version 1.1, 2004/06/17 17:36:38
|
version 1.2, 2006/01/27 20:17:50
|
Line 1
|
Line 1
|
Interpolated Marlov Chain (IMaCh program)
|
Interpolated Marlov Chain (IMaCh program)
|
|
|
This program is GPL licensed for the part the code that we wrote at
|
This program is GPL licensed for the part the code that we wrote at
|
INED. But currently (version 0.97b may 2004) maximization routines use
|
INED. But currently (version 0.98b January 2006) maximization routines use
|
a modified version of the patented routines from "Numerical Recipes in
|
a modified version of the patented routines from "Numerical Recipes in
|
C". If you have the right to use the sources of their software you can
|
C". If you have the right to use the sources of their software you can
|
get access to the source code and to the CVS tree of IMaCh.
|
get access to the source code and to the CVS tree of IMaCh.
|
Line 9 get access to the source code and to the
|
Line 9 get access to the source code and to the
|
The program estimates Health Expectancies from Cross-longitudinal
|
The program estimates Health Expectancies from Cross-longitudinal
|
studies.
|
studies.
|
The main publication is Lievre, A. Brouard. N, Heathcote
|
The main publication is Lievre, A. Brouard. N, Heathcote
|
Ch. Estimating Health Expectancies from Corss-longitudinal surveys,
|
Ch. Estimating Health Expectancies from Cross-longitudinal surveys,
|
Mathematical Population Studies, 10(1), 2003, pp. 1-52.
|
Mathematical Population Studies, 10(1), 2003, pp. 1-52.
|
|
|
To be brief, for each people interviewed we have the age and health
|
To be brief, for each people interviewed we need the age and health
|
states at each interview. Also the month of death is known. We model
|
status at each interview. Also the month and year of death should be
|
the probability to be observed in any state at any age via a
|
known as for birth. We model the probability to be observed in any
|
parametric model and estimate the parameters and their confidence interval
|
state at any age via a parametric model and estimate the parameters
|
by maximum likelihood, many tables and graphs are output in
|
and their confidence interval by maximum likelihood. Many tables and
|
HTML. Graphs are built using gnuplot.
|
graphs are output in HTML. Graphs are built using gnuplot.
|
|
|
In the Windows version, gnuplot is also distributed. Version 4 is out
|
In the Windows version, gnuplot is also distributed. Version 4 of
|
since early 2004 and is used by latest version of Imach (at least
|
gnuplot is out since early 2004 and is used by latest version of Imach
|
version 0.38i is required).
|
(at least version 0.38i of gnuplot is required). IMaCh is supposed to work on
|
IMaCh is supposed to work on Linux, MacIntosh and Windows (9X or
|
Linux, Unix, MacIntosh (Panther, Tiger) and Windows (9X or NT/XP).
|
NT/XP).
|
|
|
|
WINDOWS
|
WINDOWS
|
On Windows it is distributed via a setup file (no more a zip file). As
|
On Windows it is distributed via a setup file (no more a zip file). As
|
usual you can put the 'imach' directory in Program Files or
|
usual, you can put the 'imach' directory in Program Files or
|
wherever. The directory contains at least 2 directories, 'bin' for the
|
wherever. The directory contains at least 2 directories, 'bin' for the
|
binaries which are imach.exe and gnuplot.exe, doc for some of the
|
binaries which are imach.exe and gnuplot.exe, doc for some of the
|
documentation, and 'mytry' in order to test the program on a parameter
|
documentation, and 'mytry' in order to test the program on a parameter
|
Line 66 You may subscribe to imach-users@listes.
|
Line 65 You may subscribe to imach-users@listes.
|
imach-users-subscribe@listes.ined.fr and you can unsubscribe with a
|
imach-users-subscribe@listes.ined.fr and you can unsubscribe with a
|
mail to imach-users-unsubscribe@listes.ined.fr .
|
mail to imach-users-unsubscribe@listes.ined.fr .
|
|
|
|
Other comments are included on the Wiki server
|
|
http://euroreves.ined.fr/imach/wiki, please use it ask questions too.
|
|
|
Nicolas Brouard (brouard at ined point fr)
|
Nicolas Brouard (brouard at ined point fr)
|
Agnès Lièvre (lievre chez ined dot fr)
|
Agnès Lièvre (lievre chez ined dot fr)
|
Christopher Heathcote
|
Christopher Heathcote
|