On Mon, Jul 08, 2013 at 07:56:28PM +0200, Olivier Grisel wrote:
> > Then you'd have pdf, logpdf, cdf, logcdf, sf, rvs (not wild about this
> > one, and I think we use sample in places), etc.
> I am not found of acronyms, especially when there not common at all
> such as `rvs`. I think RVS stands f
2013/7/10 Mike Hansen :
> I have been using Scikit's text classification for several weeks, and I
> really like it. I use my own corpus (self-generated) and prepare each
> document using the NLTK. Presently I am relying on this tutorial/code-base,
> only making changes when absolutely necessary f
I have been using Scikit's text classification for several weeks, and I really
like it. I use my own corpus (self-generated) and prepare each document using
the NLTK. Presently I am relying on this tutorial/code-base, only making
changes when absolutely necessary for my documents to work.
The
2013/7/10 Josh Wasserstein :
> Thanks Olivier. Would you mind elaborating on why having a grid that is this
> fine-grained is a problem? Is it because the library runs into overflow
> problems? Something else?
No it's just that it will run forever and that you are probably just
wasting CPU time.
Ended up going with the approach of one of the SO folks.
https://github.com/scikit-learn/scikit-learn/pull/2150
On Wed, Jul 10, 2013 at 1:56 PM, Olivier Grisel wrote:
> Maybe you can try to put:
>
> from __future__ import unicode_literals
>
> in some appropriate place?
>
> Also:
> http://stacko
If you do 4 CV folds on each parameter combination of your fine grid, then:
191880 * len(skf) == 191880 * 4 == 767520
--
Olivier
2013/7/10 Olivier Grisel :
> 2013/7/10 Josh Wasserstein :
>> Thanks Olivier. Would you mind elaborating on why having a grid that is this
>> fine-grained is a problem?
On Wed, Jul 10, 2013 at 03:15:46PM -0400, Josh Wasserstein wrote:
> Thanks Olivier. Would you mind elaborating on why having a grid that is this
> fine-grained is a problem? Is it because the library runs into overflow
> problems? Something else?
Because it's going to take ages.
G
--
Thanks Olivier. Would you mind elaborating on why having a grid that is
this fine-grained is a problem? Is it because the library runs into
overflow problems? Something else?
Thanks,
Josh
On Wed, Jul 10, 2013 at 1:29 PM, Olivier Grisel wrote:
> 2013/7/10 Josh Wasserstein :
> > Thanks Robert a
Maybe you can try to put:
from __future__ import unicode_literals
in some appropriate place?
Also:
http://stackoverflow.com/questions/13473971/multi-version-support-for-python-doctests
--
See everything from the browse
2013/7/10 Justin Vincent :
> We are a little further than that unfortunately. Some failures in conf.py
> were preventing tests other than doctests from running. With conf.py and the
> doctests fixed, we are at 4 errors and 10 failures, which is still pretty
> good.
>
> What do we think should be do
2013/7/10 Josh Wasserstein :
> Thanks Robert and everyone else. I am still having a strange problem with
> GridSearchCV with 0.14-git. Even though the number returned by
>
> ===
> len(list(ParameterGrid(tuned_parameters)))
> ===
>
> is
Thanks Robert and everyone else. I am still having a strange problem with
GridSearchCV with 0.14-git. Even though the number returned by
===
len(list(ParameterGrid(tuned_parameters)))
===
is 191,880, my call to:
===
Hi pythonistas,
As there are many scientific developpers using IPython, I think some
of you may be interested in this IPython Drinkup in Paris.
Of course, we will have our own during the sprint, but if anyone wants
to meet some of the IPython dev, this is the place to be!
If you are interested in
Hi List,
I've added a paypal button to the main page of the scikit-learn website, as
well
as on the 'About us' section. As you may know, our second International
Scikit-learn
Code-sprint is around the corner, and we would be very appreciative to any
form of
donation. Donations can be made through
We are a little further than that unfortunately. Some failures in conf.py
were preventing tests other than doctests from running. With conf.py and
the doctests fixed, we are at 4 errors and 10 failures, which is still
pretty good.
What do we think should be done with u'blah' type strings in the do
On 07/09/2013 12:44 AM, Josh Wasserstein wrote:
> Peter - Yes. That also puzzles me. So odd.
>
> Thanks Olivier - I am using auc_score, not roc_curve. My scikit-learn
> installation does not complain about it. I will try to get the master
> git installed.
>
Well, it doesn't complain, but it doesn
hi everyone,
our next sprint will take place at Telecom ParisTech
(http://www.telecom-paristech.fr) located at:
46 Rue Barrault, 75013 Paris, France
Google Maps link : http://goo.gl/maps/k6QPL
We'll be in room B312 everyday starting at 9am.
See wiki page:
https://github.com/scikit-learn/sciki
17 matches
Mail list logo
