Hi David,
One simple suggestion: Did you build numpy, then scipy, and then
scikit-learn using their respective make/setup.py?
Virgile
On Mon, Sep 17, 2012 at 5:31 PM, David Montgomery davidmontgom...@gmail.com
wrote:
Hi,
I upgraded numpy to 1.6 and sci to .12 and now I get the below. I am
Gael-luke... answer faster than its shadow ;)
On Mon, Sep 17, 2012 at 5:39 PM, Virgile Fritsch
virgile.frit...@gmail.comwrote:
Hi David,
One simple suggestion: Did you build numpy, then scipy, and then
scikit-learn using their respective make/setup.py?
Virgile
On Mon, Sep 17, 2012 at 5
Hi list,
I got this error when pulling the last master state on my laptop:
~/scikit-learn/sklearn/svm [master] $ nosetests
.E
==
ERROR: Check that primal coef modification
Hi,
It may come from the fact that np.dot(X, X.T) is not invertible.
Is it possible that two (or more) features are correlated? Or that one (or
more) features have a 0 variance?
If yes, try to remove those features.
Let us know if you think this could be the reason why it fails.
Virgile
On
It may come from the fact that np.dot(X, X.T) is not invertible.
I meant, of course, np.dot(X.T, X).
Would you look at the eigenvalues of this thing please?
On Thu, Jun 28, 2012 at 10:33 AM, Virgile Fritsch virgile.frit...@gmail.com
wrote:
Hi,
It may come from the fact that np.dot(X, X.T
-14, -4.21817181e-15])
From: Virgile Fritsch virgile.frit...@gmail.com
Sent: Thu, 28 Jun 2012 17:44:35
To: scikit-learn-general@lists.sourceforge.net
Subject: Re: [Scikit-learn-general] EllipticEnvelope giving error
IndexError: 0-d arrays can't be indexed
It may come from the fact that np.dot(X
Hi all,
Going back to the .mailmap file, I would like to understand what has
changed and how Nelle has done it so that next time I may help.
Could someone please explain that to (noob) me?
Virgile
On Fri, Jun 1, 2012 at 6:54 AM, Gael Varoquaux
gael.varoqu...@normalesup.org wrote:
On Thu,
Nelle added better matching of the names to id:
And how did she achieve that? I mean, how did she know about the different
names that were used in commits?
On Fri, Jun 1, 2012 at 9:03 AM, Gael Varoquaux
gael.varoqu...@normalesup.org wrote:
On Fri, Jun 01, 2012 at 08:12:34AM +0200, Virgile
Having a look at the git shortlog -se listing, I would say that
duplicates handling requires a human knowledge about the actual set of
contributor (who would guess than there is not any Virgile but I, and that
Virgile and Virgile Fritsch should hence be merged?). Reminds me of the
Gmail find
Hi Andreas,
Indeed, it should be envelope with an e at the end.
The algorithm fits a robust covariance object to the data, compute the
(robust) observations' Mahalanobis distances from it and sets a threshold
on these distances so that a given proportion of observations are removed.
I suggest
Hi Olivier,
I don't know that work, neither the authors, so I will have a look at it.
As far as I understand from the abstract, observations should have a
low-dimensional structure in order this technique to be applicable, which
is not necessarily the case in practice.
I will read the paper
OK, but if basics methods are likely to become more efficient than already
existing ones, they should be included in more general packages at some
point (i.e. fast_svd may end into scipy.linalg).
--
Virgile Fritsch
On Mon, Dec 19, 2011 at 1:57 PM, xinfan meng mxf3...@gmail.com wrote:
Good
Virgile? I thought that you had almost isolated a simple test case.
Not really, I could have the graph_lasso crash but I was using a
configuration in which my n_features-dimensional observations actually lied
in a (n_features - 1)-dimensional subspace.
Currently, I am using the graph_lasso to
Reminds me of the PR by Robert about performing clustering from similarity
matrix or directly from the data.
So I would be in favour of having a X_is_cov keyword.
Sorry for biasing the discussion with cov_init, I answered to quikly ;)
On Wed, Nov 9, 2011 at 5:16 PM, Gael Varoquaux
14 matches
Mail list logo