6:42 p.m.
Hello, when I am trying to build a panel model with random effects either
from the gui or console gretl crashes. I haven't found any bugreports in the
list about panel models for version 1.7.9 I'm using gretl on windows XP.
It happens only with the supplyed data - it is actually not for that
specification, but crash is not the right behaviour too...
Ivan Sopov.
9:28 p.m.
On Thu, 18 Dec 2008, Ivan Sopov wrote:
Hello, when I am trying to build a panel model with random effects
either
from the gui or console gretl crashes. I haven't found any bugreports in the
list about panel models for version 1.7.9 I'm using gretl on windows XP.
Sorry, but are you using version 1.7.9 or the current snapshot?
With current CVS/snapshot gretl, taking y as dependent and x1 and
x2 as independent in your test dataset, if I try to estimate a
random effects model I get a valid error message saying I don't
have enough data, no crash.
Allin.
9:31 p.m.
On Thu, 18 Dec 2008, Allin Cottrell wrote:
On Thu, 18 Dec 2008, Ivan Sopov wrote:
> Hello, when I am trying to build a panel model with random
> effects either from the gui or console gretl crashes. I
> haven't found any bugreports in the list about panel models
> for version 1.7.9 I'm using gretl on windows XP.
Sorry, but are you using version 1.7.9 or the current snapshot?
With current CVS/snapshot gretl, taking y as dependent and x1 and
x2 as independent in your test dataset, if I try to estimate a
random effects model I get a valid error message saying I don't
have enough data, no crash.
To be a little more explicit, I think this may be a bug that Jack
or I found and already fixed in CVS, so there may have been no
public bug report.
Allin.
10:15 p.m.
> Sorry, but are you using version 1.7.9 or the current snapshot?
> With current CVS/snapshot gretl, taking y as dependent and x1 and
> x2 as independent in your test dataset, if I try to estimate a
> random effects model I get a valid error message saying I don't
> have enough data, no crash.
Allin is right, I've also reported this bug before. In current
snapshot everything works fine.
--
Mariusz Doszyń
8:09 a.m.
Sorry. I've tried google:"random effects site:
lists.wfu.edu/pipermail/gretl-devel/" and didn't found anything about it. It
was in
http://lists.wfu.edu/pipermail/gretl-users/2008-November/
I'm using 1.7.9 version... Actually I don't have access to the
ricardo.ecn.wfu.edu for downloading snapshot for last 2 or 3 months... maybe
something is wrong with my internet provider.
Many thanks.
2008/12/19 Mariusz Doszyń <mariusz.doszyn(a)gmail.com>
>> Sorry, but are you using version 1.7.9 or the current
snapshot?
>> With current CVS/snapshot gretl, taking y as dependent and x1 and
>> x2 as independent in your test dataset, if I try to estimate a
>> random effects model I get a valid error message saying I don't
>> have enough data, no crash.
Allin is right, I've also reported this bug before. In current
snapshot everything works fine.
--
Mariusz Doszyń
_______________________________________________
Gretl-devel mailing list
Gretl-devel(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-devel
11:15 p.m.
On Fri, 19 Dec 2008, Ivan Sopov wrote:
"I'm using 1.7.9 version... Actually I don't have access to the
ricardo.ecn.wfu.edu for downloading snapshot for last 2 or 3
months... maybe something is wrong with my internet provider."
Maybe. The IP number is 152.17.48.43, if that helps.
Allin Cottrell
5846
days inactive
5847
days old
5 comments
3 participants
participants (3)
-
Allin Cottrell -
Ivan Sopov -
Mariusz Doszyń