On 12/23/2013 10:47 PM, Allin Cottrell wrote:
> On Mon, 23 Dec 2013, Sven Schreiber wrote:
> 
>> Am 23.12.2013 20:57, schrieb Dominik Menno:
>>> Dear Gretl users,
>>>
>>> sorry for bothering you. I found it out myself; I guess it is an issue
>>> of portability between operation systems (in particular Linux to
>>> windows) and with file encoding/line endings of text files.
>>
>> That's actually what I had thought first, but...
>>>
>>> Here in particular, it was the file ending of the csv-files that caused
>>> the problem. I now generate the CSV file so that it has the windows file
>>> ending (encoding is now ASCII/CRLF), see attached file.
>>
>> your earlier file also had the Windows line endings (CRLF), so it seemed
>> that that was not the issue.
> 
> The file as attached to 
> http://lists.wfu.edu/pipermail/gretl-users/2013-December/009509.html
> has 0a (plain LF) line endings, according to Emacs's hex mode.
> 
>> In any case, I was under the impression that gretl would import/accept
>> text files regardless of the platform-specific line endings. (not 100%
>> sure though)
> 
> Yes -- or at least that's the intent: we aim to support CR/LF, LF, 
> CR, and also unholy mixtures of these.
> 
>> So there may still be something there, perhaps also specific to the
>> experimental 64-bit version of gretl on Win that you're using
>> apparently.
> 
> The 1.9.14 gretl release for 64-bit Windows doesn't really count as 
> "experimental": it's officially supported now.
> 
>> Could you try the precise file you sent to the list first
>> (re-grab it from the message attachment) if it really works?
> 
> I too would be interested to hear how that goes.
> 
> Allin Cottrell

Hi Sven and Allin, thanks for your help. The original file I attached to
the first mail had the unix type of line ending (is this plain??). So I
did as you suggested and the error remains under windows 64bit.  Here is
the gretl output:

parsing E:\CollegeDistOrigin.csv...
using delimiter ','
   longest line: 100 characters
   first field: 'female'
   number of columns = 14
   number of variables: 14
   number of non-blank lines: 342
scanning for variable names...
   line:
female,black,hispanic,bytest,dadcoll,momcoll,ownhome,urban,cue80,stwmfg80,dist,tuition,ed,incomehi

scanning for row labels and data...
   the cell for variable 1, obs 1 is empty: treating as missing value
treating these as undated data

Again, as I said if I generate the CSV file with CRLF file ending, the
error disappears under Windows. I can import the file both under windows
and under Linux without error message.

Best,
Dominik

> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
> 

-- 
Dominik Menno
RWTH Aachen University

Templergraben 64
52064 Aachen
Tel: +49 (0)241/80 96 286
Fax: +49 (0)241/80 92 337
dominik.menno(a)rwth-aachen.de

Reply via email to