Am 16.06.2019 um 16:34 schrieb Logan Kelly:
> Hello all,
> I am using the launch command in a loop to call out to a processor
> intensive estimation routine ( I am calling an exe file). My system only
> has 12 processor cores, so is there a way to limit the number of
> simultaneous processes to 12? Or should I just let windows handle it.
It's probably a good idea to not throw more than 12 tasks at the system
simultaneously, but if you're calling an exe via the shell, it's not
clear to me why you want to control everything from gretl (as opposed to
the shell itself).
If you want proper control from within gretl for multiprocessing,
perhaps you should look at the MPI setup (see the gretl + MPI guide
under the help menu).
I am using the launch command in a loop to call out to a processor intensive estimation routine ( I am calling an exe file). My system only has 12 processor cores, so is there a way to limit the number of simultaneous processes to 12? Or should I just let windows handle it. Here are the specs of my setup:
OS: Win 10 pro 64 bit
grelt: latest git as of 2019-06-16
CPU's: 2 x Hex-Core Intel Xeon
P.S. I apologize if this is too far off topic.
> Have a look at the attached script.
> Notable points:
> * it seems that the original income variable has to be rescaled by a=20
> number close to 10000 to match Greene's results
> * we don't do RE logit, maybe someone could write a package to do this,=20
> but I don't think they're in high demand
> * in principle, you can obtain unconditional FE logit by adding=20
> "dummify(id)" to your list of explanatory variables and dropping all unit=
> for which the dependent variable is time-invariant, but it's very=20
> computationally demanding and the estimator is inconsistent anyway, so wh=
> bother? (It's in the script, though)
> * the function package "felogit" (by yours truly) does conditional FE=20
> logit; download and enjoy
> * the --random switch to "probit" gives you the ML through Gaussian=20
> quadrature; maybe you can play with the --quadrature option
> * again, FE probit is inconsistent, so nobody uses it.
> Riccardo (Jack) Lucchetti
> Dipartimento di Scienze Economiche e Sociali (DiSES)
> Universit=C3=A0 Politecnica delle Marche
> (formerly known as Universit=C3=A0 di Ancona)
Yes, thank you so much, Jack. It does the job as advertised.
I just want to show how this is done to my students. I learned a whole lot in the process.
A minor issue with the interface, however.
When I installed the package felogit, there was a dialog box that says:
“The package felogit can be attached to the gretl menus
as ‘/menubar/Model/LimdepModels/logit/Fixed-effects logit’ in the main window.
Do you want to do this?”
My response was yes, and yet, there is no such menu item showing up.
What might be the cause of this?
I’m on a MacBook Pro, OS X 10.11.6, with gretl version 2019c-git
I was able to replicate Green’s version 6 table 23.5 pooled legit or probit using the "German Health Care Data", but not random effect logit, not fixed effect logit, nor random effect probit.
For the fixed effect legit, the error message reads: “Illegal non-positive value of the dependent variable”.
There is no random effect option for logit, it seems.
My random effect probit differ from green’s table.
What should I do?
especially those interested in panel cointegration analysis, the
KaoTest.gfn package has certain issues and will be removed from the
gretl package server soon. The author couldn't be reached to address
The good news is that work is ongoing to start a new package for panel
cointegration testing. The plan is that it will offer more tests than
the KaoTest package.
Nevertheless, if you are a user of that package and need it for your
work please keep a local copy of it to be on the safe side.
let me do some self-advertising, please: I've published an article
entitled "Practical Empirical Research Using gretl and hansl" in the
Australian Economic Review in their "student section". The journal is
hosted at the University of Melbourne. Some econometrics professors and
teachers were thinking of using gretl instead of (or jointly with) R or
Eviews back in 2017/18 -- not sure whether any decision has been made on
this already. But nevertheless, maybe the article is of interest to some
Data and replication files are available on my github repo (details can
be found in the paper).
Dear Gretl Community,
Please let me pose the following issue: the GUI does not seem to display the date on which script files were last modified. Am I right?
If I go to "File" -> "Script files" -> "User file...", then I just can see 3 columns: "Name", "Size" and "Modified". But the last column only shows time, but not date, which I think would be useful. Is this a problem with my machine (running under Windows)? Is this what it should be? Am I just looking at the wrong place?
Thanks in advance.
Dear users of collinearity analyses,
the function package bkw.gfn by Lee Adkins is going to be retired soon.
The reason is that its functionality has been integrated into gretl
itself via the new 'bkw' command and 'bkw()' function. See the
documentation in gretl 2019b (excellent as always).
Please speak up if you need a little transition period. Apart from that,
you can of course always keep a local copy of the package for your own use.
I noticed a line of code that I didn’t know was possible from your response earlier:
"ols lwage 0 log(wage) —simple”
I can’t seem to find reference to what is acceptable and what is not on the ols command line.
Here you’ve invoked a function on the command line, i.e. log(wage), not a existing series’ name.
Other transformation is also possible, such as ols y 0 y(-1 to 4) that I’ve noticed.
This raises a question about the nature of the indepvars in the ols documentation.
What other command line transformation that is possible without creating a new variable first?
Where are they documented?
Thanks in advance.