gretl crashed while searching for file within GUI
by Fred Engst

Hi Allin,
gretl crashed once again, but for a different reason this time.
It seems that whenever I tried to do a search for files within GUI after “file->Open data->User file...” or “file->Session files->Open session…”, it crashes right after it showed a list of files.
I can’t remember a time when I did a search like this without gretl crashing.
I suspect it has something to do with my character encoding. Even though I use standard English, I have Chinese on my system also.
This is more like a FYI, and it is not urgent, as I know how to get around it easily.
Attached is the crash report if it helps.
Fred
1 year

Package KaoTest.gfn to be retired soon
by Sven Schreiber

Dear all,
especially those interested in panel cointegration analysis, the
KaoTest.gfn package has certain issues and will be removed from the
gretl package server soon. The author couldn't be reached to address
those issues.
The good news is that work is ongoing to start a new package for panel
cointegration testing. The plan is that it will offer more tests than
the KaoTest package.
Nevertheless, if you are a user of that package and need it for your
work please keep a local copy of it to be on the safe side.
cheers
sven
1 year, 3 months

Export to Stata fails
by Sven Schreiber

Hi,
(with gretl snapshot Apr 4th) I tried to export the penngrow.gdt data
file to Stata format (dta). Opening it in Stata 14 works fine at first,
but the values are totally rubbish.
Going via csv works OK.
thanks
sven
1 year, 4 months

Issue with <dataset addobs> in a function
by Artur Tarassow

Hi,
I've experienced the following issue trying to augment a time-series
dataset within a function. Even though adding some observations
initially works, after restricting the sample and calling "smpl full"
again the additional observations get lost. This happens with latest git
on ubuntu:
<hansl>
open denmark.gdt -q
function void foo (series s)
eval $t2 # T=55
dataset addobs 10
eval $t2 # T=65
smpl 1974:2 1974:3
smpl full
eval $t2 # T=55
end function
foo(LRM)
</hansl>
Is this is really intended? If yo, why?
Best,
Artur
1 year, 5 months

None-uniform distribution of the p-value from ADF test in gretl when there is an intercept & trend? (Allin Cottrell)
by Fred Engst

>
> On Fri, 26 Apr 2019, Sven Schreiber wrote:
>
>>>> cheers
>>>> sven
>>>> =20
>>> =20
>>> Thanks Sven,
>>> But I=E2=80=99m more confused.
>>> If you are right, then what does the option "--ct (with constant and=20
>>> trend)=E2=80=9D as stated in gretl's documentation mean for ADF test?
>>> If I set both the intercept and trend parameters to 0.1, doesn=E2=80=99=
> t the option=20
>>> =E2=80=9C--ct=E2=80=9D in DGP become
>>> y(t)=3D0.1 + 0.1*t + y(t-1)+ e ?
>>
>> In the unit root case the DGP and the test equation are not the same as
>> regards the included terms. That's because of the implicit cumulation o=
> f
>> exogenous terms in the DGP. The ADF test options refer to the test
>> equation, not to the DGP equation (under H0).
>>
>> This is a general property and not specific to gretl.
>
> And the practical implication is that for Fred's DGP,
>
> y(t)=3D0.1 + 0.1*t + y(t-1)+ e
>
> where the 0.1*t cumulates, the adf-test option that should yield a=20
> uniformly distributed p-value is --ctt (constant and quadratic trend).
>
> Allin
You are so right Allin!
I tried and indeed, the --ctt option does yield an uniform distribution.
It is clear that I was a bit too simple minded in understanding those options when generating random walk series.
Then, again, what kind of DGP that WILL yield a distribution that IS correct to use the --ct option?
Perhaps I should read more on random walk DGP and unit root testing before asking more about this here.
Thanks for all your time.
Fred
1 year, 5 months

Re: Gretl-users Digest, Vol 147, Issue 30 (series from matrix)
by Alecos Papadopoulos

Thanks Artur I think this is it. I found out that for some reason the
first-to-be-generated vector "q" in my script below had lower dimension
than the sample size. I don't know why that happened yet. But when I
managed to obtain a "q" vector with the correct dimension, then the line
<<series qqq = sumr(q.*qq)>>
worked just fine, and gave me qqq as a series.
Thanks!
Alecos Papadopoulos PhD
Athens University of Economics and Business
web: alecospapadopoulos.wordpress.com/
cell:+30-6945-378680
skype:alecos.papadopoulos
On 27/4/2019 01:00, gretl-users-request(a)gretlml.univpm.it wrote:
>
> Date: Fri, 26 Apr 2019 21:09:27 +0200 From: Artur Tarassow
> <atecon(a)posteo.de> Subject: [Gretl-users] Re: Series from matrix To:
> gretl-users(a)gretlml.univpm.it Message-ID:
> <44aa51da-ae2c-5bdb-4362-3c61fad929f3(a)posteo.de> Content-Type:
> text/plain; charset=iso-8859-7; format=flowed Am 26.04.19 um 21:06
> schrieb Alecos Papadopoulos:
>> Good afternoon.
>>
>> Has something changed in Gretl related to how we get data out of a
>> matrix column and into a series?
>>
>> I had an old mle script that included
>>
>> series qqq = sumr(q.*qq)
>>
>> where the right hand side is a column vector. It had worked in the
>> past, but it doesn't appear to work any more, I get
>>
>> <<Incompatible types in assignment: series = matrix . The formula
>> 'series qqq = sumr(q.*qq)' produced an error on execution>>
> Hi Alecos,
>
> I've experienced some similar behaviour a while ago. Could you check
> please that the length of the resulting vector exactly meets the length
> of the currently selected data set?
>
> Artur
1 year, 5 months

Series from matrix
by Alecos Papadopoulos

Good afternoon.
Has something changed in Gretl related to how we get data out of a
matrix column and into a series?
I had an old mle script that included
series qqq = sumr(q.*qq)
where the right hand side is a column vector. It had worked in the
past, but it doesn't appear to work any more, I get
<<Incompatible types in assignment: series = matrix . The formula
'series qqq = sumr(q.*qq)' produced an error on execution>>
I run Gretl 2019a, 64-bit, Windows 7.
--
Alecos Papadopoulos PhD
Athens University of Economics and Business
web: alecospapadopoulos.wordpress.com/
1 year, 5 months

Issue with strptime()
by Artur Tarassow

Hi all,
I am a bit puzzled by the following example where I first convert a date
string into a timestamp before converting int back into a date string.
However, instead of getting returned "2005:01", gretl returns "2004:12".
It seems that the strptime() function does some weird thing, as can be
checked here: https://www.epochconverter.com/
But maybe I miss something fundamental here.
This is with latest git on ubuntu 19.04.
<hansl>
string init_obs = "01/2005"
scalar tstamp = strptime(init_obs, "%m/%Y")
eval strftime(tstamp, "%Y:%m") # returns '2004:12'
</hansl>
Best,
Artur
1 year, 5 months

Re: None-uniform distribution of the p-value from ADF test in gretl when there is an intercept or trend? (Sven Schreiber)
by Fred Engst

>
> Am 25.04.2019 um 06:17 schrieb Fred Engst:
>
>>>
>>> I guess you didn't do anything wrong. When the null hypothesis is
>>> true and a test is working as designed, its p-value will be
>>> uniformly distributed on (0,1). Then using marginal significance
>>> level alpha you'll reject 100*alpha percent of the time, and the
>>> test is properly sized. Funny how probability works.
>>>
>>> Allin
>>
>> Thanks Allin.
>> That is interesting indeed. In other words, with an uniform distribution of the p-value [when y(t) = y(-1)], there is only 5% changes that we do reject the null of random walk at 5% level of significance.
>> However, for other parameters of the intercept or trend, the p-value distribution is not uniform. So their size is incorrect?
>> For example, when I set both the intercept and trend parameters to 0.1, I get the attached table.
>
> Hi Fred, your DGP is wrong.
>
> When you apply your formula:
> y = b0 + rho*y(-1) + b1*index + e
>
> For non-zero b1 and rho=1 the trend term will be cumulated and will
> result in a quadratic trend. This is a case not covered by the standard
> ADF tests.
>
> cheers
> sven
>
Thanks Sven,
But I’m more confused.
If you are right, then what does the option "--ct (with constant and trend)” as stated in gretl's documentation mean for ADF test?
If I set both the intercept and trend parameters to 0.1, doesn’t the option “--ct” in DGP become
y(t)=0.1 + 0.1*t + y(t-1)+ e ?
Fred
1 year, 5 months

Re: gretl crashed while trying a qqplot (Allin Cottrell)
by Fred Engst

>
> On Thu, 25 Apr 2019, Fred Engst wrote:
>
>> Hi Allin,
>>
>> As I was trying to figure out the relationship between ADF reported=20
>> p-value vs. observed distribution of the slope estimation, an=20
>> attempt on qqplot crashed gretl. It is a repeatable crash (report=20
>> attached).
>
> Thanks for the report, Fred. This is now fixed in git and snapshots.
>
>> I then tried gretlcli, and the message it gave was:
>> ? qqplot ct_p b_rho
>> Gnuplot is broken or too old: must be >=3D version 5.0
>> I=E2=80=9Dm not sure this is related to the crash for gretlcli can run =
> gnuplot anyway.
>
> As for the message from gretlcli about gnuplot, I'll look into that=20
> when I can next experiment on a Mac.
>
> Allin
>
Thanks so much Allin, you are amazingly fast!
Yes, it works great now.
Fred
1 year, 5 months