Commit Graph

18 Commits

Author SHA1 Message Date
Lars Buitinck 24af95a211 ENH refactor squared-norms computation to extmath
k-means can now use the memory-efficient dense implementation from
metrics.pairwise, while pairwise can use the fast sparse implementation
from k-means.
2013-10-12 20:27:51 +02:00
Lars Buitinck eff660b0e2 ENH optimize KMeans for sparse inputs
In the document clustering example, the speedup is typically between
15× and 150×.
2013-07-08 14:18:26 +02:00
Stefano Lattarini 67ece78da4 COSMIT various typofixes
As suggested by codespell <https://github.com/lucasdemarchi/codespell>

Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
2013-04-11 21:22:14 +02:00
Lars Buitinck b37791bab9 ENH cut MiniBatchKMeans memory usage in half for large n_clusters
Also, changed logging message to reflect the (public) class that is logging,
not the (private) function.

New Cython code lives in utils.sparsefuncs, because it may prove more
generally useful.
2013-03-13 15:23:50 +01:00
Lars Buitinck 37c1d092f8 COSMIT shut up the build by calling np.import_array in Cython modules
$(make in 2>&1 | wc -l) down from 706 to 517.

Took the liberty of modernizing some of the Cython code while I was at it.
2013-01-10 18:00:21 +01:00
Alexis Mignon 4842ebb16b Unified access to 'mean_variance_axis0' for CSC and CSR matrices 2012-05-16 07:58:02 +02:00
Alexis Mignon cc01c12eb4 Regenerated with Cython 2012-05-16 07:57:51 +02:00
Alexis Mignon fb9d18d2bb Removed duplicate definition of csc_mean_variance_axis0 2012-05-16 07:57:18 +02:00
Alexis Mignon a8ed0da41d Added utility functions for csc sparse matrices 2012-05-16 07:57:09 +02:00
Olivier Grisel b2dac63feb FIX: unused cython variable 2011-12-27 10:04:53 +01:00
Olivier Grisel 1fdbbfb44b WIP: feature scaling for CSR input (lacks some tests) 2011-12-24 12:37:09 +01:00
Olivier Grisel 0cf16dc7be merge master 2011-12-23 14:28:50 +01:00
Mathieu Blondel 6e619e22bd Optimize for high dimensional data. 2011-12-21 15:56:36 +01:00
Olivier Grisel e51ceb17bb WIP: scaling CSRs 2011-12-20 15:59:28 +01:00
Olivier Grisel 494fb029cb consolidating all cython utils for sparse CSR in the same file under utils 2011-12-20 13:18:03 +01:00
Mathieu Blondel c696ad4026 Use int for indptr and indices. 2011-12-20 10:21:30 +01:00
Mathieu Blondel 8311b1bbd8 Fix bug with sparse matrices. 2011-12-19 16:28:53 +01:00
Mathieu Blondel 64adbcaddd Implement mean_variance_axis0. 2011-12-19 16:28:21 +01:00