CVS log for imach/src/imach.c

Up to [Local Repository] / imach / src

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

* imach.c (Module): Testing if conjugate gradient could be quicker when lot of variables POWELLORIGINCONJUGATE

* imach.c (Module): Testing if conjugate gradient could be quicker when lot of variables POWELLORIGINCONJUGATE

Summary: 0.99r46

* imach.c (Module): Fixed PROB_r

Summary: 0.99r46

* imach.c (Module): In the ILK....txt file, the number of columns
before the covariates values is dependent of the number of states (16+nlstate): 0.99r46

Summary: Temporary change for imachprax

*** empty log message ***

*** empty log message ***

Summary: 099r45

*** empty log message ***

Summary: Improvements in models with age*Vn*Vm

Summary: version 0.99r43

* imach.c (Module): Version 0.99r42 needed a newer version of
Gnuplot. But newer version 0.99r43 should run with the Gnuplot
version 5.0 or 5.1 distributed with IMaCh.

Summary: version 0.99r42

* src/imach.c (Module): 0.99r41 Was an error when product of timevarying and fixed. Using FixedV[of name] now. Thank you  Feinuo

Summary: Version 0.99r41

* imach.c (Module): 0.99r41 Was an error when product of timevarying and fixed. Using FixedV[of name] now. Thank you  Feinuo

Summary: version 0.99r40

* imach.c (Module): Fixing names of variables in T_ (thanks to Feinuo)

Summary: version 0.99r39

* imach.c (Module): Version 0.99r39 with colored dummy covariates
(fixed or time varying), using new last columns of
ILK_parameter.txt file.

Summary: 0.99r38

* imach.c (Module): Adding timevarying products of any kinds,
should work before shifting cotvar from ncovcol+nqv columns in
order to have a correspondance between the column of cotvar and
the id of column.
(Module): Some cleaning and adding covariates in ILK.txt

Summary: Version 0.99r38

After adding change in cotvar.

Summary: Version imach 0.99r37

* imach.c (Module): Adding timevarying products of any kinds,
should work before shifting cotvar from ncovcol+nqv columns in
order to have a correspondance between the column of cotvar and
the id of column.

Summary: version 0.99r37

* imach.c (Module): Many improvements for fixing products of fixed
timevarying as well as fixed * fixed, and test with quantitative
covariate.

Summary: 0.99r36

* imach.c (Module): Now the easy runs i.e. without result or
model=1+age only did not work. The defautl combination should be 1
and not 0 because everything hasn't been tranformed yet.

Summary: version 0.99r35

* src/imach.c: Version 0.99r35 because it outputs same results with
1+age+V1+V1*age for females and 1+age for females only
(education=1 noweight)

*** empty log message ***

Summary: improvements...

Summary: In progress for quantitative

* src/imach.c (Module): Version 0.99r33 A lot of changes in
reassigning covariates: my first idea was that people will always
use the first covariate V1 into the model but in fact they are
producing data with many covariates and can use an equation model
with some of the covariate; it means that in a model V2+V3 instead
of codtabm(k,Tvaraff[j]) which calculates for combination k, for
three covariates (V1, V2, V3) the value of Tvaraff[j], but in fact
the equation model is restricted to two variables only (V2, V3)
and the combination for V2 should be codtabm(k,1) instead of
(codtabm(k,2), and the code should be
codtabm(k,TnsdVar[Tvaraff[j]]. Many many changes have been
made. All of these should be simplified once a day like we did in
hpxij() for example by using precov[nres] which is computed in
decoderesult for each nres of each resultline. Loop should be done
on the equation model globally by distinguishing only product with
age (which are changing with age) and no more on type of
covariates, single dummies, single covariates.

Summary: Version 0.99r33

* src/imach.c (Module): Version 0.99r33 A lot of changes in
reassigning covariates: my first idea was that people will always
use the first covariate V1 into the model but in fact they are
producing data with many covariates and can use an equation model
with some of the covariate; it means that in a model V2+V3 instead
of codtabm(k,Tvaraff[j]) which calculates for combination k, for
three covariates (V1, V2, V3) the value of Tvaraff[j], but in fact
the equation model is restricted to two variables only (V2, V3)
and the combination for V2 should be codtabm(k,1) instead of
(codtabm(k,2), and the code should be
codtabm(k,TnsdVar[Tvaraff[j]]. Many many changes have been
made. All of these should be simplified once a day like we did in
hpxij() for example by using precov[nres] which is computed in
decoderesult for each nres of each resultline. Loop should be done
on the equation model globally by distinguishing only product with
age (which are changing with age) and no more on type of
covariates, single dummies, single covariates.

*** empty log message ***

Summary: last 0.99r31

*  imach.c (Module): Version of imach using partly decoderesult to rebuild xpxij function

*  imach.c (Module): Many errors in graphs fixed with Vn*age covariates.

Summary: valgrind bug fixed by initializing to zero DummyV as well as Tage

Summary: Still a problem for one-step probabilities in case of quantitative variables

Summary: some test with nres=1

Summary: r30

* imach.c (Module): Error cptcovn instead of nsd in bmij (was
coredumped, revealed by Feiuno, thank you.

*** empty log message ***

*  imach.c (Module): Output of Wald test in the htm file and not only in the log.

*  imach.c (Module): Output of Wald test in the htm file and not only in the log.

Summary: r28

*  imach.c (Module): Output of Wald test in the htm file and not only in the log.

*** empty log message ***

* imach.c (Module): Adding the Wald tests from the log to the main
htm for better display of the maximum likelihood estimators.

* imach.c (Module): Some attempts to find a bug of wrong estimates
of confidencce intervals with product in the equation modelC

* imach.c (Module):  Some minor improvements

Summary: r27

*** empty log message ***

* imach.c (Module): Adding link to text data files

* imach.c (Module): Error in rewriting the 'r' file with yearsfproj or yearsbproj fixed

*** empty log message ***

Summary: Fixed quantitative covariates

	  Fixed covariates (dummy or quantitative)
	with missing values have never been allowed but are ERRORS and
	program quits. Standard deviations of fixed covariates were
	wrongly computed. Mean and standard deviations of time varying
	covariates are still not computed.

Summary: 99r25

Improving detection of errors: result lines should be compatible with
the model.

Summary: Version 0.99r24

Summary: Version 0.99r23


* imach.c (Module): Still bugs in the result loop. Thank to Holly Benett

Summary: 0.99r22 fixed bug on result:

Summary: Version 0.99r21

* imach.c (Module): Fix bug on quitting after result lines!
(Module): Version 0.99r21

* imach.c (Module): Fix bug on quitting after result lines!

* imach.c (Module): The use of a Windows BOM (huge) file is now an error

*  (Module): imach.c Someone entered 'results:' instead of 'result:'. Now it is an error which is printed.

*  (Module): imach.c Update mle=-3 (for computing Life expectancy
and life table from the data without any state)

Summary: Error in 'r'parameter file backcast yearsbproj instead of yearsfproj

Summary: version 0.99r19 of May 2019

Summary: Cleaned 0.99r19

*** empty log message ***

Summary: Fix bug by moving date2dmy and nhstepm which gaefin=-1

Summary: Projection syntax simplified


We can now start projections, forward or backward, from the mean date
of inteviews up to or down to a number of years of projection:
prevforecast=1 yearsfproj=15.3 mobil_average=0
or
prevforecast=1 starting-proj-date=1/1/2007 final-proj-date=12/31/2017 mobil_average=0
or
prevbackcast=1 yearsbproj=12.3 mobil_average=1
or
prevbackcast=1 starting-back-date=1/10/1999 final-back-date=1/1/1985 mobil_average=1

Summary: doxygen tex bug

Summary: There was some wrong lines added

*** empty log message ***

Summary: Some updates

Summary: Before ncovmax

Summary: 0.99r18 unlimited number of individuals

The number n which was limited to 20,000 cases is now unlimited, from firstobs to lastobs. If the number is too for the virtual memory, probably an error will occur.

Summary: Bug for young ages (<-30) will be in r17

Summary: Some bugs fixed

Summary: Bug fixed by providing frequencies only for non missing covariates

Summary: some minor bugs

Summary: Some bugs fixed, valgrind tested

Summary: Computing mean and stdeviation of fixed quantitative variables

Summary: Some minor bugs fixed

*** empty log message ***

Summary: Adding second argument for quitting

Summary: 0.99r15

New Makefile with recent VirtualBox 5.26. Bug in sqrt negatve in imach.c

Summary: temporary working

Summary: Bug for mobil_average=0 and prevforecast fixed(?)

Summary: BOM files can be read now

Summary: Graphs improvements

Summary: Saito's color

Summary: Version 0.99r14

Summary: More documentation on projections

Summary: Color of backprojection changed from 6 to 5(yellow)

Summary: Some bug with rint

*** empty log message ***

Summary: Code into subroutine, cleanings

Summary: backprojection and confidence intervals of backprevalence

Summary: temporary save for backprojection

Summary: Version 0.99r13 (improvements and bugs fixed)

Summary: imach 0.99r13 Some bugs fixed

Summary: Labels in graphs

Summary: to save

*** empty log message ***

Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1

Summary: Gnuplot indexations fixed (humm)

Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3

Summary: Version 0.99r12

Some cleanings, conformed with updated documentation.

Summary: Temp

Summary: Temporary

Summary: IMaCh version 0.99r10 bugs in gnuplot fixed

Summary: Fixing data parameter line

Summary: 0.99 in progress

*** empty log message ***

Summary: not working

Summary: continue

Summary: Starting values from frequencies

*** empty log message ***

*** empty log message ***

*** empty log message ***

*** empty log message ***

*** empty log message ***

*** empty log message ***

Summary: Fixing a lots

Summary: gnuplot problem in Back projection to fix

Summary: Better

Summary: Improvement in Powell output in order to copy and paste

Author:

Summary: Starting tests of 0.99

Summary: to valgrind

*** empty log message ***

*** empty log message ***

*** empty log message ***

Summary: not working

Summary: not working

Summary: not working

Summary: Not working

Summary: Completing for func too

Summary: Fixing some arrays, still debugging

Summary: 0.99 working (more or less) for Asian Workshop on multitate methods

Summary: temp

Summary: saving but not running

Summary: Fixes

Summary: temporary

Summary: Probably last 0.98 stable version 0.98r6

Summary: minor bug

Summary: 0.99r2

*** empty log message ***

Summary: 0.99 Back projections

Summary: Experimental backcast

Summary: 0.98r4 Warning and status=-2

Version 0.98r4 is now:
 - displaying an error when status is -1, date of interview unknown and date of death known;
 - permitting a status -2 when the vital status is unknown at a known date of right truncation.
Older changes concerning s=-2, dating from 2005 have been supersed.

Summary: 0.98r4 working

Summary: temporary not working

Summary: 0.98r4

Summary: minor typo

Summary: 0.98r3 with some graph of projected cross-sectional

Author: Nicolas Brouard

Summary: Start working on projected prevalences

Summary: Adding ftolpl parameter
Author: N Brouard

We had difficulties to get smoothed confidence intervals. It was due
to the period prevalence which wasn't computed accurately. The inner
parameter ftolpl is now an outer parameter of the .imach parameter
file after estepm. If ftolpl is small 1.e-4 and estepm too,
computation are long.

Summary: temporary

*** empty log message ***

*** empty log message ***

Summary: 0.98r3 some clarification for graphs on likelihood contributions

Summary: Some new graphs of contribution to likelihood

Summary: looking at better estimation of the hessian

Also a better criteria for convergence to the period prevalence And
therefore adding the number of years needed to converge. (The
prevalence in any alive state shold sum to one

Summary: Adding some overall graph on contribution to likelihood. Might change

Summary: 0.98r0

- Some new graphs like suvival functions
- Some bugs fixed like model=1+age+V2.

Summary: Big bug thanks to Flavia

Even model=1+age+V2. did not work anymore

Summary: 0.98q6 changing default small png format for graph to vectorized svg.

Summary: 0.98q5 Flavia

*** empty log message ***

Summary: 0.98q5

Summary: Adding a hack for testing purpose

After reading the title, ftol and model lines, if the comment line has
a q, starting with #q, the answer at the end of the run is quit. It
permits to run test files in batch with ctest. The former workaround was
$ echo q | imach foo.imach

Summary:  Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.

Summary: 0.98q4

Summary: Fixing some outputs

Summary: Some fixes

Summary: Adding digits in output parameters (7 digits instead of 6)

Fix 1+age+.

Summary: 0.98q2

*** empty log message ***

*** empty log message ***

Summary: V1*age is working now, version 0.98q1

Some codes had been disabled in order to simplify and Vn*age was
working in the optimization phase, ie, giving correct MLE parameters,
but, as usual, outputs were not correct and program core dumped.

Summary: Inclusion of compile and links command line for Intel Compiler

Summary: Back from Windows 8. Intel Compiler

Summary: 0.98q0, trying with directest, mnbrak fixed

We use directest instead of original Powell test; probably no
incidence on the results, but better justifications;
We fixed Numerical Recipes mnbrak routine which was wrong and gave
wrong results.

Summary: Trying to keep directest which seems simpler and more general
Author: Nicolas Brouard

Summary: Comments on Powell added

Author:

Summary: Finishing move from main to function (hpijx and prevalence_limit)

Summary: back to OS/X

*** empty log message ***

Summary: Still testing ilc32 on OSX

*** empty log message ***

*** empty log message ***

Summary: Still in cross-compilation

Summary: trying to detect cross-compilation

Summary: Back from Visual Studio and Intel, options for compiling for Windows XP

Summary: Back from Visual C

Still problem with utsname.h on Windows

Summary: Cleaning some \%% back to %%

The escape was mandatory for a specific compiler (which one?), but too many warnings.

Summary: 0.98p

Outputs some informations on compiler used, OS etc. Testing on different platforms.

Summary: udate

Summary: Testing uname and compiler version and if compiled 32 or 64

Testing on Linux 64

*** empty log message ***

Summary: After compiling on Visual C

* imach.c (Module): Merging 1.61 to 1.162

Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn

* imach.c (Module): Merging 1.61 to 1.162

* imach.c (Module): Merging 1.61 to 1.162

Summary: temporary backup 0.99!

Summary: Problem with macro SQR on Intel compiler

*** empty log message ***

Summary: WIN32
Author: Brouard

*** empty log message ***

Summary: Preparing windows Visual studio version
Author: Brouard

In order to compile on Visual studio, time.h is now correct and time_t
and tm struct should be used. difftime should be used but sometimes I
just make the differences in raw time format (time(&now).
Trying to suppress #ifdef LINUX
Add xdg-open for __linux in order to open default browser.

*** empty log message ***

Summary: New compile, minor changes
Author: Brouard

Summary: Outputs now all graphs of convergence to period prevalence

Summary: If 3 live state, convergence to period prevalence on same graph
Author: Brouard

Summary: open browser, use gnuplot on same dir than imach if not found in the path

*** empty log message ***

Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
Author: brouard

Summary: Some fixes in parameter files errors
Author: Nicolas Brouard

Summary: Nothing new
Author: Brouard

Just a new packaging for OS/X version 0.98nS

*** empty log message ***

Summary: Merge
Author: Brouard

Merge, before building revised version.

Summary: Debugging with valgrind
Author: Nicolas Brouard

Lot of changes in order to output the results with some covariates
After the Edimburgh REVES conference 2014, it seems mandatory to
improve the code.
No more memory valgrind error but a lot has to be done in order to
continue the work of splitting the code into subroutines.
Also, decodemodel has been improved. Tricode is still not
optimal. nbcode should be improved. Documentation has been added in
the source code.

*** empty log message ***

Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising

* imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
(Module): Version 0.98nR Running ok, but output format still only works for three covariates.

Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2

* imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...

* imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...

Summary: times.h is ok with mingw32 now.

After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
I remember having already fixed agemin agemax which are pointers now but not cvs saved.

*** empty log message ***

(Module): Checking covariates for more complex models
than V1+V2. A lot of change to be done. Unstable.

(Module): merging some libgsl code. Fixing computation
of likelione (using inter/intrapolation if mle = 0) in order to
get same likelihood as if mle=1.
Some cleaning of code and comments added.

(Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.

(Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.

just nforces

Many tings

Some dimensions resccaled

(Module): Max Covariate is now set to 20 instead of 8. A
lot of cleaning with variables initialized to 0. Trying to make
V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.

Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting

(Module): Clarifications on computing e.j

(Module): Yes the sum of survivors was wrong since
imach-114 because nhstepm was no more computed in the age
loop. Now we define nhstepma in the age loop.
(Module): In order to speed up (in case of numerous covariates) we
compute health expectancies (without variances) in a first step
and then all the health expectancies with variances or standard
deviation (needs data from the Hessian matrices) which slows the
computation.
In the future we should be able to stop the program is only health
expectancies and graph are needed without standard deviations.

(Module): Yes the sum of survivors was wrong since
imach-114 because nhstepm was no more computed in the age
loop. Now we define nhstepma in the age loop.
Version 0.98h

Errors in calculation of health expectancies. Age was not initialized.
Forecasting file added.

Parameters are printed with %lf instead of %f (more numbers after the comma).
The log-likelihood is printed in the log file

* imach.c (Module): <title> changed, corresponds to .htm file
name. <head> headers where missing.

* imach.c (Module): Weights can have a decimal point as for
English (a comma might work with a correct LC_NUMERIC environment,
otherwise the weight is truncated).
Modification of warning when the covariates values are not 0 or
1.
Version 0.98g

(Module): Weights can have a decimal point as for
English (a comma might work with a correct LC_NUMERIC environment,
otherwise the weight is truncated).
Modification of warning when the covariates values are not 0 or
1.
Version 0.98g

* imach.c (Module): Comments concerning covariates added

* imach.c (Module): refinements in the computation of lli if
status=-2 in order to have more reliable computation if stepm is
not 1 month. Version 0.98f

(Module): refinements in the computation of lli if
status=-2 in order to have more reliable computation if stepm is
not 1 month. Version 0.98f

(Module): Bug if status = -2, the loglikelihood was
computed as likelihood omitting the logarithm. Version O.98e

(Module): varevsij Comments added explaining the second
table of variances if popbased=1 .
(Module): Covariances of eij, ekl added, graphs fixed, new html link.
(Module): Function pstamp added
(Module): Version 0.98d

(Module): varevsij Comments added explaining the second
table of variances if popbased=1 .
(Module): Covariances of eij, ekl added, graphs fixed, new html link.
(Module): Function pstamp added
(Module): Version 0.98d

(Module): Variance-covariance wrong links and
varian-covariance of ej. is needed (Saito).

(Module): One freematrix added in mlikeli! 0.98c

(Module): Some improvements in processing parameter
filename with strsep.

(Module): Memory leaks checks with valgrind and:
datafile was not closed, some imatrix were not freed and on matrix
allocation too.

(Module): Back to gnuplot.exe instead of wgnuplot.exe

(Module): Lots of cleaning and bugs added (Gompertz)
(Module): Comments can be added in data file. Missing date values
can be a simple dot '.'.

(Module): Lots of cleaning and bugs added (Gompertz)

(Module): Comments (lines starting with a #) are allowed in data.

Gnuplot problem appeared...
To be fixed

Test existence of gnuplot in imach path

Some cleaning and links added in html output

*** empty log message ***

(Module): sump fixed, loop imx fixed, and simplifications.
(Module): If the status is missing at the last wave but we know
that the person is alive, then we can code his/her status as -2
(instead of missing=-1 in earlier versions) and his/her
contributions to the likelihood is 1 - Prob of dying from last
health status (= 1-p13= p11+p12 in the easiest case of somebody in
the healthy state at last known wave). Version is 0.98

(Module): sump fixed, loop imx fixed, and simplifications.

Add the possibility to read data file including tab characters.

Fix on curr_time

Add version for Mac OS X. Just define UNIX in Makefile

*** empty log message ***

New version 0.97 . First attempt to estimate force of mortality
directly from the data i.e. without the need of knowing the health
state at each age, but using a Gompertz model: log u =a + b*age .
This is the basic analysis of mortality and should be done before any
other analysis, in order to test if the mortality estimated from the
cross-longitudinal survey is different from the mortality estimated
from other sources like vital statistic data.

The same imach parameter file can be used but the option for mle should be -3.

Agnès, who wrote this part of the code, tried to keep most of the
former routines in order to include the new code within the former code.

The output is very simple: only an estimate of the intercept and of
the slope with 95% confident intervals.

Current limitations:
A) Even if you enter covariates, i.e. with the
model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
B) There is no computation of Life Expectancy nor Life Table.

Version 0.96d. Population forecasting command line is (temporarily)
suppressed.

* imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
rewritten within the same printf. Workaround: many printfs.

* imach.c (Repository):
(Repository): Using imachwizard code to output a more meaningful covariance
matrix (cov(a12,c31) instead of numbers.

Just cleaning

(Module): On windows (cygwin) function asctime_r doesn't
exist so I changed back to asctime which exists.
(Module): Version 0.96b

(Module): On windows (cygwin) function asctime_r doesn't
exist so I changed back to asctime which exists.

* imach.c (Repository): Duplicated warning errors corrected.
(Repository): Elapsed time after each iteration is now output. It
helps to forecast when convergence will be reached. Elapsed time
is stamped in powell.  We created a new html file for the graphs
concerning matrix of covariance. It has extension -cov.htm.

(Module): Some bugs corrected for windows. Also, when
mle=-1 a template is output in file "or"mypar.txt with the design
of the covariance matrix to be input.

(Module): Some bugs corrected for windows. Also, when
mle=-1 a template is output in file "or"mypar.txt with the design
of the covariance matrix to be input.

* 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.

Version 0.96

(Module): Change position of html and gnuplot routines and added
routine fileappend.

* imach.c (Repository): Check when date of death was earlier that
current date of interview. It may happen when the death was just
prior to the death. In this case, dh was negative and likelihood
was wrong (infinity). We still send an "Error" but patch by
assuming that the date of death was just one stepm after the
interview.
(Repository): Because some people have very long ID (first column)
we changed int to long in num[] and we added a new lvector for
memory allocation. But we also truncated to 8 characters (left
truncation)
(Repository): No more line truncation errors.

* imach.c (Repository): Replace "freqsummary" at a correct
place. It differs from routine "prevalence" which may be called
many times. Probs is memory consuming and must be used with
parcimony.
Version 0.95a2 (should output exactly the same maximization than 0.8a2)

We decided to release, in June 2003, a version 0.8a1 (May 2002) of
former version 0.8a to correct a bug highlighted by Chris Jackson.
It can be useful until version 0.9x will be released.

In version 0.8a, likelihood was computed as if date of death was unknown. Death
was treated as any other health state, i.e the date of the interview
describes the actual state and not the date of a change in health
state.

It means that in the early version of IMaCh we were considering that
at each interview the current state was recorded (healthy, disable or
death).  And IMaCh was correct. But, when we introduced the exact date
of death then we should have modified the contribution of an exact
(more precise) date of death to the likelihood. This new contribution
is smaller and very dependent of the step unit stepm. It is no more
the probability to die between last interview and month of death
(which is the way version 0.8a of IMaCh was doing) but the probability
to survive from last interview up to one month before death multiplied
by the probability to die within a month. Thanks to Chris Jackson for
correcting this bug.  Former versions increased mortality
artificially. The bad side is that we add another loop which slows
down the processing. The difference can be up to 10% lower mortality.

Minor changes to run version 0.8a on Linux. Bad ifdef linux.

*** empty log message ***

Add log in  imach.c and  fullversion number is now printed.

*** empty log message ***

Trying to add the true revision is the program and log

*** empty log message ***

minor changes

(Module): Insert of a warning if the delay between two waves is
negative.

(Module):

Some cleaning

Check memory allocation through valgrind on Linux and libnjamd .
So it is free of malloc bugs now!
Version 0.95

In freqsummary, the observed prevalence is now computed from all the
waves and the bug in the forecasting routine is corrected. We can
forecast prevalence if ncov=0.

(Module): In version up to 0.92 likelihood was computed
as if date of death was unknown. Death was treated as any other
health state: the date of the interview describes the actual state
and not the date of a change in health state. The former idea was
to consider that at each interview the state was recorded
(healthy, disable or death) and IMaCh was corrected; but when we
introduced the exact date of death then we should have modified
the contribution of an exact death to the likelihood. This new
contribution is smaller and very dependent of the step unit
stepm. It is no more the probability to die between last interview
and month of death but the probability to survive from last
interview up to one month before death multiplied by the
probability to die within a month. Thanks to Chris
Jackson for correcting this bug.  Former versions increased
mortality artificially. The bad side is that we add another loop
which slows down the processing. The difference can be up to 10%
lower mortality. Version 0.93

(Module): In version up to 0.92 likelihood was computed
as if date of death was unknown. Death was treated as any other
health state: the date of the interview describes the actual state
and not the date of a change in health state. The former idea was
to consider that at each interview the state was recorded
(healthy, disable or death) and IMaCh was corrected; but when we
introduced the exact date of death then we should have modified
the contribution of an exact death to the likelihood. This new
contribution is smaller and very dependent of the step unit
stepm. It is no more the probability to die between last interview
and month of death but the probability to survive from last
interview up to one month before death multiplied by the
probability to die within a month. Thanks to Chris
Jackson for correcting this bug.  Former versions increased
mortality artificially. The bad side is that we add another loop
which slows down the processing. The difference can be up to 10%
lower mortality.

Preforecast routine cleaned. Change of notation, comments added.
Version 0.92

*** empty log message ***

Correction of initialisations in case of several death states.

Fixed probability of death over estepm months. We don't calculate the
probability of death over a year anymore. Fix estepm as 12 and you
will get it.

I changed some comments

Correction of bug related to the covariates

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.

Trying a linear interpolation extrapolation in the computing of
likelihood for large stepm.

just a sign error in mlikeli

mle=1 corrects bias
mle=2 no correction

Allocation bug

Major change: likelihood is modified in order to unbiased the duration
between 2 waves which is rounded (biased) to the nearest multiple of
stepm.
Some cleaning of the code. Many allocations were not freed correctly.

Comments added and corrections in routine moving average done.

*** empty log message ***

*** empty log message ***

Bug in ***movingaverage

mobilav can be 3, 5 or 7 instead of only 1 (which defaults to 5).

Moving average of observed prevalence moved

Some improvments

Some errors corrected in diagonalisation of covariance-matrice

*** empty log message ***

Changes in the htm file.

version 0.8h

Many changes:
(a) We look at covariance between step probabilities by drawing
confidence interval ellipsoids of two step probabilities pij versus
pkl.
It is done by diagonalizing the inverse of the convariance matrix.

(b) In order to write to the gnuplot file or even to the html file
without using to much memory we are now reopening ficgp and fichtm
with an "append" option each time we need them into the main program.
It is better than keeping matrices in memory until the end.

(Module): Add correlation matrix of one-step
probabilities and covariance matrix

The bug in printinghtm is corrected

Change version number ot 0.8e

Change of gif to png.
Clarify some of the outputs.
Add a new graph: pij*YEARM/stepm in order to get pseudo-incidence per
year.

Change gif images to png format because of gif copyright.
Change wgnuplot to gnuplot
Correct a small bug

Version 0.8a. Standard deviation of health expectancies eij.

*** empty log message ***

*** empty log message ***

The bug in printinghtml is corrected.

Changed  version to 0.8b

Estepm added. We are calculating Life expectancy by interpolating the
survival  function pijx linearly every estepm month. Estepm should be
higher than stepm. It is a new parameter (sorry).

Version used for article20.tex Changes in varevsij.

CHANGE ncov to ncovcol
There was a confusion with older ncov parameter. In fact it was the
number of columns, between id and date of birth, which can be used for
covariates. In the program we use ncovmodel for the real number of
covariates. Version 0.8 !

Some cleaning in esij and nhstepm

Supress a bug in esij

0.71a cleaning done

Version 0.71

Output mle=0 in the Rparameter file so thart clicking on
rparameterfil.imach will skip the optimization process.

Some improvements in forecasting and the bug relating to agemax is
corrected.

Some changes in population forecasting

Add date in the freqsummary file.

Cleaning of the main function: gnuplot and html file are now
subroutines; the bug in the forecasting is corrected; the moving
average statement moved to the last command line.

*** empty log message ***

Merge OK

First merge between Agn�ès and Nicolas!!

Error 108 if stepm is not equal to 1 (month) because forecasting
is only possible with a former optimization with a step of one month.

Prevalence and forecasting

Forecasting and calculation of the mean date of interview

Forecasting

Some bugs corrected

Version 0.7
hPijx are now computed for h=1 year instead of 2 years

Moving average  on cross-sectional prevalences

Forecasting

Variance of one-step transitions added and forecasting of prevalences

Remise �à jour du serveur CVS peu de changements

Updated to Windows2000: changes in the path name to execute wgnuplot.

Changes in calculation spaces between two waves

Change of index.htm to parameterfile.htm
Version 0.64b

Changes index.htm: title, model, ...

Add age*V1

some changes index.htm
Version16.

suppressing cygwinpath. split directory instead.
version 15

*** empty log message ***

version 13

Version Rostock: covariables + graphique proba elementaires

Import de imach064

Initial revision