27;s working for me.
>>>>>
>>>>> Thanks again,
>>>>>
>>>>> Ariel
>>>>>
>>>>>
>>>>> On Thu, Jan 24, 2013 at 5:35 AM, Bertrand Thirion <
>>>>> bertrand.thir...@inria.fr> wrote:
Hi Denis,
On Thu, Jan 31, 2013 at 8:48 AM, denis wrote:
> Ariel,
> what's k, how many data points do you have ?
>
I have something between approximately 2 and 150 data points.
What's k? I guess that's my next question, right? For now, I am trying to
use an AIC criterion to determine how hig
Ariel,
what's k, how many data points do you have ?
There's a trivial k-means under
http://stackoverflow.com/questions/5529625/is-it-possible-to-specify-your-own-distance-function-using-scikits-learn-k-means
"verbose=1" gives you some idea of how clusters converge (or not).
(If you have even a
by minimizing the sum of squared distances to a given set of points within
>>>>> each cluster. However, it is true that re-projecting the euclidean mean
>>>>> to
>>>>> the sphere would approximate well the theoretical solution in most cases.
>>&g
t;> by minimizing the sum of squared distances to a given set of points within
>>>> each cluster. However, it is true that re-projecting the euclidean mean to
>>>> the sphere would approximate well the theoretical solution in most cases.
>>>>
>>>> A standard alternative
andard alternative to k-means is Von Mises-Fisher distribution.
>>>
>>> Bertrand
>>>
>>> --
>>>
>>> *De: *"Vince Fernando"
>>> *À: *math...@mblondel.org, scikit-learn-general@lists.sourceforge.net
gt;> Bertrand
>>
>> --
>>
>> *De: *"Vince Fernando"
>> *À: *math...@mblondel.org, scikit-learn-general@lists.sourceforge.net
>> *Envoyé: *Jeudi 24 Janvier 2013 09:55:46
>> *Objet: *Re: [Scikit-learn-general] K mean
;
> --
>
> *De: *"Vince Fernando"
> *À: *math...@mblondel.org, scikit-learn-general@lists.sourceforge.net
> *Envoyé: *Jeudi 24 Janvier 2013 09:55:46
> *Objet: *Re: [Scikit-learn-general] K means on a sphere
>
>
> Are there any theoretica
.
A standard alternative to k-means is Von Mises-Fisher distribution.
Bertrand
- Mail original -
> De: "Vince Fernando"
> À: math...@mblondel.org, scikit-learn-general@lists.sourceforge.net
> Envoyé: Jeudi 24 Janvier 2013 09:55:46
> Objet: Re: [Scikit-learn-genera
On Jan 24, 2013 5:35 PM, "Charles-Pierre Astolfi" wrote:
>
> There's no projection that conserves the distance wrt to any pair of
> points on the sphere (although there are some that conserves the
> distance wrt 1 or 2 specific points on the sphere)
>
> BUT the gnomonic project conserves the short
I'm a noob when it comes to data on a sphere, but is there any issue
with preprocessing the data to project it on a place, run kmeans in
the plane and the reproject it back on the sphere?
There's no projection that conserves the distance wrt to any pair of
points on the sphere (although there are
hi Ariel,
what I would do first, if the data are not too big, is reimplement my kmeans in
10 lines and after you update the centers, normalize them to put them back
on the sphere. I don't think you can say much about convergence but
it might work well enough in practice.
HTH
Alex
On Thu, Jan 24,
Are there any theoretical problems if one uses the great circle
(orthodromic) distance on a sphere in k-means or any other clustering
algorithm?
vince
On 24 January 2013 07:11, Mathieu Blondel wrote:
> On Thu, Jan 24, 2013 at 9:24 AM, Gael Varoquaux
> wrote:
>
> > Yes, there is a massive diffe
On Thu, Jan 24, 2013 at 9:24 AM, Gael Varoquaux
wrote:
> Yes, there is a massive difference in amount of work and performance when
> you try to replace the Euclidean distance. Amongst other problems, the
> mean is no longer the sum divided by the number of points, but the
> Frechet mean, which re
hi Ariel,
what I would do, if the data are not too big, is reimplement my kmeans in
10 lines and after you update the centers, normalize them to put them back
on the sphere. I don't think you can say much about convergence but
it might work in practice.
HTH
Alex
On Thu, Jan 24, 2013 at 1:24 AM,
On Thu, Jan 24, 2013 at 12:34:31AM +0100, Andreas Mueller wrote:
> Sorry, custom metrics for K means are not possible at the moment.
Yes, there is a massive difference in amount of work and performance when
you try to replace the Euclidean distance. Amongst other problems, the
mean is no longer th
Hi Ariel.
Sorry, custom metrics for K means are not possible at the moment.
If you wanted to tweak the sklearn implementation, you would have to
look into this file:
https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/cluster/k_means_.py#L413
In particular the function _labels_inert
Hi everyone,
I am interested in using the sklearn implementation of k means to estimate
clusters of unit vectors on the surface of a sphere.
This requires that the distance metric be changed from the current
Euclidean distance metric to angles.
Is there any easy way to achieve that with the curr
18 matches
Mail list logo
