I did implement a weighted mean shift (see issue 442 on the github
repository). It may or may not be weighted in the sense that you mean.
Rather only having a flat kernel, there is an option to use a gaussian
kernel. I did go so far as to make a pull request, but we decided my
branch needed a bit more work to not change the previous interface. I
never got around to making those small changes, and so it has not been
merged.
Here's a link to my branch with the gaussian kernel:
https://github.com/conradlee/scikit-learn/tree/442-meanshift-gauss
On Wed, Apr 11, 2012 at 5:51 PM, Alexandre Gramfort <
[email protected]> wrote:
> a sample_weight param seems reasonable to me
>
> Alex
>
> On Wed, Apr 11, 2012 at 5:10 PM, Olivier Grisel
> <[email protected]> wrote:
> > Le 11 avril 2012 16:59, Michael Selik <[email protected]> a écrit :
> >> Certainly. It looks like a good approach would be to break out line 121
> in mean_shift_.py:
> >>> my_mean = np.mean(points_within, axis=0)
> >>
> >> And provide a function instead that allows several methods of mean
> calculation -- flat kernel (current method), gaussian kernel, and/or
> accuracy-weighted kernel.
> >>
> >> Any thoughts before I get started?
> >
> > Have a look at other estimators that use stuff like precomputed
> > kernels, class_weight and sample_weight and try to reuse the idioms of
> > the rest of the library where applicable for consistency.
> >
> > git grep precomputed
> > git grep kernel
> > git grep class_weight
> > git grep sample_weight
> >
> > --
> > Olivier
> > http://twitter.com/ogrisel - http://github.com/ogrisel
> >
> >
> ------------------------------------------------------------------------------
> > Better than sec? Nothing is better than sec when it comes to
> > monitoring Big Data applications. Try Boundary one-second
> > resolution app monitoring today. Free.
> > http://p.sf.net/sfu/Boundary-dev2dev
> > _______________________________________________
> > Scikit-learn-general mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
>
> ------------------------------------------------------------------------------
> Better than sec? Nothing is better than sec when it comes to
> monitoring Big Data applications. Try Boundary one-second
> resolution app monitoring today. Free.
> http://p.sf.net/sfu/Boundary-dev2dev
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
------------------------------------------------------------------------------
For Developers, A Lot Can Happen In A Second.
Boundary is the first to Know...and Tell You.
Monitor Your Applications in Ultra-Fine Resolution. Try it FREE!
http://p.sf.net/sfu/Boundary-d2dvs2
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
