Commit Graph

72 Commits

Author SHA1 Message Date
Lars Buitinck 551b6c2f06 ENH honor Y_norm_squared when X=Y in euclidean_distances
Also a micro-optimization to row_norms.
2013-10-13 18:54:21 +02:00
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 edf747e6a2 DOC docstring for extmath.norm 2013-10-12 20:17:33 +02:00
dengemann c1df89edf2 ... add the fix to import check 2013-10-08 23:01:48 +02:00
dengemann add591cbf7 FIX scipy API 2013-10-08 22:54:14 +02:00
Lars Buitinck ceffed1d6e MAINT remove useless deprecation in sklearn.utils
This module is off-limits to users anyway.
2013-09-04 18:43:56 +02:00
balu fcce3032c0 fix visual indentation 2013-08-31 12:27:16 +05:30
dengemann fbd6d9c8f6 ENH: update version checking 2013-08-25 12:17:33 +02:00
dengemann f91fba2f01 ENH: restrict fast_dot to np < 1.8 2013-08-25 11:16:22 +02:00
dengemann 1322da977c ENH/FIX: address @ogrisel comments 2013-08-25 11:16:21 +02:00
dengemann f36cba5b05 FIX: superfluous check 2013-08-25 11:16:21 +02:00
dengemann b86f3d5208 ENH: add new Warning class, improve tests, update docs 2013-08-25 11:16:21 +02:00
dengemann dcd6292095 ENH: cover another cornercase + fix messages 2013-08-25 11:16:21 +02:00
dengemann dbc751fd50 ENH: add fast_dot shape check 2013-08-25 11:16:21 +02:00
dengemann 69fc2970e7 FIX: remove spurious test 2013-08-25 11:16:21 +02:00
dengemann bc15f3019d CLEANUP 2013-08-25 11:16:21 +02:00
dengemann 5a3a432a79 ENH: return np.dot if ndim does not match 2013-08-25 11:16:21 +02:00
dengemann e98d1ccf42 FIX: workaround for missing BLAS 2013-08-25 11:16:21 +02:00
dengemann 19e0401c12 employ DataConversionWarning 2013-08-25 11:16:20 +02:00
dengemann 2240e44f31 ENH: add fast_dot function 2013-08-25 11:16:20 +02:00
Lars Buitinck 74ec752e3c ENH speed up logistic_sigmoid (using less code)
Timings, before:

>>> x = np.linspace(-1000, 1000, 10000)
>>> %timeit logistic_sigmoid(x)
1000 loops, best of 3: 570 us per loop

After:

>>> %timeit logistic_sigmoid(x)
1000 loops, best of 3: 286 us per loop

For comparison:

>>> %timeit np.tanh(x)
10000 loops, best of 3: 117 us per loop
>>> %timeit .5 * (1 + np.tanh(x/2.))
1000 loops, best of 3: 301 us per loop

TODO: optimize log_logistic_sigmoid.
2013-07-28 16:44:13 +02:00
Kemal Eren 395b231774 moved and renamed _make_nonnegative() and _safe_min() 2013-07-25 19:19:19 +02:00
syhw ac1e1a9c87 squeezing logistic_sigmoid result only on 1D arrays 2013-07-24 17:27:23 +02:00
Vlad Niculae e488649af4 ENH Add fast and stable logistic sigmoid to utils and RBM 2013-07-24 17:27:22 +02:00
Lars Buitinck 77695ee27a COSMIT pep8 2013-07-11 11:50:23 +02:00
Lars Buitinck 34815a35c8 COSMIT omit unused parameter/return value in svd_flip 2013-06-12 13:39:39 +02:00
Robert Layton dacfd8bd5d DOC: Replaced all BSD style licenses with "BSD 3 clause"
Replaced all BSD style licenses with "BSD 3 clause"
Not checked yet!

Removed duplicate "3 clause, 3 clause"

Removed trailing period if exists

Fixed some missed licences, still about 50 to do, but those can be automated

Think I got the last of them.

Apparently me and sed have different ideas of regex.

Found a few more
2013-04-30 08:34:46 +02:00
Lars Buitinck 541987802f P3K range vs. xrange 2013-02-14 02:05:35 +01:00
Vlad Niculae 4040d42830 Make randomized_svd flipped by default 2012-12-24 15:03:49 +01:00
Vlad Niculae 0755f156f8 Add sign flip as flag in randomized_svd 2012-12-24 15:03:49 +01:00
Vlad Niculae cf5d34e776 Make sign flip in place 2012-12-24 15:02:24 +01:00
Vlad Niculae d5857bbb60 Sign flipping as suggested by @ogrisel, not in place 2012-12-24 15:02:24 +01:00
Vlad Niculae 0df8fb20ba Add failing test and no-op flip 2012-12-24 15:02:24 +01:00
Andreas Mueller 754c2d9533 COSMIT pep8 2012-12-22 16:48:17 +01:00
Peter Prettenhofer 52ba220aae empty instead of zeros 2012-11-11 18:12:43 +01:00
Peter Prettenhofer 0530c106c8 pep8 2012-11-10 12:59:46 +01:00
Peter Prettenhofer 77b585f9d2 Merge branch 'master' into gbrt-interactions 2012-10-22 08:24:18 +02:00
Mathieu Blondel a2cd5fa94c Fix density utility when input is sparse. 2012-10-19 16:53:36 +09:00
Peter Prettenhofer 35860aa42d Merge branch 'master' into gbrt-interactions
Conflicts:
	sklearn/ensemble/gradient_boosting.py
2012-10-19 09:21:10 +02:00
Peter Prettenhofer 4bd2e9b5f8 add cartesian product 2012-10-14 19:58:50 +02:00
Andreas Mueller 9bd0aadaa3 MISC renamed n_iterations to n_iter in all other places. 2012-10-10 19:10:29 +01:00
Jake Vanderplas 776b060098 @jakevdp's version of pinvh
speed up symmetric_pinv
2012-09-04 19:58:46 +01:00
Vlad Niculae bfac9f2597 Vectorize singular value inversion 2012-09-04 19:58:46 +01:00
Vlad Niculae 4a5a1d6d84 Compute pseudoinverse using eigendecomposition 2012-09-04 19:58:46 +01:00
Vlad Niculae 2516b6217f Remove fast_svd.
This was deprecated in version 0.9.
2012-09-01 20:02:13 +02:00
Andreas Mueller eba29c2116 FIX copy/paste error 2012-05-04 23:16:31 +02:00
Andreas Mueller 667ef37714 DOC move references from Notes to References section in docstrings 2012-03-04 15:16:51 +01:00
Andreas Mueller 5c2a8696e3 COSMIT removed unused imports 2012-01-23 21:13:55 +01:00
Olivier Grisel 0f6c525d7e ENH: renamed fast_svd to randomized_svd + related improvements 2011-12-31 14:31:55 +01:00
Olivier Grisel 73e65c2107 PEP257: docstring cosmits in utils.extmath 2011-12-30 22:52:35 +01:00