Re: [Gretl-users] Recode variable
by Allin Cottrell
On Thu, 16 Dec 2010, Laurent Schuermans wrote:
> I have a bunch of dichotomious variables with categories 1 and 2
> that I want to recode to 0 and 1, which is easier for
> interpretation, but I don't know how to do this.
If you have a series y containing just 1s and 2s, and you want the
1s to become zeros and the 2s to become 1s then just subtract 1:
y -= 1
Take a look at the "dummify" command for more complex recoding
needs.
Allin Cottrell
14 years
Recode variable
by Laurent Schuermans
Dear all,
In econometric classen, we use gretl along with the Principles of
Econometrics handbook. Now we have to write a paper and do research about a
subject of own interest while using gretl.
The problem however is that I don't know how to recode variables in gretl. I
have a bunch of dichotomious variables with categories 1 and 2 that I want
to recode to 0 and 1, which is easier for interpretation, but I don't know
how to do this.
Can somebody help me out?
Kind Regards,
Laurent Schuermans
Master Economics
University of Antwerp
14 years
Re: [Gretl-users] breusch-godfrey test of autocorrelation
by Allin Cottrell
On Tue, 14 Dec 2010, Samrat Sanyal wrote:
> Thanks for the help.I am basically looking for a suitable
> autocorrelation test in lagged independent variable-included
> regression for a few financial time series,so the first 'general' test
> that came to me was breusch godfrey.are there any other tests
> available in gretl for the purpose?
The test available in gretl is indeed Breusch-Godfrey. However,
the test statistic that we report first in relation to this test
is Kiviet's LMF, which has better small-sample properties that
TR^2.
Allin Cottrell
14 years
Re: [Gretl-users] breusch-godfrey test of autocorrelation
by Allin Cottrell
On Mon, 13 Dec 2010, Sven Schreiber wrote:
> Am 13.12.2010 07:48, schrieb Samrat Sanyal:
> > Gretl users,
> >
> > Is there any option to execute breusch-godfrey autocorrelation test in
> > gretl?
> >
>
> check out the documentation for the 'modtest --autocorr' command. Do you
> need "just" a valid test for autocorrelation, or do you really need
> breusch-godfrey?
Also, in the graphical interface after estimating a model, see
"Autocorrelation" under the tests menu.
Allin Cottrell
14 years
define range
by Bernadette Scully
Hello
Using the menu driven regression page, a pop up question asked for the
forecast range. I entered a range which was fine and then wished to
lengthen the range, which it did not allow. Can anyone direct me to the
drop down menu that allows the range to be extended. I've tried
Forecast - range boxes are there but not changeable.
Grateful for help, Bernadette
<P><p><font face="Arial, Helvetica, sans-serif" size="2" style="font-size:13.5px">_______________________________________________________________<BR>Care2 makes it easy for everyone to live a healthy, green lifestyle
and impact the causes you care about most. Over 12 Million members!
http://www.care2.com
Feed a child by searching the web! Learn how http://www.care2.com/toolbar</font>
14 years
Re: [Gretl-users] cochrane orcutt/Prais-winsten models
by Allin Cottrell
On Mon, 13 Dec 2010, Sven Schreiber wrote:
> Am 13.12.2010 08:36, schrieb Samrat Sanyal:
> > Gretl users,
> > I have another query.Is there a way out to execute a cochrane
> > orcutt/Prais-winsten model in gretl? I basically have to run a
> > regression where i have a financial time series variable dependent on
> > lagged financial variables.
> >
>
> It's right in the menus, did you actually look for it?
Under /Model/Time series. For this to be active your data have to
be recognized as time series. If that's not the case, use the
menu item /Data/Dataset structure.
Allin Cottrell
14 years
cochrane orcutt/Prais-winsten models
by Samrat Sanyal
Gretl users,
I have another query.Is there a way out to execute a cochrane
orcutt/Prais-winsten model in gretl? I basically have to run a regression
where i have a financial time series variable dependent on lagged financial
variables.
--
Thanks and Regards,
Samrat Sanyal
14 years
Calculating Critical Values of Statistical Distributions
by Henrique Andrade
Dear Gretl Community,
Is there a way to calculate critical values of the t statistic using the
command line interface? I'm asking this because I would like to insert this
kind of calculation inside some scripts.
Thanks in advance,
--
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
14 years
Re: [Gretl-users] Gretl-users Digest, Vol 47, Issue 12
by Robert Kirkby
On Wed, 8 Dec 2010, Allin Cottrell wrote
>
> On Tue, 7 Dec 2010, Robert Kirkby wrote:
>
> [ that he'd like to be able to retrieve the p-values from
> gretl's Johansen cointegration test programatically ]
>
> This is now implemented in CVS and snapshots. The $test and
> $pvalue accessors will give you matrices with the same layout as
> the printed results; that is, column 1 of each matrix pertains to
> the trace test and column 2 to the Lambda-max test.
>
> Allin Cottrell
>
Thanks for the quick response. I guess I will just wait till this
makes the next stable version then. Thanks again.
Robert
14 years
