Annotation of imach/src/ChangeLog, revision 1.2
1.2 ! brouard 1: 2003-05-03 <brouard@BROUARD>
! 2:
! 3: * imach.c: Some cleaning
1.1 brouard 4: 2003-03-28 <brouard@BROUARD>
5:
6: * imach.c (Module): In version up to 0.92 likelihood was computed
7: as if date of death was unknown. Death was treated as any other
8: health state: the date of the interview describes the actual state
9: and not the date of a change in health state. The former idea was
10: to consider that at each interview the state was recorded
11: (healthy, disable or death) and IMaCh was corrected; but when we
12: introduced the exact date of death then we should have modified
13: the contribution of an exact death to the likelihood. This new
14: contribution is smaller and very dependent of the step unit
15: stepm. It is no more the probability to die between last interview
16: and month of death but the probability to survive from last
17: interview up to one month before death multiplied by the
18: probability to die within a month. Thanks to Chris
19: Jackson for correcting this bug. Former versions increased
20: mortality artificially. The bad side is that we add another loop
21: which slows down the processing. The difference can be up to 10%
22: lower mortality.
23:
24: 2002-05-30 <brouard@BROUARD>
25:
26: *
27:
28: * imach.c (Module): Add correlation matrix of one-step
29: probabilities and covariance matrix
30:
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>