List from string in a function
by Michael Boldin
I am using GRETL 1.912 and ran into a problem using strings in a user
defined function
I wanted to create a function that accepts a bundle with a set of
specification terms.
but I understand bundles can not hold lists since the last step in the
code below fails
list xlist= dy(-1) de(-1)
bundle amodel=null
amodel.xlist= xlist
So I thought I could pass a string of variable names through the bundle
since this code works
string xstring= "dy(-1) de(-1)"
list xlist= @xstring
ols dy const xlist
However, I get a error message with a function version
function matrix olslist1(string xstring)
list xlist= @xstring #fails
ols dy const xlist
#ols dy const @xstring #also fails
return $coeff
end function
The message is
> list xlist= dy(
The symbol 'dy' is undefined
The problems seems fundamental to strings in a function since
print @xstring
works outside of function but fails inside a function
Or am I using strings and functions incorrectly?
10 years, 10 months
question
by Salvador Rodriguez
Because you can use x12 arima or x13arima for modeling data are weekly, is
that I can not, so I've seen can only be used for weekly, quarterly, data
etc. .... I need help with this, in advance thank you! ...[?].
10 years, 10 months
command log window
by artur.bala.tn
Dear Allin,
I noticed an unusual behavior of the command log window (the "open" command is duplicated, the "smpl" command is truncated...).
Eg, if you run the following commands
open "C:\Program Files\gretl\data\misc\credscore.gdt"
smpl Selfempl --dummy
smpl Selfempl=0 --restrict --replace
here's what the command log window will look like:
Best,
Artur
10 years, 10 months
subsampling in a progressive loop
by artur.bala.tn
Dear all,
I encounter a problem in a progressive loop while estimating probit models on different subsamples.
In its first run the loop works fine. In the second run gretl stops executing pointing out that "dataset is subsampled, model is not. Can't retrieve yhat: data set has changed".
As a roundabout solution I attempted to run the same script several times outside the loop and there's absolutely no error coming out: it just works fine.
The script looks like this :
\<hansl>
loop 250 --progressive
smpl full
matrix alldata = {dataset}
alldata = resample(alldata)
# some coding to create series from the alldata matrix
...
# estimate the first probit model and retrieve $yhat
probit V1 const V2
genr resid1 = $yhat
...
# subsample, estimate the 2nd probit model en retrieve $yhat
smpl V4 = 1 --restrict
probit V5 const V6
genr resid2 = $yhat
...
endloop
Best,
Artur
10 years, 10 months
No text output from x12
by Bob McCall
I recently installed gretl 1.9.14 and added x13as. It runs and produces
graphical
output but no text output, just a blank term. This is running on Mint 16.
Any suggestions?
thanks,
Bob
10 years, 10 months
monthly price index
by dilek guler
Dear All,
I have a problem. I want to construct a monthly price index.
Orginal question is:
"Convert the coefficients of the monthly dummy variables you use for your
index construction into
large percentage change format with the transformation exp(b)-1 Provide a
table
with summary statistics on all variables in your regression other than
neighborhood
dummies."
I created and describe the dummies. I could not find the next steps.
How can construct a montly price index?
Best,
10 years, 10 months
(no subject)
by dilek guler
Hello,
I want to make log variable and i see a message
it is:
"Generated missing values"
How can i correct it?
And the second question is How can i make separate dummy variables for
another one?
Best!
10 years, 10 months
Test for AR(2) errors: z = NA
by Anutechia Asongu
Dear All,
Good morning,
I have an important issue troubling me. I had some prior trouble with the AR(2) test in my initial dataset. To this end, I
have increased the N (53) and reduced the T (4 and 3) of my dataset to mitigate potential issues of instrument proliferation. Hence, I'm now working with 53 cross-sections of 4-year and 3-year non-overlapping intervals. I am now more confident with the identification restrictions (more or less four degrees of freedom in overidentification).
But my concern persists: after performing the two-step System GMM, the results for the AR(2) errors test reads 'not applicable'. Can I use the AR(1) test as the information criterion for the Arellano & Bond autocorrelation test?
Thank you for the opportunity of guidance.
Best,
Simplice
10 years, 10 months
gretl discriptive statistics for MBA students
by Logan Kelly
Hello all,
This is a question that probably has not come up before (and may be a new feature request) because I doubt that many researchers using gretl need it much.
Is there an easy way for my MBA student to generate the descriptive statistics including the standard error of the mean and confidence intervals for the mean. There are a number of ways to do this already, for example one could regress a series on a constant. But this is a bit more difficult to explain to my MBA students who still do not understand why they can't just use Excel.
Thanks,
Logan
10 years, 10 months
on logit/probit models
by artur.bala.tn
Dear all,
Is there any way (or post-estimation internal variable) to retrieve the % of cases correctly predicted for a logit/probit model?
Best,
Artur
10 years, 10 months
