Annotation of imach/src/ChangeLog, revision 1.33
1.33 ! brouard 1: 2010-04-29 brouard <brouard@ined.fr>
! 2:
! 3: * imach.c (Module): Checking covariates for more complex models
! 4: than V1+V2. A lot of change to be done. Unstable.
! 5:
! 6: 2010-04-26 brouard <brouard@ined.fr>
! 7:
! 8: * imach.c (Module): merging some libgsl code. Fixing computation
! 9: of likelione (using inter/intrapolation if mle = 0) in order to
! 10: get same likelihood as if mle=1.
! 11: Some cleaning of code and comments added.
! 12:
1.32 brouard 13: 2009-10-29 brouard <brouard@ined.fr>
14:
15: * imach.c (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
16:
1.31 brouard 17: 2006-06-30 Brouard Nicolas <brouard@localhost>
18:
19: * imach.c (Module): Clarifications on computing e.j
20:
1.30 brouard 21: 2006-04-28 Brouard Nicolas <brouard@localhost>
22:
23: * imach.c (Module): Yes the sum of survivors was wrong since
24: imach-114 because nhstepm was no more computed in the age
25: loop. Now we define nhstepma in the age loop.
1.31 brouard 26: (Module): In order to speed up (in case of numerous covariates) we
27: compute health expectancies (without variances) in a first step
28: and then all the health expectancies with variances or standard
29: deviation (needs data from the Hessian matrices) which slows the
30: computation.
31: In the future we should be able to stop the program is only health
32: expectancies and graph are needed without standard deviations.
1.30 brouard 33:
1.29 lievre 34: 2006-04-04 LIEVRE Agnes <lievre@PC_00697.ined.fr>
35:
36: * imach.iss (Module):
37:
38: * imach.c (Module): bugs in evsi + population file added in htm output file
39:
40: 2006-03-22 LIEVRE Agnes <lievre@PC_00697.ined.fr>
41:
42: * imach.c (Module):
43:
1.27 brouard 44: 2006-03-20 Brouard Nicolas <brouard@localhost>
45:
1.28 brouard 46: * imach.c (Module): <title> changed, corresponds to .htm file
47: name. <head> headers where missing.
48:
1.27 brouard 49: * Makefile (Module): 0.98g
50:
51: * imach.c (Module): Weights can have a decimal point as for
52: English (a comma might work with a correct LC_NUMERIC environment,
53: otherwise the weight is truncated).
54: Modification of warning when the covariates values are not 0 or
55: 1.
56: Version 0.98g
57:
1.24 lievre 58: 2006-03-16 LIEVRE Agnes <lievre@PC_00697.ined.fr>
59:
1.26 lievre 60: * imach.c (Module): Comments concerning covariates added
61:
1.25 lievre 62: * noreg-setup.iss (Module): Version 0.98f
63:
1.24 lievre 64: * Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup
65:
66: * imach.c (Module): refinements in the computation of lli if
67: status=-2 in order to have more reliable computation if stepm is
68: not 1 month. Version 0.98f
69:
1.22 brouard 70: 2006-03-15 Brouard Nicolas <brouard@localhost>
71:
72: * imach.c (Module): Bug if status = -2, the loglikelihood was
73: computed as likelihood omitting the logarithm. Version O.98e
74:
1.21 brouard 75: 2006-03-14 Brouard Nicolas <brouard@localhost>
76:
77: * imach.c (Module): varevsij Comments added explaining the second
78: table of variances if popbased=1 .
79: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
80: (Module): Function pstamp added
81: (Module): Version 0.98d
82:
1.20 brouard 83: 2006-03-06 Brouard Nicolas <brouard@localhost>
84:
85: * imach.c (Module): Variance-covariance wrong links and
86: varian-covariance of ej. is needed (Saito).
87:
1.19 brouard 88: 2006-02-27 Brouard Nicolas <brouard@localhost>
89:
90: * imach.c (Module): One freematrix added in mlikeli!
91:
92: 2006-02-26 Brouard Nicolas <brouard@localhost>
93:
94: * imach.c (Module): Some improvements in processing parameter
95: filename with strsep.
96:
97: 2006-02-24 Brouard Nicolas <brouard@localhost>
98:
99: * imach.c (Module): Memory leaks checks with valgrind and:
100: datafile was not closed, some imatrix were not freed and on matrix
101: allocation too.
102:
1.16 lievre 103: 2006-02-09 LIEVRE Agnes <lievre@PC_00697.ined.fr>
104:
1.18 lievre 105: * strsep.c (Module): Inclusion because not in mingw32!!!
106:
1.17 lievre 107: * setup.iss.in (Module): IMACHSETUPVERSION added
108:
109: * noreg-setup.iss (Module): added to cvs tree
110:
1.16 lievre 111: * Makefile (Module): IMACHSETUPVERSION added
112:
1.15 brouard 113: 2006-01-30 BROUARD Nicolas <brouard@localhost>
114:
1.23 brouard 115: * Makefile (Module, Module): Latest Makefile
116:
1.15 brouard 117: * imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe
118:
119: 2006-01-27 BROUARD Nicolas <brouard@localhost>
120:
121: * noreg-setup.iss (Module): ADDED
122:
123: 2006-01-24 BROUARD Nicolas <brouard@localhost>
124:
125: * imach.c (Module): Comments (lines starting with a #) are allowed in data.
126:
1.14 brouard 127: 2005-10-25 Nicolas Brouard <brouard@dhcp171.recherche.ined.fr>
128:
129: * Makefile (Module): Added Imach98.pmsp and howto crete a dmg
130:
131: * createdmg.sh (Module): Creates a dmg disk for MaC OS/X
132:
1.12 lievre 133: 2005-09-30 <lievre@PC_00595>
134:
135: * imach.c (Module): sump fixed, loop imx fixed, and simplifications.
1.13 lievre 136: (Module): If the status is missing at the last wave but we know
137: that the person is alive, then we can code his/her status as -2
138: (instead of missing=-1 in earlier versions) and his/her
139: contributions to the likelihood is 1 - Prob of dying from last
140: health status (= 1-p13= p11+p12 in the easiest case of somebody in
141: the healthy state at last known wave). Version is 0.98
1.12 lievre 142:
1.11 brouard 143: 2004-05-20 Brouard Nicolas <brouard@localhost>
144:
145: * imach.c (Repository):
146: Agnes added a direct estimation of mortality (without the need of
147: computing period prevalence and differential mortality). Thus here
148: is version 0.97a which has been distributed to some people at
149: REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of
150: mortality using covariates is not done today. Estimating direct
151: mortality is a very different process because it doesn't need
152: interpolation because it is easy to get the lx from the force of
153: the mortality mux in the simplest case as for a Gompertz (log mux
154: = a + b*x . But we have been able to incorporate the new code
155: within former imach program (0.96d) without deteriorating too much
156: the understanding of the program.
157:
1.10 brouard 158: 2003-06-25 <brouard@BROUARD>
159:
160: * imach.c (Module): On windows (cygwin) function asctime_r doesn't
161: exist so I changed back to asctime which exists.
162: (Module): Version 0.96b
163:
164: 2003-06-24 <brouard@BROUARD>
165:
166: * imach.c (Module): Some bugs corrected for windows. Also, when
167: mle=-1 a template is output in file "or"mypar.txt with the design
168: of the covariance matrix to be input.
169:
1.9 brouard 170: 2003-06-25 Brouard Nicolas <brouard@homebrou>
171:
172: * imach.c (Repository): Duplicated warning errors corrected.
173: (Repository): Elapsed time after each iteration is now output. It
174: helps to forecast when convergence will be reached. Elapsed time
175: is stamped in powell. We created a new html file for the graphs
176: concerning matrix of covariance. It has extension -cov.htm.
177:
1.8 brouard 178: 2003-06-23 Brouard Nicolas <brouard@homebrou>
179:
180: * 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.
181:
182: 2003-06-18 Brouard Nicolas <brouard@homebrou>
183:
184: * imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96
185:
1.7 brouard 186: 2003-06-17 Brouard Nicolas <brouard@homebrou>
187:
188:
189: * imach.c (Repository): Check when date of death was earlier that
190: current date of interview. It may happen when the death was just
191: prior to the death. In this case, dh was negative and likelihood
192: was wrong (infinity). We still send an "Error" but patch by
193: assuming that the date of death was just one stepm after the
194: interview.
195: (Repository): Because some people have very long ID (first column)
196: we changed int to long in num[] and we added a new lvector for
197: memory allocation. But we also truncated to 8 characters (left
198: truncation)
199:
200: (Repository): No more line truncation errors.
201:
202: 2003-06-14 Brouard Nicolas <brouard@homebrou>
203:
204: * imach.c (Repository): Add of a routine likelione (likelihood
205: only once) which prints on a text file
206: (ilk) the contributions to the likelihood for each
207: individual/wave.
208:
209: 2003-06-13 Brouard Nicolas <brouard@homebrou>
210: * imach.c (Repository): Replace "freqsummary" at a correct
211: place. It differs from routine "prevalence" which may be called
212: many times. Probs is memory consuming and must be used with
213: parcimony.
214:
215: 2003-06-17 <brouard@BROUARD>
216:
217: * timeval.h (Module): Added included file to make use of
218: gettimeofday working on win32 with cygwin.
219:
220: 2003-05-16 Brouard Nicolas <brouard@brouard>
221: * imach.c (Module): Insert of a warning if the delay between two
222: waves is negative.
223:
224: * imach.c (Module):
225:
226: 2003-05-03 <brouard@BROUARD>
227:
228: * imach.c: Some cleaning
229:
230: 2003-03-28 <brouard@BROUARD>
231: * imach.c (Module): In version up to 0.92 likelihood was computed
232: as if date of death was unknown. Death was treated as any other
233: health state: the date of the interview describes the actual state
234: and not the date of a change in health state. The former idea was
235: to consider that at each interview the state was recorded
236: (healthy, disable or death) and IMaCh was corrected; but when we
237: introduced the exact date of death then we should have modified
238: the contribution of an exact death to the likelihood. This new
239: contribution is smaller and very dependent of the step unit
240: stepm. It is no more the probability to die between last interview
241: and month of death but the probability to survive from last
242: interview up to one month before death multiplied by the
243: probability to die within a month. Thanks to Chris
244: Jackson for correcting this bug. Former versions increased
245: mortality artificially. The bad side is that we add another loop
246:
247:
248: which slows down the processing. The difference can be up to 10%
249:
250:
251: lower mortality.
252:
253:
254:
255:
256:
257: 2002-05-30 <brouard@BROUARD>
258:
259:
260:
261:
262:
263: *
264:
265:
266:
267:
268:
269: * imach.c (Module): Add correlation matrix of one-step
270:
271:
272: probabilities and covariance matrix
273:
274:
275:
276:
277:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/small/back.gif){kind=icon}
Return to ChangeLog CVS log ![[TXT]](/icons/small/text.gif){kind=icon}
![[DIR]](/icons/small/dir.gif){kind=icon}
Up to [Local Repository] / imach / src