Re: [Gretl-users] IRFs and CI plots
by Riccardo (Jack) Lucchetti
On Wed, 13 Mar 2013, Gabriela Nodari wrote:
> Dear Riccardo
>
> I guess there is a problem in gretl plot controls when using the svar
> package.
> After the command IRFplot, when I try to modify the default plot (like
> colors, etc.) it does not work very well.. For example if you do nothing
> and just click on apply in the plot controls windows.. The confidence
> interval turns blue!!!
> Could be a gretl problem or my pc?
>
> Thank you in advance!!
> Gabriela
The problem is not easy to solve, but there's a way: let me add a boring
but short preface. All gretl's graphics are handled by creating a gnuplot
script, executing it and then sending the result to the display. All this
is done transparently. When you edit a graph, you modify the underlying
gnuplot script via some GUI elements, so when you click "Apply" the
graphic gets re-generated. This is where the problem is: gretl's GUI
interface for modifying graphics can't handle arbitrary gnuplot scripts
(nor we want it to: we're not interested in building a GUI interface to
gnuplot, which is a vast and complex language), but only those generated
internally.
The figures generated by the SVAR package contain a few extra features
that the GUI editor doesn't "understand", so if you invoke the GUI
controls, it messes up the graph. However, you can customise the graph by
editing the gnuplot script directly. Here's how: right-click on it and
"Save [it] to session as icon". Then, in the icon view, right click on the
graph icon and choose "Edit plot commands": you'll have the gnuplot source
to the graph, that you can customise in any way you want. Of course, this
entails some knowledge of gnuplot. However, if you click on the help icon
while editing a gnuplot script the gnuplot manual in pdf will be
downloaded and opened for you.
PS: I hope you don't mind if I send this message to the user-list too, as
others may want to ask the same question.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
13 years
Gretl for Mac OS X 10.8.2
by Daniel Veiga de Carvalho
Hi eveybody,
I am trying to download gretl for my mac, but i think gretl doesn`t run
with Xquartz, so i need to download X11. Someone knows a version of X11
that works normally at Mac OS X 10.8.2 because the version that i tried
doesn`t work also.
Thanks.
--
Daniel Carvalho
13 years
Help on compiling gretl in ubuntu
by yinung＠Gmail
Dear all,
I just wanted to compile gretl in my ubuntu 10.04 after these commands in
terminal
cvs update -d -P
./configure
But I got following messages:
~/gretl/cvs/gretl-old$ make
./builddate
build.h is current
make -C lib
make[1]: Entering directory `/home/yinung/gretl/cvs/gretl-old/lib'
make[1]: *** No rule to make target `../lib/src/gretl_scalar.h', needed by
`compat.lo'. Stop.
make[1]: Leaving directory `/home/yinung/gretl/cvs/gretl-old/lib'
make: *** [lib] Error 2
Any one can help, please?
Yi-Nung Yang
13 years
possible bug in cvs (and ods?)
by Leandro Zipitria
Dear Gretl community:
I have problems loading a database in cvs format. The error display
"non-numercial value" when indeed its a numerical value. Could this be a
bug, or just a problem of observation format or conversion?
Also, I have a second question related to loading databases. When I try to
load a dataset with strings in CVS I have no problem, but when I try to
load the same dataset in ods or xls format I found the following error
"Problem in column X: numerical value expected and found string".
Why is there such different behaviour for the same database? If strings are
possible in CVS why not in different format?
Regards,
Leandro
13 years
Structural shocks VAR
by Gabriela Nodari
Dear all,
I would like to know if there is a way to get the residuals from a SVAR
model.
In particular, I would like to save (as a time series) the structural
residuals from
just one equation, since I am interested in only one structural shock, in
order to
make some tests on the "innovations". But I only know how to estimate the
reduced
form VAR and then compute IRFs and FEVDs.
Thank you for your time and attention,
Gabriela
13 years
Fixed effect logit for panel datasets
by Riccardo (Jack) Lucchetti
I uploaded earlier today a function package for estimating fixed-effects
logit models via conditional ML. If people could test it, I'd appreciate
it. Thanks!
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
13 years
isnull for matrices
by Pindar
Hi,
I encountered that a matrix that 'is null' cannot be filtered by 'isnull':
<hanls>
matrix test = {1,0,3;1,1,100}
eval isnull(test)
matrix seek = selifr(test,test[,1].=1)
matrix seek = selifr(seek,seek[,2].=2)
eval seek
eval isnull(seek)
# and also not working
eval ok(test)
eval ok(seek)
<hanls>
Best
Leon
13 years
Starting Gretl from another program
by Data Analytics Corp.
Hi,
I'd like to start Gretl from another program. In particular, I'd like
to start Gretl from a JMP script, send it a data file in CSV format, and
have the Gretl GUI appear with that data loaded. No Gretl scripts are
involved; I just want the data loaded and ready for analysis. I'm aware
of gretlcli but this seems to start gretl, execute a script file, and
then send the results to an output file. I just want to start it, have
it read a CSV file, and show the data in the GUI. Can this be done?
And if so, how?
Thanks for any help,
Walt
________________________
Walter R. Paczkowski, Ph.D.
Data Analytics Corp.
44 Hamilton Lane
Plainsboro, NJ 08536
________________________
(V) 609-936-8999
(F) 609-936-3733
walt(a)dataanalyticscorp.com
www.dataanalyticscorp.com
_____________________________________________________
13 years
