Hi Allin and Jack,
I’m a real trouble maker, it seems. :)

OK, here is a test file with the troubled data with only the hprice and inc 
variables. In my current gretl, it will read the data as string-valued series, 
and a freq on hprice will crash my gretl.
If I shorten the dataset, gretl will not read it and will give an error message 
instead, so I have to send you the full sample, which is much larger than the 
one I send before.

Fred

Attachment: test3.xlsx
Description: MS-Excel 2007 spreadsheet

> 
> Hi Jack,
> Attached is a small sample dataset with 3 string-valued variables, CITY, =
> hprice, & inc that crashes on CITY, hprice, but not on inc.
> 
> Fred
> 
> 
>> 
>> On Thu, 24 Dec 2020, Fred Engst wrote:
>> 
>>> Hi Allin,
>>> Another repeatable crash today. This is from a dataset that my student =
>> gave me.
>>> After looking into the issue, I found that the problematic data series =
>> was actually read as a string-valued
>>> series that looked like numbers.
>>> Once I corrected this data issue, the frequency plot was fine.
>>> However, it still crashes on any string-valued data series.
>>> Attached is the crash report.
>> 
>> This is weird. Normally, there is no problem with "freq" on string-valued=
>> =20
>> variables. Example:
>> 
>> <hansl>
>> open grunfeld.gdt
>> freq ticker --plot=3Ddisplay
>> </hansl>
>> 
>> There must be something funny in your student's dataset. Would you mind=20
>> sending it over? Thanks.
> 
> 
> --Apple-Mail=_F6A326C4-1187-46BA-BB14-9E5641F06218--
> 
> ------------------------------
> 
> Date: Thu, 24 Dec 2020 08:55:07 -0500 (EST)
> From: Allin Cottrell <[email protected]>
> Subject: [Gretl-users] Re: Repeatable crash on frequency distribution
>       on a string-valued data series from GUI
> To: Gretl list <[email protected]>
> Message-ID:
>       <[email protected]>
> Content-Type: text/plain; charset=US-ASCII; format=flowed
> 
> On Thu, 24 Dec 2020, Fred Engst wrote:
> 
>> Another repeatable crash today. This is from a dataset that my 
>> student gave me.
>> 
>> After looking into the issue, I found that the problematic data 
>> series was actually read as a string-valued series that looked 
>> like numbers.
>> 
>> Once I corrected this data issue, the frequency plot was fine.
>> However, it still crashes on any string-valued data series.
>> Attached is the crash report.
> 
> Thanks, Fred. That's now fixed in git, snapshots to follow.
> 
> Could you please send me the xlsx file that was imported to create 
> the test3.gdt file that you posted? Besides the crashing problem 
> itself there's something funny happening on importing those data: we 
> end up with one string value consisting of 34 spaces, for hprice.
> 
> Allin
> 
> ------------------------------
> 
> Date: Thu, 24 Dec 2020 16:20:18 +0100 (CET)
> From: "Riccardo (Jack) Lucchetti" <[email protected]>
> Subject: [Gretl-users] Re: Repeatable crash on frequency distribution
>       on a string-valued data series from GUI
> To: Gretl list <[email protected]>
> Message-ID: <alpine.DEB.2.21.2012241618280.2199@newvaio>
> Content-Type: multipart/mixed;
>       boundary="8323329-1436782324-1608823222=:2199"
> 
> --8323329-1436782324-1608823222=:2199
> Content-Type: text/plain; charset=ISO-8859-15; format=flowed
> Content-Transfer-Encoding: quoted-printable
> 
> On Thu, 24 Dec 2020, Allin Cottrell wrote:
> 
>> On Thu, 24 Dec 2020, Fred Engst wrote:
>> 
>>> Another repeatable crash today. This is from a dataset that my student=
> gave=20
>>> me.
>>> =20
>>> After looking into the issue, I found that the problematic data series=
> was=20
>>> actually read as a string-valued series that looked like numbers.
>>> =20
>>> Once I corrected this data issue, the frequency plot was fine.
>>> However, it still crashes on any string-valued data series.
>>> Attached is the crash report.
>> 
>> Thanks, Fred. That's now fixed in git, snapshots to follow.
>> 
>> Could you please send me the xlsx file that was imported to create the=20
>> test3.gdt file that you posted? Besides the crashing problem itself the=
> re's=20
>> something funny happening on importing those data: we end up with one s=
> tring=20
>> value consisting of 34 spaces, for hprice.
> 
> I just pushed another commit to git to fix a bug exposed by the data file=
> =20
> that Fred sent. You'd get a segfault if you tried to apply the atof()=20
> function to the hprice series. Now this is fixed.
> 
> Thanks, Fred!
> 
> -------------------------------------------------------
>  Riccardo (Jack) Lucchetti
>  Dipartimento di Scienze Economiche e Sociali (DiSES)
> 
>  Universit=E0 Politecnica delle Marche
>  (formerly known as Universit=E0 di Ancona)
> 
>  [email protected]
>  http://www2.econ.univpm.it/servizi/hpp/lucchetti
> -------------------------------------------------------
> --8323329-1436782324-1608823222=:2199--
> 
> ------------------------------
> 
> Subject: Digest Footer
> 
> _______________________________________________
> Gretl-users mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
> Website: 
> https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/
> 
> 
> ------------------------------
> 
> End of Gretl-users Digest, Vol 167, Issue 30
> ********************************************

_______________________________________________
Gretl-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Website: 
https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/

Reply via email to