[prev in list] [next in list] [prev in thread] [next in thread] 

List:       wekalist
Subject:    [Wekalist] Re: weka load error: unable to determine structure as arff (reason: java.ioexception keyw
From:       Jeff Pattillo <jeff.pattillo () gmail ! com>
Date:       2012-10-18 20:49:14
Message-ID: loom.20121018T223701-418 () post ! gmane ! org
[Download RAW message or body]

> Hi. I can't load a dataset. I receive the error "unable to determine structure
as arff (reason: java.ioexception keyword  <at> relation expected, read
Token[ï » ¿], line 1)". I have googled this problem and not found a resolution.

I posted about the exact same problem on Oct 8 on this list.  I never figured
out if I could fix it by changing the encoding but I do have a incredibly stupid
hack that works.  If you erase the line with the problem (line 1 in your case),
I have found that the remaining lines do not have a  ¿ appended to them and the
import works.

Erasing the line can obviously cause problems so here is what I have done:
1) When the problematic line is the header, such as  ¿@relation, I've found if
you saved the header in a separate file, and then copy and paste it in your data
file, for some reason the  ¿ does not come with it.
2) When the problematic line is in the data, force the program that generated
the data to output a dummy line so you can always erase as the first line.

Obviously these are stupid solutions but they work.  Does anyone have legitimate
solution to this problem?  Because I've found it to be true for notepad,
notepad++, and vim.




_______________________________________________
Wekalist mailing list
Send posts to: Wekalist@list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/=
listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.=
html


[prev in list] [next in list] [prev in thread] [next in thread]