[Gretl-devel] problem with CSV import procedure in gretl (v. 1.9.14)

Hélio Guilherme helioxentric at gmail.com
Thu Mar 27 08:49:08 EDT 2014


The file opens OK in gretl 1.9.15CVS in Linux (Fedora 19).

parsing /home/helio/tmp/aapl.csv...
using delimiter ','
   longest line: 55 characters
   first field: 'date'
   seems to be observation label
   number of columns = 7
   number of variables: 6
   number of non-blank lines: 250
scanning for variable names...
   line: date,Open,High,Low,Close,Volume,AClose
scanning for row labels and data...
   first row label "2014-03-25", last label "2013-04-01"
trying to parse row labels as dates...
Trying date order YYYYMMDD
Could be 2014-03-25 - 2013-04-01
Observations: 249; days in sample: 257
Missing daily observations: 8
reversing the data!
taking date information from row labels

----
The same with Sven's file:

parsing /home/helio/tmp/aapl_sven.csv...
using delimiter ','
   longest line: 56 characters
   first field: 'date'
   seems to be observation label
   number of columns = 7
   number of variables: 6
   number of non-blank lines: 250
scanning for variable names...
   line: date,Open,High,Low,Close,Volume,AClose
scanning for row labels and data...
   first row label "2014-03-25", last label "2013-04-01"
trying to parse row labels as dates...
Trying date order YYYYMMDD
Could be 2014-03-25 - 2013-04-01
Observations: 249; days in sample: 257
Missing daily observations: 8
reversing the data!
taking date information from row labels


----

I think that the fix was to deal with an invisible code BOM indicator.

Maybe you should test with latest Windows snapshot.

Hélio


On Wed, Mar 26, 2014 at 9:13 PM, Evgenii Gilenko <e.gilenko at spbu.ru> wrote:

> Dear developers of gretl!
>
>
>
>
>
> Thanks again for the wonderful software! It helps a lot both in scientific
> research and in educational process!
>
>
>
> Unfortunately, I’ve discovered a problem with importing data from a CSV
> file to a gretl data file.
>
> Specifically, when I try to import  daily data from a CSV file (I attach
> the CSV-file to this letter), I receive the following:
>
>
>
> +++++++++++++++++++++++++++++++++++++++++
>
> parsing D:\aapl.csv...
>
> using delimiter ','
>
>    longest line: 56 characters
>
>    first field: 'date'
>
>    seems to be observation label
>
>    number of columns = 7
>
>    number of variables: 6
>
>    number of non-blank lines: 250
>
> scanning for variable names...
>
>    line: date,Open,High,Low,Close,Volume,AClose
>
> scanning for row labels and data...
>
>    first row label *"014-03-25",* last label "2013-04-01"
>
> treating these as undated data
>
> +++++++++++++++++++++++++++++++++++++++++
>
>
>
> For some reasons gretl cuts off the first digit of the date in the “first
> row label” line (I stressed in red).
>
> And thus it interprets the data as cross-sectional, not as time-series.
>
>
>
> I’m using gretl of version 1.9.14. I tried on version 1.9.13 – the same
> problem.
>
> But, on version 1.9.12 there was no such problem with importing data from
> the attached CSV file – I double-checked it.
>
>
>
> Could you please give any comments on how to override the problem on
> version 1.9.14?
>
> Downgrade to version 1.9.12 is not an option for me now.
>
>
>
> Thanks,
>
>
>
>
>
> Evgenii Gilenko
>
> _______________________________________________
> Gretl-devel mailing list
> Gretl-devel at lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.wfu.edu/pipermail/gretl-devel/attachments/20140327/a0f005da/attachment-0001.html 


More information about the Gretl-devel mailing list