Questions re: Chow tests
by James Wells
I'm working with time series data (annual from 1959 to 2004) and
regressing my dependent variable on five explanatories, plus one lagged
values of the dependent variable. When I run a Chow test for structural
break, it creates only three "sd_" variables. How come?
Also, are there any plans to implement a Quandt likelihood ratio test in
gretl?
Thanks,
James
18 years, 10 months
Re:Re: [Gretl-users] Installing gretl from source
by Talha Yalta
> From: Allin Cottrell <cottrell(a)wfu.edu>
> To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu>
> Date: Fri, 9 Jun 2006 13:34:48 -0400 (EDT)
> Subject: Re: [Gretl-users] Installing gretl from source
> On Fri, 9 Jun 2006, Talha Yalta wrote:
>
> > I was fed up with all the windows stuck inside a single desktop
> > so I decided to try installing gretl snapshot in my newly
> > installed Kubuntu Dapper system. I am not very experienced in
> > compiling software on my own and here is what happened:
> >
> > After running ./configure everything goes fine except one error message:
> >
> > *** The gtk-config script installed by GTK could not be found
> > *** If GTK was installed in PREFIX, make sure PREFIX/bin is in
> > *** your path, or set the GTK_CONFIG environment variable to the
> > *** full path to gtk-config.
>
> Hmm. The "gtk-config" script is the configuration mechanism for
> gtk version 1.2, which is very old by now. On a reasonably
> up-to-date system one would expect to find gtk 2.0, and it does
> its configuration using the pkg-config program.
>
> I suspect that gretl's "configure" is falling back on gtk-1.2
> because it can't find the relevant gtk-2.0 files (and then
> failing, of course).
>
> I recommend trying this command:
>
> pkg-config --libs gtk+-2.0
>
> (note the slightly odd string "gtk+-2.0").
>
> If this command doesn't succeed, the "devel" packages for gtk 2.0
> are presumably not installed. They probably shouldn't be hard to
> find.
>
> Congratulations on joining the compiling club! I trust the
> initial problems should be sorted out without too much grief.
>
> Allin Cottrell
>
>
Well, turns out that libgtk-devel was the missing package. After
running the ./configure command maybe 20 times, I was finally able to
locate and install all the dependencies one by one so that now I have
a working gretl with everything (latex etc.) :) I later realized that
there is also a list of dependencies in gretl website. I learn
something every day :)
There is just one problem though: I realized that the source code that
I downloaded is for the official version 1.5.1. While this is newer
than the package offered in ubuntu repos, it is not the latest
snapshot that I wanted. How do I download the tar.bz2 file for the
latest development version?
--
Don't go around saying the world owes you a living. The world owes you
nothing. It was here first. - Mark Twain
--
18 years, 10 months
HP-filter
by Daniel
Hi,
Im a novice at this so my question might be kind of simple but does anyone
know how to apply the HP-filter in gretl to alot of time series. I have like
200 series to filter so doing it manually is pretty time consuming...
Thanks!
/Daniel
18 years, 10 months
Re: [Gretl-users] Installing gretl from source
by Peter Davidoff
Know that GRETL can be installed from the synaptic package manager
for Ubuntu Dapper, version 1.5.0-2. It is in the Mathematics (universe)
repository. Less fun then installing from the source...
Peter
At 10:49 AM 6/9/2006, you wrote:
>I was fed up with all the windows stuck inside a single desktop so I
>decided to try installing gretl snapshot in my newly installed Kubuntu
>Dapper system. I am not very experienced in compiling software on my
>own and here is what happened:
18 years, 10 months
Installing gretl from source
by Talha Yalta
I was fed up with all the windows stuck inside a single desktop so I
decided to try installing gretl snapshot in my newly installed Kubuntu
Dapper system. I am not very experienced in compiling software on my
own and here is what happened:
After running ./configure everything goes fine except one error message:
*** The gtk-config script installed by GTK could not be found
*** If GTK was installed in PREFIX, make sure PREFIX/bin is in
*** your path, or set the GTK_CONFIG environment variable to the
*** full path to gtk-config.
I tried to locate this file using the "locate gtk-config" command and
apparently it is not available in the system. I know I have several
gtk based applications installed in my system such as firefox and
gnumeric. Is there anyone who knows which package(s) need to get
installed so that I have gtk-config?
Ps: Alternatively, I know I can try installing using ./configure
--without-gnome but I didn't do this not knowing what functionality I
will be missing. Thanks for any directions.
Sincerely,
A. Talha YALTA
--
Don't go around saying the world owes you a living. The world owes you
nothing. It was here first. - Mark Twain
--
18 years, 10 months
variance in mean test
by Sven Schreiber
Well, as you can see from my recent posts I have been using the test
statistic calculator for the last couple of days. Here's another
question about it:
When testing the difference of two means (Utilities -> test statistic
calculator), I cannot reproduce the standard deviation used in the
denominator. For example, using
mu1=1, sigma1=2, n1=25,
mu2=1, sigma2=3, n2=36
gretl's test statistic output is 0/0.687129. What's the formula for that
value? I would use \sqrt(4/25 + 9/36) = 0.6403... in the denominator
instead, and I couldn't get gretl's ouput by some variations.
Thanks for your help,
Sven
18 years, 10 months
Estimating ARMA model with GARCH errors
by sudip mukherjee
Hi,
I want to know how can I estimate the parameters of an
ARMA(1,1) model with Garch(1,1) errors. I believe all
the parameters will need to be estimate
simultaneously.
Any help will be greatly appreciated.
Thanks
Sudip
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
18 years, 10 months
translation issues
by Sven Schreiber
Hi,
some of the following issues may even be of interest for other
translators as well:
The dialog string "There are no new files on the server right now."
doesn't seem to exist in the .po file.
Same with "no user-defined functions...", but I guess that's just too new.
A different thing is happening with strings from the frequency
distribution results: "Frequency distribution for...", "interval,
midpt...", are in the .po file and there exist German translations, but
they're still showing up in English (latest windows snapshot). But "rel.
cum." seems missing again.
Thanks,
Sven
18 years, 10 months
proportion test bug
by Sven Schreiber
When using the test statistic calculator to perform a proportion test
and requesting a graph of the test distribution, there's an error
message "bad png header..." and a black window. gretl 1.5.1 as well as
latest snapshot on windows.
cheers,
sven
18 years, 10 months
I think I found a bug with mpols and genr
by Talha Yalta
This happens in gretl 1.5.1 and the latest snapshot (6.1.2006),
(windows version):
1- First, I run a high precision regression: mpols y 0 x
2- I run: genr cfx = $coeff(x) to get the internal variable and
command returns no comments (no errors or messages saying scalar cfx
generated)
3- Consequently, running "print cfx" returns an error saying no
variable named cfx.
4- Next, I run a regular ols this time: ols y 0 x
5- Running genr cfx = $coeff(x) creates the scalar OK this time and I
can see the value using the "print cfx" command.
6- Now, I run the initial high precision regression once again: mpols
y 0 x and suprisingly running genr cfx = $coeff(x) suddenly starts
working by giving the message "replaced scalar cfx ..."
7- Finally, after running a mpols and running genr to get $uhat,
$yhat, $stderr, $rho, $vcv also return no feedback (as described
above) while $T, $rsq, $ess, $df, $ncoeff, $trsq, $sigma, $aic, $bic,
$lnl return "The statistic you requested is not available. On the
other hand $nobs, $pd, $nvars, $test, $pvalue seem to work OK with the
genr command after running the mpols.
--
Don't go around saying the world owes you a living. The world owes you
nothing. It was here first. - Mark Twain
--
18 years, 10 months
