1:36 p.m.
Dear Gretl users,
I was trying to retrieve some data from an estimated model but I'm not sure that's
possible.
An example:
<hansl>
open australia.gdt
# simple ols model
ols E 0 PAU PUS
# scalars for the coefficients
coef1 = $coeff[1]
coef2 = $coeff[2]
# scalars for the standard errors
stderr1 = $stderr[1]
stderr2 = $stderr[2]
</hansl>
Is it possible to retrieve data for the test statistic and p-value for the independent
variables like I do for the coefficient and standard error estimations?
Thanks,
Filipe
2:15 p.m.
On Mon, 28 Sep 2020, Filipe Costa wrote:
Dear Gretl users,
I was trying to retrieve some data from an estimated model but I'm not sure
that's possible.
An example:
<hansl>
open australia.gdt
# simple ols model
ols E 0 PAU PUS
# scalars for the coefficients
coef1 = $coeff[1]
coef2 = $coeff[2]
# scalars for the standard errors
stderr1 = $stderr[1]
stderr2 = $stderr[2]
</hansl>
Is it possible to retrieve data for the test statistic and p-value
for the independent variables like I do for the coefficient and
standard error estimations?
Not as such, but it's straightforward to reconstruct those values
given $coeff and $stderr.
tstat1 = coef1 / stderr1
# P-value for two tailed test
pval1 = 2 * pvalue(t, $df, abs(tstat1))
Allin
3:10 p.m.
Ok, thanks for the quick reply. Just another question related to these models. Let's
say I run several models in one session and save them in this way:
Model1 <- ols y 0 x1
Model2 <- ols y 0 x2
I can retrieve the models with the code:
Model1.show
Model2.show
But a question I have is the following: can I by then retrieve the coefficients and
standard errors of the regressions using the console? The variables $coeff and $stderr
store info just for the last model run, so can't use them after running everything. If
not the best is then to create scalars just after each model to store the variables to
then retrieve them. Is there other way?
> On 28/09/2020 15:15 Allin Cottrell <cottrell(a)wfu.edu> wrote:
>
>
> On Mon, 28 Sep 2020, Filipe Costa wrote:
>
> > Dear Gretl users,
> >
> > I was trying to retrieve some data from an estimated model but I'm not sure
that's possible.
> >
> > An example:
> >
> > <hansl>
> > open australia.gdt
> >
> > # simple ols model
> > ols E 0 PAU PUS
> >
> > # scalars for the coefficients
> > coef1 = $coeff[1]
> > coef2 = $coeff[2]
> >
> > # scalars for the standard errors
> > stderr1 = $stderr[1]
> > stderr2 = $stderr[2]
> > </hansl>
> >
> > Is it possible to retrieve data for the test statistic and p-value
> > for the independent variables like I do for the coefficient and
> > standard error estimations?
>
> Not as such, but it's straightforward to reconstruct those values
> given $coeff and $stderr.
>
> tstat1 = coef1 / stderr1
> # P-value for two tailed test
> pval1 = 2 * pvalue(t, $df, abs(tstat1))
>
> Allin
> _______________________________________________
> Gretl-users mailing list -- gretl-users(a)gretlml.univpm.it
> To unsubscribe send an email to gretl-users-leave(a)gretlml.univpm.it
> Website: https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/
4:10 p.m.
Am 28.09.2020 um 17:10 schrieb Filipe Costa:
Ok, thanks for the quick reply. Just another question related to
these models. Let's say I run several models in one session and save them in this
way:
Model1 <- ols y 0 x1
Model2 <- ols y 0 x2
I can retrieve the models with the code:
Model1.show
Model2.show
But a question I have is the following: can I by then retrieve the coefficients and
standard errors of the regressions using the console? The variables $coeff and $stderr
store info just for the last model run, so can't use them after running everything. If
not the best is then to create scalars just after each model to store the variables to
then retrieve them. Is there other way?
There are several ways; you could stack the vector resuls ($coeff,
$stderr) one after another into one growing matrix.
Or you could grab the whole $model bundle after each estimation and
store that one. You could even put those bundles together into an array
of bundles, like this:
<hansl>
bundles mymodelbundles
loop foreach i x1 x2
ols y 0 $i --quiet
mymodelbundles += $model
endloop
</hansl>
Then you have the coeff and stderr and lots more inside each of the
$model bundles.
cheers
sven
12:38 p.m.
On Mon, 28 Sep 2020, Filipe Costa wrote:
Ok, thanks for the quick reply. Just another question related to
these models. Let's say I run several models in one session and save them in this
way:
Model1 <- ols y 0 x1
Model2 <- ols y 0 x2
I can retrieve the models with the code:
Model1.show
Model2.show
But a question I have is the following: can I by then retrieve the
coefficients and standard errors of the regressions using the
console? The variables $coeff and $stderr store info just for the
last model run, so can't use them after running everything.
For named models:
matrix b = Model1.$coeff
matrix se = Model1.$stderr
Allin
1:06 p.m.
Am 29.09.2020 um 14:38 schrieb Allin Cottrell:
On Mon, 28 Sep 2020, Filipe Costa wrote:
> But a question I have is the following: can I by then retrieve the
> coefficients and standard errors of the regressions using the
> console? The variables $coeff and $stderr store info just for the
> last model run, so can't use them after running everything.
For named models:
matrix b = Model1.$coeff
matrix se = Model1.$stderr
Hm, interesting, I didn't know this syntax. (Model1.coeff doesn't work,
although $model.coeff does...)
I'm noticing that the following doesn't work, though:
- Model1.$sample_t2 and .$sample_t1
-...$t2 and $t1
Not sure if there's a deeper rationale for that.
thanks
sven
1567
days inactive
1568
days old
5 comments
4 participants
participants (4)
-
Allin Cottrell -
Filipe Costa -
Filipe Costa -
Sven Schreiber