mle derivatives
by Gordon Hughes
As I understand, when mle is given derivatives of the log-likelihood,
it checks the computed derivatives against its own numerical
derivatives at the first iteration. If they are different, the
program reports that there is a difference and stops. This is a very
useful test, but the diagnostic information could be a bit more helpful.
Would there be any problem in reporting the computed derivatives and
the numerical derivatives, because this would help the programmer
identify where the error in programming the derivatives has
occurred? At the moment I am struggling with an unusually messy set
of derivatives. While I can, I think, persuade the program to
generate the comparison, it would be much easier if this were done
automatically before the program stops.
As a further thought: what is the test that the numerical and
computed derivatives are different? I can imagine circumstances
where one might start the estimation from a potentially degenerate
set of initial parameter values that might lead to a false positive
for the difference between the two.
Gordon
17 years
Strange output of validate gretl_commands.xml
by Hélio Guilherme
Allin,
This wonderful tool you made, "validate" is giving some strange
characters in the output when gretl_commands.xml is tested. It messes
the terminal output, by turning on graphic chars mode.
Please use:
validate gretl_commands.xml | less (instead of using more)
Attached are examples of the files for you to see using an editor.
Regards,
Hélio Guilherme
Found 121 commands in 'gretl_commands.xml'
* libgretl command 'error' is not in the reference
* libgretl command 'omitfrom' is not in the reference
* libgretl command '' is not in the reference
* libgretl command '' is not in the reference
* 'append' appears to have 2 options in lib, 1 in manual
* 'append': lib option '--quiet' unmatched in ref
* 'ar1' appears to have 5 options in lib, 4 in manual
* 'ar1': lib option '--quiet' unmatched in ref
* 'arbond': ref option '--quiet' unmatched in lib
* 'arbond': lib option '--orthdev' unmatched in ref
* 'arima' appears to have 8 options in lib, 6 in manual
* 'arima': lib option '--hessian' unmatched in ref
* 'arima': lib option '--quiet' unmatched in ref
* 'corrgm' appears to have 1 options in lib, 0 in manual
* 'corrgm': lib option '--quiet' unmatched in ref
* 'data' appears to have 1 options in lib, 0 in manual
* 'data': lib option '--odbc' unmatched in ref
* 'dataset' appears to have 1 options in lib, 0 in manual
* 'dataset': lib option '--preserve' unmatched in ref
* 'eqnprint' appears to have 2 options in lib, 1 in manual
* 'eqnprint': lib option '--t-ratios' unmatched in ref
* 'estimate' appears to have 5 options in lib, 4 in manual
* 'estimate': lib option '--quiet' unmatched in ref
* 'gmm' appears to have 5 options in lib, 4 in manual
* 'gmm': lib option '--quiet' unmatched in ref
* 'gnuplot' appears to have 10 options in lib, 9 in manual
* 'gnuplot': ref option '--suppress-fit' unmatched in lib
* 'gnuplot': lib option '--suppress-fitted' unmatched in ref
* 'gnuplot': lib option '--control' unmatched in ref
* 'heckit' appears to have 4 options in lib, 3 in manual
* 'heckit': lib option '--ml' unmatched in ref
* 'hsk' appears to have 2 options in lib, 1 in manual
* 'hsk': lib option '--quiet' unmatched in ref
* 'logit' appears to have 5 options in lib, 4 in manual
* 'logit': lib option '--quiet' unmatched in ref
* 'mpols' appears to have 4 options in lib, 3 in manual
* 'nls' appears to have 5 options in lib, 4 in manual
* 'omit' appears to have 8 options in lib, 7 in manual
* 'omit': lib option '--bootstrap' unmatched in ref
* 'open' appears to have 5 options in lib, 2 in manual
* 'open': ref option 'š°áááx0·x0·ZZÈ[@-0·0·0·0· 0· 0· g °0·°0·p"à3š1À€Ð0·Ð0·Ø0·Ø0·`Kž3X,X,ðK`WÈ!È!' unmatched in lib
* 'open': lib option '--coded' unmatched in ref
* 'open': lib option '--odbc' unmatched in ref
* 'open': lib option '--preserve' unmatched in ref
* 'open': lib option '--quiet' unmatched in ref
* 'panel' appears to have 14 options in lib, 9 in manual
* 'panel': lib option '--hausman-reg' unmatched in ref
* 'panel': lib option '--pooled' unmatched in ref
* 'panel': lib option '--robust' unmatched in ref
* 'panel': lib option '--silent' unmatched in ref
* 'print' appears to have 4 options in lib, 3 in manual
* 'print': lib option '--ten' unmatched in ref
* 'probit' appears to have 5 options in lib, 4 in manual
* 'probit': lib option '--quiet' unmatched in ref
* 'quantreg' appears to have 4 options in lib, 3 in manual
* 'quantreg': lib option '--no-df-corr' unmatched in ref
* 'quit' appears to have 1 options in lib, 0 in manual
* 'quit': lib option '--exit' unmatched in ref
* 'restrict' appears to have 7 options in lib, 0 in manual
* 'restrict': lib option '--bootstrap' unmatched in ref
* 'restrict': lib option '--full' unmatched in ref
* 'restrict': lib option '--jitter' unmatched in ref
* 'restrict': lib option '--quiet' unmatched in ref
* 'restrict': lib option '--verbose' unmatched in ref
* 'restrict': lib option '--lbfgs' unmatched in ref
* 'restrict': lib option '--no-scaling' unmatched in ref
* 'setobs' appears to have 7 options in lib, 5 in manual
* 'setobs': lib option '--restructure' unmatched in ref
* 'setobs': lib option '--special-time-series' unmatched in ref
* 'smpl' appears to have 6 options in lib, 0 in manual
* 'smpl': lib option '--full' unmatched in ref
* 'smpl': lib option '--dummy' unmatched in ref
* 'smpl': lib option '--no-missing' unmatched in ref
* 'smpl': lib option '--random' unmatched in ref
* 'smpl': lib option '--replace' unmatched in ref
* 'smpl': lib option '--restrict' unmatched in ref
* 'system' appears to have 2 options in lib, 0 in manual
* 'system': lib option '--iterate' unmatched in ref
* 'system': lib option '--verbose' unmatched in ref
* 'tabprint' appears to have 2 options in lib, 3 in manual
* 'tabprint': ref option '--format="f1|f2|f3|f4"' unmatched in lib
* 'tsls' appears to have 4 options in lib, 2 in manual
* 'tsls': lib option '--quiet' unmatched in ref
* 'tsls': lib option '--save' unmatched in ref
* 'var' appears to have 8 options in lib, 7 in manual
* 'var': lib option '--quiet' unmatched in ref
* 'vecm' appears to have 9 options in lib, 7 in manual
* 'vecm': lib option '--quiet' unmatched in ref
* 'vecm': lib option '--verbose' unmatched in ref
* 'xcorrgm' appears to have 1 options in lib, 0 in manual
* 'xcorrgm': lib option '--quiet' unmatched in ref
gretl_commands.xml:
library commands missing from reference: 4
gretl_commands.xml:
extra commands in ref but not in library: 0
Found 121 commands in 'gretl_commands_pt.xml'
* libgretl command 'error' is not in the reference
* libgretl command 'omitfrom' is not in the reference
* libgretl command '' is not in the reference
* libgretl command '' is not in the reference
* 'adf' appears to have 10 options in lib, 9 in manual
* 'adf': lib option '--gls' unmatched in ref
* 'append' appears to have 2 options in lib, 0 in manual
* 'append': lib option '--time-series' unmatched in ref
* 'append': lib option '--quiet' unmatched in ref
* 'ar1' appears to have 5 options in lib, 4 in manual
* 'ar1': lib option '--quiet' unmatched in ref
* 'arbond' appears to have 5 options in lib, 4 in manual
* 'arbond': lib option '--orthdev' unmatched in ref
* 'arima' appears to have 8 options in lib, 5 in manual
* 'arima': lib option '--opg' unmatched in ref
* 'arima': lib option '--hessian' unmatched in ref
* 'arima': lib option '--quiet' unmatched in ref
* 'chow' appears to have 1 options in lib, 0 in manual
* 'chow': lib option '--quiet' unmatched in ref
* 'corr' appears to have 4 options in lib, 0 in manual
* 'corr': lib option '--kendall' unmatched in ref
* 'corr': lib option '--spearman' unmatched in ref
* 'corr': lib option '--uniform' unmatched in ref
* 'corr': lib option '--verbose' unmatched in ref
* 'corrgm' appears to have 1 options in lib, 0 in manual
* 'corrgm': lib option '--quiet' unmatched in ref
* 'cusum' appears to have 2 options in lib, 1 in manual
* 'cusum': lib option '--quiet' unmatched in ref
* 'data' appears to have 1 options in lib, 0 in manual
* 'data': lib option '--odbc' unmatched in ref
* 'dataset' appears to have 1 options in lib, 0 in manual
* 'dataset': lib option '--preserve' unmatched in ref
* 'delete' appears to have 1 options in lib, 0 in manual
* 'delete': lib option '--db' unmatched in ref
* 'eqnprint' appears to have 2 options in lib, 1 in manual
* 'eqnprint': lib option '--t-ratios' unmatched in ref
* 'estimate' appears to have 5 options in lib, 3 in manual
* 'estimate': lib option '--quiet' unmatched in ref
* 'estimate': lib option '--verbose' unmatched in ref
* 'fcast' appears to have 4 options in lib, 2 in manual
* 'fcast': lib option '--quiet' unmatched in ref
* 'fcast': lib option '--out-of-sample' unmatched in ref
* 'freq' appears to have 4 options in lib, 2 in manual
* 'freq': lib option '--silent' unmatched in ref
* 'freq': lib option '--normal' unmatched in ref
* 'garch' appears to have 6 options in lib, 4 in manual
* 'garch': lib option '--fcp' unmatched in ref
* 'garch': lib option '--nc' unmatched in ref
* 'gmm' appears to have 5 options in lib, 4 in manual
* 'gmm': lib option '--quiet' unmatched in ref
* 'gnuplot' appears to have 10 options in lib, 8 in manual
* 'gnuplot': ref option '--suppress-fit' unmatched in lib
* 'gnuplot': lib option '--suppress-fitted' unmatched in ref
* 'gnuplot': lib option '--time-series' unmatched in ref
* 'gnuplot': lib option '--control' unmatched in ref
* 'heckit' appears to have 4 options in lib, 3 in manual
* 'heckit': lib option '--ml' unmatched in ref
* 'help' appears to have 1 options in lib, 0 in manual
* 'help': lib option '--func' unmatched in ref
* 'hsk' appears to have 2 options in lib, 1 in manual
* 'hsk': lib option '--quiet' unmatched in ref
* 'lmtest' appears to have 10 options in lib, 7 in manual
* 'lmtest': lib option '--breusch-pagan' unmatched in ref
* 'lmtest': lib option '--robust' unmatched in ref
* 'lmtest': lib option '--white-nocross' unmatched in ref
* 'logit' appears to have 5 options in lib, 3 in manual
* 'logit': lib option '--p-values' unmatched in ref
* 'logit': lib option '--quiet' unmatched in ref
* 'mle' appears to have 5 options in lib, 4 in manual
* 'mle': lib option '--quiet' unmatched in ref
* 'mpols' appears to have 4 options in lib, 3 in manual
* 'nls' appears to have 5 options in lib, 3 in manual
* 'nls': lib option '--quiet' unmatched in ref
* 'nulldata' appears to have 1 options in lib, 0 in manual
* 'nulldata': lib option '--preserve' unmatched in ref
* 'omit' appears to have 8 options in lib, 6 in manual
* 'omit': lib option '--auto' unmatched in ref
* 'omit': lib option '--bootstrap' unmatched in ref
* 'open' appears to have 5 options in lib, 1 in manual
* 'open': lib option '--coded' unmatched in ref
* 'open': lib option '--odbc' unmatched in ref
* 'open': lib option '--preserve' unmatched in ref
* 'open': lib option '--quiet' unmatched in ref
* 'panel' appears to have 14 options in lib, 8 in manual
* 'panel': lib option '--fixed-effects' unmatched in ref
* 'panel': lib option '--hausman-reg' unmatched in ref
* 'panel': lib option '--pooled' unmatched in ref
* 'panel': lib option '--robust' unmatched in ref
* 'panel': lib option '--silent' unmatched in ref
* 'pca' appears to have 3 options in lib, 2 in manual
* 'pca': lib option '--covariance' unmatched in ref
* 'print' appears to have 4 options in lib, 3 in manual
* 'print': lib option '--ten' unmatched in ref
* 'probit' appears to have 5 options in lib, 3 in manual
* 'probit': lib option '--p-values' unmatched in ref
* 'probit': lib option '--quiet' unmatched in ref
* 'quantreg' appears to have 4 options in lib, 3 in manual
* 'quantreg': lib option '--no-df-corr' unmatched in ref
* 'quit' appears to have 1 options in lib, 0 in manual
* 'quit': lib option '--exit' unmatched in ref
* 'reset' appears to have 3 options in lib, 0 in manual
* 'reset': lib option '--cubes-only' unmatched in ref
* 'reset': lib option '--quiet' unmatched in ref
* 'reset': lib option '--squares-only' unmatched in ref
* 'restrict' appears to have 7 options in lib, 0 in manual
* 'restrict': lib option '--bootstrap' unmatched in ref
* 'restrict': lib option '--full' unmatched in ref
* 'restrict': lib option '--jitter' unmatched in ref
* 'restrict': lib option '--quiet' unmatched in ref
* 'restrict': lib option '--verbose' unmatched in ref
* 'restrict': lib option '--lbfgs' unmatched in ref
* 'restrict': lib option '--no-scaling' unmatched in ref
* 'runs' appears to have 2 options in lib, 0 in manual
* 'runs': lib option '--difference' unmatched in ref
* 'runs': lib option '--equal' unmatched in ref
* 'setobs' appears to have 7 options in lib, 5 in manual
* 'setobs': lib option '--restructure' unmatched in ref
* 'setobs': lib option '--special-time-series' unmatched in ref
* 'smpl' appears to have 6 options in lib, 0 in manual
* 'smpl': lib option '--full' unmatched in ref
* 'smpl': lib option '--dummy' unmatched in ref
* 'smpl': lib option '--no-missing' unmatched in ref
* 'smpl': lib option '--random' unmatched in ref
* 'smpl': lib option '--replace' unmatched in ref
* 'smpl': lib option '--restrict' unmatched in ref
* 'system' appears to have 2 options in lib, 0 in manual
* 'system': lib option '--iterate' unmatched in ref
* 'system': lib option '--verbose' unmatched in ref
* 'tabprint': ref option '--format="f1|f2|f3|f4"' unmatched in lib
* 'tabprint': lib option '--rtf' unmatched in ref
* 'tsls' appears to have 4 options in lib, 2 in manual
* 'tsls': lib option '--quiet' unmatched in ref
* 'tsls': lib option '--save' unmatched in ref
* 'var' appears to have 8 options in lib, 6 in manual
* 'var': lib option '--quiet' unmatched in ref
* 'var': lib option '--trend' unmatched in ref
* 'vecm' appears to have 9 options in lib, 7 in manual
* 'vecm': lib option '--quiet' unmatched in ref
* 'vecm': lib option '--verbose' unmatched in ref
* 'xcorrgm' appears to have 1 options in lib, 0 in manual
* 'xcorrgm': lib option '--quiet' unmatched in ref
gretl_commands_pt.xml:
library commands missing from reference: 4
gretl_commands_pt.xml:
extra commands in ref but not in library: 0
17 years
Request for help troubleshooting LaTex for gretl-ref-pt.pdf.
by Hélio Guilherme
Dear Allin, Jack, et al.,
As you most certainly know there is now in CVS the LaTex structure for
the Portuguese translations of the manuals.
This translation can now speed up, because Henrique offered his help.
I did some updates to the files, and now I cannot obtain the
gretl-ref-pt.pdf properly generated. It is missing the table of contents.
The LaTex script stops and waits for user input.
If someone could help me to troubleshoot this, I would appreciate very
much :D.
Thanks,
Hélio Guilherme
17 years
Latex in Gretl not working
by abdilahi osman hussein
The Latex menu in my gretl is not working. I uninstalled and installed the
windows version several times with the same result. Can someone help with
this?
17 years, 1 month
Using the smpl command in functions
by Gordon Hughes
At the beginning of August I reported a problem that I had
encountered with use of the smpl and setobs commands inside
functions. From the change log for Version 1.7.7 (items 4 & 5) I had
understood that this problem had been corrected in the new version,
but I have not had time to test it before now. Sadly, my test
scripts still fail in the same way as before.
To recap briefly. My function is designed to deal with unbalanced
panel data, whereas Gretl's standard panel functions insist on
balanced panel data. It proceeds as follows: (a) check that the data
supplied is panel data, (b) extract the panel & time ids to new
variables, (c) define the data as cross section and drop cases with
missing values, (d) carry out the function estimation, (e) attempt to
reset sample to the full data supplied (smpl full), (f) redefine the
data as panel data using the panel & time ids extracted in step (b).
This works as intended when it is supplied with my full test
dataset. However, the program falls over in this final step,
reporting that missing values were encountered, when the sample is
restricted before the function is called. The reason is that the
full data contains 91 panels containing up to 14 time periods in each
panel with 1274 cases in total. Before executing the function I have
dropped all panels with panel_id > 9 (smpl id < 10
--restrict). However, after the "smpl full" command in the function
is run the dataset has reverted to 1274 cases instead of the 126
cases supplied - the first 126 cases correspond to the original cases
but cases 127 onwards consist entirely of missing values because the
function does not know about them. I may be able to deal with this
by dropping cases with missing values for the panel & time ids, but I
understand that "smpl full" should not work in this way.
I can provide my test programs and data but I have had problems in
uploading Zip archives in the past so I would prefer to send them
direct to someone's e-mail address rather than to the mailing list.
Gordon
17 years, 1 month
Compiling gretl on Windows
by andreas.rosenblad＠ltv.se
I have started learning C in order to be able to help in the development of
gretl. My earlier programming experinces are mostly MATLAB, R, GAUSS and
FORTRAN. I am working on Windows, and wants to compile and develop gretl on
Windows.
Well, now I want all useful help to get this project working. What are your
experiences on compiling gretl on Windows?
Which compiler is best to use? MinGW? Or is it possible to use the
Microsoft Visual C++ compiler?
Right now I am in the beginning, trying to figure out which IDE is best to
use. I am choosing between Code::Blocks, Dev-C++, wxDev-C++ and Microsoft
Visual C++ 2008 Express Edition.
Any suggestion about which to use?
Dev-C++ and wxDev-C++ seems to have integrated support for CVS, which
should be good when working with gretl. But it looks like they are not in
active development anymore. Code::Blocks is free, cross-platform, in active
development, and supports multiple compilers, including MinGW and Microsoft
Visual C++. On the other hand, Microsoft Visual C++ 2008 Express Edition
seems to have more help files and many other good features.
Best regards
Andreas
17 years, 1 month
new formatting of models
by Allin Cottrell
In CVS and the Windows snapshot I've switched to a new format for
the plain text output of models. It's not radically different,
but avoids the situation where you'd get a lot of white space
between columns. (Now we calculate the column widths in advance
rather than always allowing for the "worst case".)
Comments welcome, also any reports of cases where the new
formatting screws up.
Allin.
17 years, 1 month
