Re: [Scikit-learn-general] Interpretation of LogisticRegression coefficients in multiclass case

Hi Michael, > I agree it would be nice to see a multinomial logistic regreesion in the > scikit :).  Any plans for it? not on my side but this question comes so often that it might become a priority for the project. > Is your warning about interpreting the weights just because in a penalized > m

Re: [Scikit-learn-general] Cython public API

2012/3/6 Olivier Grisel > 2012/3/6 Mathieu Blondel : > > Hello, > > > > This was probably obvious but I recently realized that Cython pyd > > files can be shared across projects. For example, one can do: > > > >from sklearn.linear_model.sgd_fast cimport WeightVector > > > > For it to work, on

Re: [Scikit-learn-general] Cython public API

2012/3/6 Mathieu Blondel : > Hello, > > This was probably obvious but I recently realized that Cython pyd > files can be shared across projects. For example, one can do: > >    from sklearn.linear_model.sgd_fast cimport WeightVector > > For it to work, one currently has to compile the corresponding

[Scikit-learn-general] Cython public API

Hello, This was probably obvious but I recently realized that Cython pyd files can be shared across projects. For example, one can do: from sklearn.linear_model.sgd_fast cimport WeightVector For it to work, one currently has to compile the corresponding pyx file with the -I option to indicat

[Scikit-learn-general] Text feature extraction: documentation and refactoring

Hi all, I think this PR is ready for a round of final review: https://github.com/scikit-learn/scikit-learn/pull/668 -- Olivier http://twitter.com/ogrisel - http://github.com/ogrisel -- Try before you buy = See our ex

Re: [Scikit-learn-general] Interpretation of LogisticRegression coefficients in multiclass case

Hi Alex, I agree it would be nice to see a multinomial logistic regreesion in the scikit :). Any plans for it? Is your warning about interpreting the weights just because in a penalized model they're not best estimators? Mostly I just want to take a look to see in general what's happening -- e.g

Re: [Scikit-learn-general] Not all plots generated on website

I know that several of the plots fail with Matplotlib < 1.0. It's some problem that comes up when certain lines have a dashed linestyle. Jake Andreas wrote: > Hi everybody. > I noticed that not all example plots are present on the > website. > Does anyone know why that is? > I would guess tha

Re: [Scikit-learn-general] Is there a branch support sparse feature decision tree?

On Mon, Mar 5, 2012 at 3:15 PM, Jinhui Li wrote: > The fit function of basedecisionregressor convert the X to dense format. > > In my case, there are 100M train samples, 30K features, most data are > zeros. No one is working on sparse decision trees AFAIK. I'd suggest trying a dimensionality redu

Re: [Scikit-learn-general] Is there a branch support sparse feature decision tree?

Hi, Unfortunately, our implementation of decision trees indeed does not support sparse datasets. Best, Gilles On 5 March 2012 15:15, Jinhui Li wrote: > Hi there, > > Again, I have a question about tree.py. > > The fit function of basedecisionregressor convert the X to dense format. > > In my c

Re: [Scikit-learn-general] Is there a branch support sparse feature decision tree?

2012/3/5 Jinhui Li : > Hi there, > > Again, I have a question about tree.py. > > The fit function of basedecisionregressor convert the X to dense format. > > In my case, there are 100M train samples, 30K features, most data are > zeros. > > please give some suggestions? AFAIK nobody has been worki

[Scikit-learn-general] Is there a branch support sparse feature decision tree?

Hi there, Again, I have a question about tree.py. The fit function of basedecisionregressor convert the X to dense format. In my case, there are 100M train samples, 30K features, most data are zeros. please give some suggestions? -

[Scikit-learn-general] Not all plots generated on website

Hi everybody. I noticed that not all example plots are present on the website. Does anyone know why that is? I would guess that some version of Matplotlib is used that doesn't work with the examples. If there was a way to find out which version was installed, we could easily test that (and make sur