A couple of things to note in current CVS and the current
1) I have "rationalized" the syntax for the "restrict" command:
all coefficient arrays are now accessed using a 1-based index
2) I've added support for reading PcGive databases in the form
of in7/bn7 files. This is under the Databases item in the File
menu. I haven't seen a public specification of in7/bn7 format
so I've stuck to what I've been able to infer from examples.
That means that at present only annual, quarterly and monthly
data are supported. If anyone knows how, e.g. daily data are
handled in these files, please let me know.
Translators: The IRO robot seems to be "on holiday" -- at
least, I've had no response to my last submission of a gretl.pot
file. The current .pot (last updated 2006/08/03) is available
I'm happy to accept .po updates via CVS or via email.
a strange thing just happened to me while looking at some data in gretl:
After hitting some key combinations (wrong ones), gretl suddenly started
reading the numbers -- I mean, literally reading, with an American
accent, in English, over the loudspeaker.
And closing gretl didn't stop it! I had to kill the gretlw32.exe process
(on Windows, obviously) in the task manager...
That was really spooky...
What's behind this?
I was wondering if it was possible, when constructing loops, to have 2 or
several indexes in order to differenciate between different sets of
variables. For instance, is it possible to run OLS estimation repetitively
taking each time a different dependent variable indexed by i, and a
different independent variable indexed by j in another list of variables ?
This would amount to having a loop on j within a loop concerning i.
The supplementary material for our forthcoming paper titled "GRETL
1.6.0 and Its Numerical Accuracy" is made available in the Journal of
Applied Econometrics Data Archive at the following url:
This supplement explains the testing procedure of gretl in terms of
numerical accuracy and also provides several tables that compare
gretl's performance with the published results of various commercial
econometric programs. It is expected to be published next year.
A positive attitude may not solve all your problems, but it will annoy
enough people to make it worth the effort. - Herm Allbright
Thanks. Here is an example script: if I run it twice successively, the
estimates change as well as the size of the sample.
Your suggestion to introduce missing values seems to have stabilized things
(the problem seems to have occured though, but only once in several tests.)