[Corpora-List] python's NLTK vs R's TM

Christian Pietsch chr.pietsch at googlemail.com
Tue Aug 21 11:52:27 CEST 2012

Hi Daniel,

On Mon, Aug 20, 2012 at 08:57:02PM +0200, Daniel Riaņo wrote:
> Does anybody know if there's any chance that the NLTK be translated into Python3 anytime soon?

This is a question for the NLTK users mailing list <http://groups.google.com/group/nltk-users>, but I might as well share here what I read there and tried out recently.

As you know, a Python 3 version of NLTK has not been released yet. Early last year, the NLTK team said they were waiting for NumPy to be ported to Python 3. Then somebody told them that had already happened. I do not know what their current excuse is ;-)

The good news is, there is an unstable fork that runs on Python 3: https://github.com/kmike/nltk/

On my OpenSuse Linux box, I had to install the python3-2to3 and python3-xml packages first (and python3-nose for running the tests). Then I could do this: sudo python3 setup.py install Now “import nltk” works without complaints inside Python 3. To find out if the modules you need work in the Python 3 version, you could run their doctests. Trying to run all tests, I got caught in an infinite loop (or ran out of patience). YMMV.

Cheers, Christian


Christian Pietsch · http://purl.org/net/pietsch

LibTec · Library Technology and Knowledge Management

University Library and CRC 882

Bielefeld University, Bielefeld, Germany -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 490 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20120821/142d6291/attachment.sig>

More information about the Corpora mailing list