fprintf(ficlog," %.5e",matcov[jj][ll]);
10296: fprintf(ficres," %.5e",matcov[jj][ll]);
10297: }
10298: }else{
10299: if(itimes==1){
10300: if(mle>=1)
10301: printf(" Var(%s%1d%1d)",ca,i,j);
10302: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
10303: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
10304: }else{
10305: if(mle>=1)
10306: printf(" %.7e",matcov[jj][ll]);
10307: fprintf(ficlog," %.7e",matcov[jj][ll]);
10308: fprintf(ficres," %.7e",matcov[jj][ll]);
10309: }
10310: }
10311: }
10312: } /* end lk */
10313: } /* end lj */
10314: } /* end li */
10315: if(mle>=1)
10316: printf("\n");
10317: fprintf(ficlog,"\n");
10318: fprintf(ficres,"\n");
10319: numlinepar++;
10320: } /* end k*/
10321: } /*end j */
1.126 brouard 10322: } /* end i */
10323: } /* end itimes */
10324:
10325: fflush(ficlog);
10326: fflush(ficres);
1.225 brouard 10327: while(fgets(line, MAXLINE, ficpar)) {
10328: /* If line starts with a # it is a comment */
10329: if (line[0] == '#') {
10330: numlinepar++;
10331: fputs(line,stdout);
10332: fputs(line,ficparo);
10333: fputs(line,ficlog);
10334: continue;
10335: }else
10336: break;
10337: }
10338:
1.209 brouard 10339: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
10340: /* ungetc(c,ficpar); */
10341: /* fgets(line, MAXLINE, ficpar); */
10342: /* fputs(line,stdout); */
10343: /* fputs(line,ficparo); */
10344: /* } */
10345: /* ungetc(c,ficpar); */
1.126 brouard 10346:
10347: estepm=0;
1.209 brouard 10348: 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 10349:
10350: if (num_filled != 6) {
10351: 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);
10352: 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);
10353: goto end;
10354: }
10355: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
10356: }
10357: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
10358: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
10359:
1.209 brouard 10360: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 10361: if (estepm==0 || estepm < stepm) estepm=stepm;
10362: if (fage <= 2) {
10363: bage = ageminpar;
10364: fage = agemaxpar;
10365: }
10366:
10367: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 10368: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
10369: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 10370:
1.186 brouard 10371: /* Other stuffs, more or less useful */
1.126 brouard 10372: while((c=getc(ficpar))=='#' && c!= EOF){
10373: ungetc(c,ficpar);
10374: fgets(line, MAXLINE, ficpar);
1.141 brouard 10375: fputs(line,stdout);
1.126 brouard 10376: fputs(line,ficparo);
10377: }
10378: ungetc(c,ficpar);
10379:
10380: 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);
10381: 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);
10382: 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);
10383: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
10384: 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);
10385:
10386: while((c=getc(ficpar))=='#' && c!= EOF){
10387: ungetc(c,ficpar);
10388: fgets(line, MAXLINE, ficpar);
1.141 brouard 10389: fputs(line,stdout);
1.126 brouard 10390: fputs(line,ficparo);
10391: }
10392: ungetc(c,ficpar);
10393:
10394:
10395: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
10396: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
10397:
10398: fscanf(ficpar,"pop_based=%d\n",&popbased);
1.193 brouard 10399: fprintf(ficlog,"pop_based=%d\n",popbased);
1.126 brouard 10400: fprintf(ficparo,"pop_based=%d\n",popbased);
10401: fprintf(ficres,"pop_based=%d\n",popbased);
10402:
10403: while((c=getc(ficpar))=='#' && c!= EOF){
10404: ungetc(c,ficpar);
10405: fgets(line, MAXLINE, ficpar);
1.141 brouard 10406: fputs(line,stdout);
1.126 brouard 10407: fputs(line,ficparo);
10408: }
10409: ungetc(c,ficpar);
10410:
10411: 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);
10412: 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);
10413: 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);
10414: 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);
10415: 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);
10416: /* day and month of proj2 are not used but only year anproj2.*/
10417:
1.217 brouard 10418: while((c=getc(ficpar))=='#' && c!= EOF){
10419: ungetc(c,ficpar);
10420: fgets(line, MAXLINE, ficpar);
10421: fputs(line,stdout);
10422: fputs(line,ficparo);
10423: }
10424: ungetc(c,ficpar);
10425:
10426: 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 10427: 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);
10428: 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);
10429: 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 10430: /* day and month of proj2 are not used but only year anproj2.*/
1.126 brouard 10431:
1.230 brouard 10432: /* Results */
10433: while(fgets(line, MAXLINE, ficpar)) {
10434: /* If line starts with a # it is a comment */
10435: if (line[0] == '#') {
10436: numlinepar++;
10437: fputs(line,stdout);
10438: fputs(line,ficparo);
10439: fputs(line,ficlog);
10440: continue;
10441: }else
10442: break;
10443: }
10444: while((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
10445: if (num_filled == 0)
10446: resultline[0]='\0';
10447: else if (num_filled != 1){
10448: printf("ERROR %d: result line should be at minimum 'result=' %s\n",num_filled, line);
10449: }
10450: printf("Result %d: result line should be at minimum 'line=' %s, result=%s\n",num_filled, line, resultline);
10451: decoderesult(resultline);
10452: while(fgets(line, MAXLINE, ficpar)) {
10453: /* If line starts with a # it is a comment */
10454: if (line[0] == '#') {
10455: numlinepar++;
10456: fputs(line,stdout);
10457: fputs(line,ficparo);
10458: fputs(line,ficlog);
10459: continue;
10460: }else
10461: break;
10462: }
10463: if (feof(ficpar))
10464: break;
10465: else{ /* Processess output results for this combination of covariate values */
10466: }
10467: }
10468:
10469:
1.126 brouard 10470:
1.230 brouard 10471: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 10472: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 10473:
10474: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 10475: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 10476: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 10477: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10478: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 10479: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 10480: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10481: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10482: }else{
1.218 brouard 10483: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 10484: }
10485: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.225 brouard 10486: model,imx,jmin,jmax,jmean,rfileres,popforecast,prevfcast,backcast, estepm, \
10487: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 10488:
1.225 brouard 10489: /*------------ free_vector -------------*/
10490: /* chdir(path); */
1.220 brouard 10491:
1.215 brouard 10492: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
10493: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
10494: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
10495: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 10496: free_lvector(num,1,n);
10497: free_vector(agedc,1,n);
10498: /*free_matrix(covar,0,NCOVMAX,1,n);*/
10499: /*free_matrix(covar,1,NCOVMAX,1,n);*/
10500: fclose(ficparo);
10501: fclose(ficres);
1.220 brouard 10502:
10503:
1.186 brouard 10504: /* Other results (useful)*/
1.220 brouard 10505:
10506:
1.126 brouard 10507: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 10508: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
10509: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 10510: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 10511: fclose(ficrespl);
10512:
10513: /*------------- h Pij x at various ages ------------*/
1.180 brouard 10514: /*#include "hpijx.h"*/
10515: hPijx(p, bage, fage);
1.145 brouard 10516: fclose(ficrespij);
1.227 brouard 10517:
1.220 brouard 10518: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 10519: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 10520: k=1;
1.126 brouard 10521: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 10522:
1.219 brouard 10523: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 10524: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 10525: for(i=1;i<=AGESUP;i++)
1.219 brouard 10526: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 10527: for(k=1;k<=ncovcombmax;k++)
10528: probs[i][j][k]=0.;
1.219 brouard 10529: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
10530: if (mobilav!=0 ||mobilavproj !=0 ) {
10531: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 10532: for(i=1;i<=AGESUP;i++)
10533: for(j=1;j<=nlstate;j++)
10534: for(k=1;k<=ncovcombmax;k++)
10535: mobaverages[i][j][k]=0.;
1.219 brouard 10536: mobaverage=mobaverages;
10537: if (mobilav!=0) {
1.227 brouard 10538: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
10539: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
10540: printf(" Error in movingaverage mobilav=%d\n",mobilav);
10541: }
1.219 brouard 10542: }
10543: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
10544: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
10545: else if (mobilavproj !=0) {
1.227 brouard 10546: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
10547: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
10548: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
10549: }
1.219 brouard 10550: }
10551: }/* end if moving average */
1.227 brouard 10552:
1.126 brouard 10553: /*---------- Forecasting ------------------*/
10554: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
10555: if(prevfcast==1){
10556: /* if(stepm ==1){*/
1.225 brouard 10557: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 10558: }
1.217 brouard 10559: if(backcast==1){
1.219 brouard 10560: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10561: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10562: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
10563:
10564: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
10565:
10566: bprlim=matrix(1,nlstate,1,nlstate);
10567: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
10568: fclose(ficresplb);
10569:
1.222 brouard 10570: hBijx(p, bage, fage, mobaverage);
10571: fclose(ficrespijb);
1.219 brouard 10572: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
10573:
10574: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 10575: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 10576: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10577: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10578: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
10579: }
1.217 brouard 10580:
1.186 brouard 10581:
10582: /* ------ Other prevalence ratios------------ */
1.126 brouard 10583:
1.215 brouard 10584: free_ivector(wav,1,imx);
10585: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
10586: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
10587: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 10588:
10589:
1.127 brouard 10590: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 10591:
1.201 brouard 10592: strcpy(filerese,"E_");
10593: strcat(filerese,fileresu);
1.126 brouard 10594: if((ficreseij=fopen(filerese,"w"))==NULL) {
10595: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
10596: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
10597: }
1.208 brouard 10598: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
10599: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.219 brouard 10600:
1.227 brouard 10601: for (k=1; k <= (int) pow(2,cptcoveff); k++){ /* For any combination of dummy covariates, fixed and varying */
1.219 brouard 10602: fprintf(ficreseij,"\n#****** ");
1.225 brouard 10603: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 10604: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.219 brouard 10605: }
10606: fprintf(ficreseij,"******\n");
10607:
10608: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10609: oldm=oldms;savm=savms;
10610: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart);
1.127 brouard 10611:
1.219 brouard 10612: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 10613: }
10614: fclose(ficreseij);
1.208 brouard 10615: printf("done evsij\n");fflush(stdout);
10616: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 10617:
1.227 brouard 10618: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 10619:
10620:
1.201 brouard 10621: strcpy(filerest,"T_");
10622: strcat(filerest,fileresu);
1.127 brouard 10623: if((ficrest=fopen(filerest,"w"))==NULL) {
10624: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
10625: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
10626: }
1.208 brouard 10627: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
10628: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 10629:
1.126 brouard 10630:
1.201 brouard 10631: strcpy(fileresstde,"STDE_");
10632: strcat(fileresstde,fileresu);
1.126 brouard 10633: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 10634: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
10635: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 10636: }
1.227 brouard 10637: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
10638: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 10639:
1.201 brouard 10640: strcpy(filerescve,"CVE_");
10641: strcat(filerescve,fileresu);
1.126 brouard 10642: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 10643: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
10644: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 10645: }
1.227 brouard 10646: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
10647: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 10648:
1.201 brouard 10649: strcpy(fileresv,"V_");
10650: strcat(fileresv,fileresu);
1.126 brouard 10651: if((ficresvij=fopen(fileresv,"w"))==NULL) {
10652: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
10653: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
10654: }
1.227 brouard 10655: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
10656: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 10657:
1.145 brouard 10658: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
10659: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
10660:
1.225 brouard 10661: for (k=1; k <= (int) pow(2,cptcoveff); k++){
1.227 brouard 10662: printf("\n#****** ");
1.208 brouard 10663: fprintf(ficrest,"\n#****** ");
1.227 brouard 10664: fprintf(ficlog,"\n#****** ");
10665: for(j=1;j<=cptcoveff;j++){
10666: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10667: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10668: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10669: }
1.208 brouard 10670: fprintf(ficrest,"******\n");
1.227 brouard 10671: fprintf(ficlog,"******\n");
10672: printf("******\n");
1.208 brouard 10673:
10674: fprintf(ficresstdeij,"\n#****** ");
10675: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 10676: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 10677: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10678: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 10679: }
10680: fprintf(ficresstdeij,"******\n");
10681: fprintf(ficrescveij,"******\n");
10682:
10683: fprintf(ficresvij,"\n#****** ");
1.225 brouard 10684: for(j=1;j<=cptcoveff;j++)
1.227 brouard 10685: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 10686: fprintf(ficresvij,"******\n");
10687:
10688: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10689: oldm=oldms;savm=savms;
1.227 brouard 10690: printf(" cvevsij combination#=%d, ",k);
10691: fprintf(ficlog, " cvevsij combination#=%d, ",k);
1.208 brouard 10692: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart);
10693: printf(" end cvevsij \n ");
10694: fprintf(ficlog, " end cvevsij \n ");
10695:
10696: /*
10697: */
10698: /* goto endfree; */
10699:
10700: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
10701: pstamp(ficrest);
10702:
10703:
10704: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 10705: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
10706: cptcod= 0; /* To be deleted */
10707: printf("varevsij vpopbased=%d \n",vpopbased);
10708: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
10709: varevsij(optionfilefiname, vareij, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, estepm, cptcov,cptcod,vpopbased,mobilav, strstart); /* cptcod not initialized Intel */
10710: 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 ");
10711: if(vpopbased==1)
10712: 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);
10713: else
10714: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
10715: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
10716: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
10717: fprintf(ficrest,"\n");
10718: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
10719: epj=vector(1,nlstate+1);
10720: printf("Computing age specific period (stable) prevalences in each health state \n");
10721: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
10722: for(age=bage; age <=fage ;age++){
10723: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k); /*ZZ Is it the correct prevalim */
10724: if (vpopbased==1) {
10725: if(mobilav ==0){
10726: for(i=1; i<=nlstate;i++)
10727: prlim[i][i]=probs[(int)age][i][k];
10728: }else{ /* mobilav */
10729: for(i=1; i<=nlstate;i++)
10730: prlim[i][i]=mobaverage[(int)age][i][k];
10731: }
10732: }
1.219 brouard 10733:
1.227 brouard 10734: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
10735: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
10736: /* printf(" age %4.0f ",age); */
10737: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
10738: for(i=1, epj[j]=0.;i <=nlstate;i++) {
10739: epj[j] += prlim[i][i]*eij[i][j][(int)age];
10740: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
10741: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
10742: }
10743: epj[nlstate+1] +=epj[j];
10744: }
10745: /* printf(" age %4.0f \n",age); */
1.219 brouard 10746:
1.227 brouard 10747: for(i=1, vepp=0.;i <=nlstate;i++)
10748: for(j=1;j <=nlstate;j++)
10749: vepp += vareij[i][j][(int)age];
10750: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
10751: for(j=1;j <=nlstate;j++){
10752: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
10753: }
10754: fprintf(ficrest,"\n");
10755: }
1.208 brouard 10756: } /* End vpopbased */
10757: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
10758: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
10759: free_vector(epj,1,nlstate+1);
10760: printf("done \n");fflush(stdout);
10761: fprintf(ficlog,"done\n");fflush(ficlog);
10762:
1.145 brouard 10763: /*}*/
1.208 brouard 10764: } /* End k */
1.227 brouard 10765:
10766: printf("done State-specific expectancies\n");fflush(stdout);
10767: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
10768:
1.126 brouard 10769: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 10770:
1.201 brouard 10771: strcpy(fileresvpl,"VPL_");
10772: strcat(fileresvpl,fileresu);
1.126 brouard 10773: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
10774: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
10775: exit(0);
10776: }
1.208 brouard 10777: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
10778: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 10779:
1.145 brouard 10780: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
10781: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 10782:
1.225 brouard 10783: for (k=1; k <= (int) pow(2,cptcoveff); k++){
1.227 brouard 10784: fprintf(ficresvpl,"\n#****** ");
10785: printf("\n#****** ");
10786: fprintf(ficlog,"\n#****** ");
10787: for(j=1;j<=cptcoveff;j++) {
10788: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10789: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10790: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
10791: }
10792: fprintf(ficresvpl,"******\n");
10793: printf("******\n");
10794: fprintf(ficlog,"******\n");
10795:
10796: varpl=matrix(1,nlstate,(int) bage, (int) fage);
10797: oldm=oldms;savm=savms;
10798: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart);
10799: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 10800: /*}*/
1.126 brouard 10801: }
1.227 brouard 10802:
1.126 brouard 10803: fclose(ficresvpl);
1.208 brouard 10804: printf("done variance-covariance of period prevalence\n");fflush(stdout);
10805: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 10806:
10807: free_vector(weight,1,n);
10808: free_imatrix(Tvard,1,NCOVMAX,1,2);
10809: free_imatrix(s,1,maxwav+1,1,n);
10810: free_matrix(anint,1,maxwav,1,n);
10811: free_matrix(mint,1,maxwav,1,n);
10812: free_ivector(cod,1,n);
10813: free_ivector(tab,1,NCOVMAX);
10814: fclose(ficresstdeij);
10815: fclose(ficrescveij);
10816: fclose(ficresvij);
10817: fclose(ficrest);
10818: fclose(ficpar);
10819:
10820:
1.126 brouard 10821: /*---------- End : free ----------------*/
1.219 brouard 10822: if (mobilav!=0 ||mobilavproj !=0)
10823: 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 10824: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 10825: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
10826: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 10827: } /* mle==-3 arrives here for freeing */
1.227 brouard 10828: /* endfree:*/
10829: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
10830: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
10831: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
10832: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 ! brouard 10833: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 10834: free_matrix(coqvar,1,maxwav,1,n);
10835: free_matrix(covar,0,NCOVMAX,1,n);
10836: free_matrix(matcov,1,npar,1,npar);
10837: free_matrix(hess,1,npar,1,npar);
10838: /*free_vector(delti,1,npar);*/
10839: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
10840: free_matrix(agev,1,maxwav,1,imx);
10841: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
10842:
10843: free_ivector(ncodemax,1,NCOVMAX);
10844: free_ivector(ncodemaxwundef,1,NCOVMAX);
10845: free_ivector(Dummy,-1,NCOVMAX);
10846: free_ivector(Fixed,-1,NCOVMAX);
10847: free_ivector(Typevar,-1,NCOVMAX);
10848: free_ivector(Tvar,1,NCOVMAX);
1.231 brouard 10849: free_ivector(TvarFD,1,NCOVMAX);
10850: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 10851: free_ivector(TvarF,1,NCOVMAX);
10852: free_ivector(TvarFind,1,NCOVMAX);
10853: free_ivector(TvarV,1,NCOVMAX);
10854: free_ivector(TvarVind,1,NCOVMAX);
10855: free_ivector(TvarA,1,NCOVMAX);
10856: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 10857: free_ivector(TvarFQ,1,NCOVMAX);
10858: free_ivector(TvarFQind,1,NCOVMAX);
10859: free_ivector(TvarVD,1,NCOVMAX);
10860: free_ivector(TvarVDind,1,NCOVMAX);
10861: free_ivector(TvarVQ,1,NCOVMAX);
10862: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 10863: free_ivector(Tvarsel,1,NCOVMAX);
10864: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 10865: free_ivector(Tposprod,1,NCOVMAX);
10866: free_ivector(Tprod,1,NCOVMAX);
10867: free_ivector(Tvaraff,1,NCOVMAX);
10868: free_ivector(invalidvarcomb,1,ncovcombmax);
10869: free_ivector(Tage,1,NCOVMAX);
10870: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 10871: free_ivector(TmodelInvind,1,NCOVMAX);
10872: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 10873:
10874: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
10875: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 10876: fflush(fichtm);
10877: fflush(ficgp);
10878:
1.227 brouard 10879:
1.126 brouard 10880: if((nberr >0) || (nbwarn>0)){
1.216 brouard 10881: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
10882: 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 10883: }else{
10884: printf("End of Imach\n");
10885: fprintf(ficlog,"End of Imach\n");
10886: }
10887: printf("See log file on %s\n",filelog);
10888: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 10889: /*(void) gettimeofday(&end_time,&tzp);*/
10890: rend_time = time(NULL);
10891: end_time = *localtime(&rend_time);
10892: /* tml = *localtime(&end_time.tm_sec); */
10893: strcpy(strtend,asctime(&end_time));
1.126 brouard 10894: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
10895: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 10896: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 10897:
1.157 brouard 10898: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
10899: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
10900: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 10901: /* printf("Total time was %d uSec.\n", total_usecs);*/
10902: /* if(fileappend(fichtm,optionfilehtm)){ */
10903: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
10904: fclose(fichtm);
10905: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
10906: fclose(fichtmcov);
10907: fclose(ficgp);
10908: fclose(ficlog);
10909: /*------ End -----------*/
1.227 brouard 10910:
10911:
10912: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 10913: #ifdef WIN32
1.227 brouard 10914: if (_chdir(pathcd) != 0)
10915: printf("Can't move to directory %s!\n",path);
10916: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 10917: #else
1.227 brouard 10918: if(chdir(pathcd) != 0)
10919: printf("Can't move to directory %s!\n", path);
10920: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 10921: #endif
1.126 brouard 10922: printf("Current directory %s!\n",pathcd);
10923: /*strcat(plotcmd,CHARSEPARATOR);*/
10924: sprintf(plotcmd,"gnuplot");
1.157 brouard 10925: #ifdef _WIN32
1.126 brouard 10926: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
10927: #endif
10928: if(!stat(plotcmd,&info)){
1.158 brouard 10929: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 10930: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 10931: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 10932: }else
10933: strcpy(pplotcmd,plotcmd);
1.157 brouard 10934: #ifdef __unix
1.126 brouard 10935: strcpy(plotcmd,GNUPLOTPROGRAM);
10936: if(!stat(plotcmd,&info)){
1.158 brouard 10937: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 10938: }else
10939: strcpy(pplotcmd,plotcmd);
10940: #endif
10941: }else
10942: strcpy(pplotcmd,plotcmd);
10943:
10944: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 10945: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 10946:
1.126 brouard 10947: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 10948: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 10949: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 10950: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 10951: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 10952: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 10953: }
1.158 brouard 10954: printf(" Successful, please wait...");
1.126 brouard 10955: while (z[0] != 'q') {
10956: /* chdir(path); */
1.154 brouard 10957: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 10958: scanf("%s",z);
10959: /* if (z[0] == 'c') system("./imach"); */
10960: if (z[0] == 'e') {
1.158 brouard 10961: #ifdef __APPLE__
1.152 brouard 10962: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 10963: #elif __linux
10964: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 10965: #else
1.152 brouard 10966: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 10967: #endif
10968: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
10969: system(pplotcmd);
1.126 brouard 10970: }
10971: else if (z[0] == 'g') system(plotcmd);
10972: else if (z[0] == 'q') exit(0);
10973: }
1.227 brouard 10974: end:
1.126 brouard 10975: while (z[0] != 'q') {
1.195 brouard 10976: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 10977: scanf("%s",z);
10978: }
10979: }
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