Re: Import large field-delimited file with strings and numbers

[CdeMills]

Joao Rodrigues wrote
> I got this:
> 
> octave:1> pkg install -forge dataframe
> error: could not download file dataframe-1.0.0.tar.gz from url 
> http://packages.octave.org/download/dataframe-1.0.0.tar.gz
> error: called from:
> error:   /usr/share/octave/3.8.1/m/pkg/pkg.m at line 390, column 13
> 
> While browsing in the web I did not find any documentation (e.g., what 
> is the command to import a file or the format of objects), only a minor 
> description stating that it is analogue to R's data.frame. Can you point 
> me in the right direction?

Strange. I'll ask Carnë to look at this issue. He uploaded this version at
the end of march. My guess is that you need to download the package manually
and  install it from the local file.

For the  use, the main information is contained in the main file itself.
Some simple call:

resu=dataframe('some_data.csv');

There are a few options to struggle with issue like numbers encoded as
'3,14' where the comma is used as decimal separator, skip some initial
garbage (free format header f.i.), and so on. 

Once the data are loaded, the object tries to mimic an ordinary array;
except that the only constraint is that each column must be from an
homogenous type. You may have a global object with the first colum a string,
the second a char, the third an integer, and so on.

regards

Pascal



--
View this message in context: 
http://octave.1599824.n4.nabble.com/Import-large-field-delimited-file-with-strings-and-numbers-tp4666380p4666438.html
Sent from the Octave - General mailing list archive at Nabble.com.