]> henry.ined.fr Git - .git/commit
git://henry.ined.fr / .git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: fdddea1) | patch
Extrapolation doesn't seem to work fine, so we decided to use
authorAgnès Lièvre <agnes.lievre@education.gouv.fr>
Thu, 21 Nov 2002 08:46:21 +0000 (08:46 +0000)
committerAgnès Lièvre <agnes.lievre@education.gouv.fr>
Thu, 21 Nov 2002 08:46:21 +0000 (08:46 +0000)
commitbc90b1f981c43e1eb9086e405033c6eb7e4e9c6a
treef40cc4ce3b09fe12df17a0d6888ae154fcfb6eectree | snapshot
parentfdddea19e3421cd2c4effa4a62cd6d3ace00be2bcommit | diff
Extrapolation doesn't seem to work fine, so we decided to use
interpolation only ie at the cost of an additional matrix product if
the bias, bbh, is positive, in order to make it negative
(interpolation).
We keep other methods mle=4 no adjustment for bias, mle=3 inter-extra
exponential, mle=2 inter-extra linear.
src/imach.c diff | blob | history
Current version 0.99s? : IMaCh program estimates Health Expectancies from Cross-longitudinal surveys using Interpolated Markov Chains.