this may help
https://github.com/pymc-devs/pymc they have quite a bit in the way of bayes
things
On Fri, Jan 18, 2013 at 5:52 PM, wrote:
> Dear Scikit-learn team,
>
> First, you guys have built an amazing set of ML tools. I am constantly
> impressed by your feedback to the mailing list.
>
> I
Dear Scikit-learn team,
First, you guys have built an amazing set of ML tools. I am constantly
impressed by your feedback to the mailing list.
I wonder if sklearn has a PGM in the works? I've been trying to understand
how to code one up, do you have any suggestions where I could look for a
python
On 19 January 2013 06:47, Satrajit Ghosh wrote:
> hi jaques,
>
> in that case i like it (so it stays always on the top of the screen. i
> would still recommend for something simpler perhaps without the shading,
> just an outline of the box.
>
> cheers,
>
> satra
>
>
> On Fri, Jan 18, 2013 at 2:26
hi jaques,
in that case i like it (so it stays always on the top of the screen. i
would still recommend for something simpler perhaps without the shading,
just an outline of the box.
cheers,
satra
On Fri, Jan 18, 2013 at 2:26 PM, Jaques Grobler wrote:
> Regarding the banner, the first screens
Regarding the banner, the first screenshot I posted is actually a fixed one
that scrolls along with the page. I like the idea of it being less
disco-intense
but catches your eye as you scroll
2013/1/18 Mathieu Blondel
> On Sat, Jan 19, 2013 at 1:38 AM, Andreas Mueller
> wrote:
> > I don't see
Scikit Mailing List,
Can someone please elaborate on the difference between pre_dispatch and
njobs in
http://scikit-learn.org/dev/modules/generated/sklearn.grid_search.GridSearchCV.htmlspecifically,
what is a 'job' as used by pre_dispatch.
Thanks,
Ronnie
--
On Sat, Jan 19, 2013 at 1:38 AM, Andreas Mueller
wrote:
> I don't see the problem with semi-supervised algorithm actually. The API
> is pretty straight-forward
> and there are well-established algorithms like self-taught learning.
The API feels natural and is flexible but retrieving all labeled
e
+1 for removing HMM from 1.0 I would still like them accessible somewhere
though.
+1 for semi-supervised to be kept in 1.0 release. It doesn't seem very
developed. Seems self-explanatory.
On Fri, Jan 18, 2013 at 11:38 AM, Andreas Mueller
wrote:
> Am 18.01.2013 17:25, schrieb Lars Buitinck:
> >
Am 18.01.2013 17:25, schrieb Lars Buitinck:
> 2013/1/18 Mathieu Blondel :
>> Modules I would consider for removal are gaussian processes,
>> semi-supervised learning and HMMs. We could move them to a
>> scikit-learn-bleeding-edge repository.
> +1 for getting rid of the HMMs; I never liked their API
2013/1/18 Mathieu Blondel :
> Modules I would consider for removal are gaussian processes,
> semi-supervised learning and HMMs. We could move them to a
> scikit-learn-bleeding-edge repository.
+1 for getting rid of the HMMs; I never liked their API, and the
MultinomialHMM in particular is quite us
http://jsfiddle.net/Fgc39/1/
all controlled via the css 'alert'.
cheers,
satra
On Fri, Jan 18, 2013 at 10:18 AM, Lars Buitinck wrote:
> 2013/1/18 Jaques Grobler :
> > @Ronnie - here are some colours that don't really fit the theme versions
> >
> > pink http://i49.tinypic.com/286zac1.png
> > c
For the 1.0 release, I would concentrate on machine learning classics,
essentially supervised and unsupervised learning methods (i.e. the
methods for which our API is robust and well-tested). I would keep
structured prediction, semi-supervised learning, active learning or
anything that requires API
2013/1/18 Jaques Grobler :
> @Ronnie - here are some colours that don't really fit the theme versions
>
> pink http://i49.tinypic.com/286zac1.png
> cyan http://i50.tinypic.com/ao8d42.png
> green http://i46.tinypic.com/2rc05si.png
> red http://i50.tinypic.com/nd91le.png
-1, all of these stand out a
Btw, if we do keep 7 in some form, it should be checkbox, not a radio
button.
--
Master HTML5, CSS3, ASP.NET, MVC, AJAX, Knockout.js, Web API and
much more. Get web development skills now with LearnDevNow -
350+ hours of
Thanks
--
Olivier
http://twitter.com/ogrisel - http://github.com/ogrisel
--
Master HTML5, CSS3, ASP.NET, MVC, AJAX, Knockout.js, Web API and
much more. Get web development skills now with LearnDevNow -
350+ hours of step
hi jaques,
here is a quick fiddle to play around with the markup/colors. i just used
twitter bootstrap's alert css embedded it in a div after the header.
http://jsfiddle.net/Fgc39/
cheers,
satra
On Fri, Jan 18, 2013 at 9:43 AM, Jaques Grobler wrote:
> @Ronnie - here are some colours that don'
> Question 7:
> Question mark.
> I'm not sure question 7 is needed in its current form. If the options
> were broad concepts like "Coding quality", "Integration with other
> libraries", "scalability", "efficiency", then the question would work.
> As it stands, this question is answered by the a
+1 red or green since they stick out really obviously
On Fri, Jan 18, 2013 at 9:43 AM, Jaques Grobler wrote:
> @Ronnie - here are some colours that don't really fit the theme versions
>
> pink http://i49.tinypic.com/286zac1.png
> cyan http://i50.tinypic.com/ao8d42.png
> green http://i46.tinypic.
@xinfan - good question :) Not sure how long it would take yet - but I'll
make a note of that
2013/1/18 Jaques Grobler
> @Ronnie - here are some colours that don't really fit the theme versions
>
> pink http://i49.tinypic.com/286zac1.png
> cyan http://i50.tinypic.com/ao8d42.png
> green http://
Am 18.01.2013 15:40, schrieb xinfan meng:
> How long would the survey take? If it takes less than 2 minutes, then
> I think it is better to say "a one minute survey" instead of "a quick
> survey".
>
Not sure. Take it and we will know ;)
---
@Ronnie - here are some colours that don't really fit the theme versions
pink http://i49.tinypic.com/286zac1.png
cyan http://i50.tinypic.com/ao8d42.png
green http://i46.tinypic.com/2rc05si.png
red http://i50.tinypic.com/nd91le.png
Stil using layout from Satrajit's suggestion here
J
2013/1/18 R
How long would the survey take? If it takes less than 2 minutes, then I
think it is better to say "a one minute survey" instead of "a quick survey".
On Fri, Jan 18, 2013 at 10:29 PM, Ronnie Ghose wrote:
> h what about one that does not share color with the scikit webpage at
> all. it would l
h what about one that does not share color with the scikit webpage at
all. it would look horrible - and as a result stand out
0 - neutral
On Fri, Jan 18, 2013 at 9:27 AM, Andreas Mueller
wrote:
> Am 18.01.2013 15:06, schrieb Ronnie Ghose:
>
> +1 to a different color. The blue blends i
Yeah the first colour was just a random choice :) Was just to get things
going with this
2013/1/18 Andreas Mueller
> Am 18.01.2013 15:06, schrieb Ronnie Ghose:
>
> +1 to a different color. The blue blends in with the bg.
>
> +1
>
>
> --
Am 18.01.2013 15:06, schrieb Ronnie Ghose:
+1 to a different color. The blue blends in with the bg.
+1
--
Master HTML5, CSS3, ASP.NET, MVC, AJAX, Knockout.js, Web API and
much more. Get web development skills now with L
Sorry about the embedded one - wasn't thinking there. French weather is
freezing my brain
Here's the image I sent earlier http://i48.tinypic.com/24xgcd0.png
Here's the new one with Satrajit's suggestions
http://i45.tinypic.com/11uztbs.png
2013/1/18 Satrajit Ghosh
> sending without embedded
+1 to a different color. The blue blends in with the bg.
On Fri, Jan 18, 2013 at 8:57 AM, Satrajit Ghosh wrote:
> sending without embedded image.
>
> On Fri, Jan 18, 2013 at 8:56 AM, Satrajit Ghosh wrote:
>
>> hi jaques,
>>
>> looks good. two possible options
>>
>> 1. put it right below the cu
sending without embedded image.
On Fri, Jan 18, 2013 at 8:56 AM, Satrajit Ghosh wrote:
> hi jaques,
>
> looks good. two possible options
>
> 1. put it right below the current banner (across the page, below the
> logo).
> 2. change the background to the orange or red. something that brings
> atte
cppcheck reported:
[sklearn/svm/src/libsvm/svm.cpp:1880] -> [sklearn/svm/src/libsvm/svm.cpp:1875]:
(style) Found duplicate branches for 'if' and 'else'.
Signed-off-by: Thomas Jarosch
---
sklearn/svm/src/libsvm/svm.cpp | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
I just downloaded the
cppcheck reported:
[sklearn/svm/src/liblinear/linear.cpp:2240]: (error) Mismatching allocation and
deallocation: C
[sklearn/svm/src/liblinear/linear.cpp:2258]: (error) Mismatching allocation and
deallocation: C
[sklearn/svm/src/liblinear/linear.cpp:2303]: (error) Mismatching allocation and
deall
On Fri, Jan 18, 2013 at 4:08 PM, Olivier Grisel wrote:
> Thanks for the details. So if I understand correctly you can only see
> the problem on a pipeline that is loaded from a joblib pickle on the
> harddrive?
>
> If so you should be able to reproduce the issue with the 20 newsgroups
> dataset.
>
Thanks for the details. So if I understand correctly you can only see
the problem on a pipeline that is loaded from a joblib pickle on the
harddrive?
If so you should be able to reproduce the issue with the 20 newsgroups dataset.
Could you please try to push a self-hosting script that uses the 20
32 matches
Mail list logo
