Re: [Gretl-users] A strange results in the icon view
On Thu, 11 Mar 2010, Allin Cottrell wrote:
On Fri, 12 Mar 2010, yinung at Gmail wrote:
> I have a mydata1.gdt and a script named test.inp as attached. After I
> re-run the script, a strange result in the icon view occurs. That is,
> there are two model1, model2, model2, ... and so on. If I re-run the
> script again. these icons of models will be replicated accumulatively,
> though every individual icon works normally if you click on it. (the
> screenshot of the icon view is also attached)
>
> Is there anything wrong with my script? Or, is this a minor bug?
It's a bug, and it'll be fixed shortly.
Now fixed in CVS.
It's unrelated, but there's a bug in your script. Your dataset
contains a variable named "weekday" which holds values 1 to 5.
You create a set of 5 dummies via
series dum1=(weekday-1=1)
series dum2=(weekday-1=2)
series dum3=(weekday-1=3)
series dum4=(weekday-1=4)
series dum5=(weekday-1=5)
and you then include all 5 dummies (plus the constant) in an OLS
regression. In fact the variable dum5 will contain all zeros, and
hence will be automatically deleted from the regression list.
If the dummies were defined "correctly" (?) as, e.g.,
series dum1=(weekday=1)
then dum5 would still be deleted, since the 5 dummies would be
jointly perfectly collinear with the constant.
One smaller point: the recommended syntax uses "==" for testing
for equality, as in
series dum1=(weekday==1)
Allin