59b5602346
DOC more accurate comment in class weight computation
2013-01-12 18:37:08 +01:00
bd5e502278
P3K fix incorrect import
...
(it breaks e.g. "import sklearn.metrics" under Python 3.3)
2013-01-11 05:02:51 +06:00
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
7819e182e2
DOC try to be as clear as possible.
2013-01-05 19:59:46 +01:00
de9bcc9fdb
FIX do checks even if allow_lists
2013-01-05 19:59:46 +01:00
03048ddb5f
ENH more logical behavior, better docstring, tests
2013-01-05 19:59:46 +01:00
76c66a0dab
BUG Cholesky delete routines wouldn't compile on Solaris
...
Fixed that and cleaned it up. Should solve #1510 .
2013-01-03 15:03:26 +01:00
1c1538af87
ENH move utility function into dedicated file, not __init__.py
2013-01-03 13:28:10 +01:00
956975a70e
DOC changed docstring to be more clear.
2013-01-03 13:21:51 +01:00
3be8a23d2d
ENH addressed @ogrisel's comments.
2013-01-03 13:21:51 +01:00
843d795a94
DOC docstring for compute_class_weight
2013-01-03 13:21:51 +01:00
127c44b730
ENH refactoring class weights for SVM and SGD
2013-01-03 13:21:50 +01:00
b91f5c03a8
FIX i should trust my past self a bit more
2012-12-24 15:42:58 +01:00
5b948bbcf3
TEST flipped svd remains correct
2012-12-24 15:04:36 +01:00
641d0a7e5b
Cannot easily ensure non-uniqueness without the fix, just test uniqueness
2012-12-24 15:03:49 +01:00
5aed364db5
svd_flip test fails on Travis. Change random seed, see if it helps
2012-12-24 15:03:49 +01:00
4040d42830
Make randomized_svd flipped by default
2012-12-24 15:03:49 +01:00
160648d8cc
Make sure svd_flip test actually tests something
2012-12-24 15:03:49 +01:00
0755f156f8
Add sign flip as flag in randomized_svd
2012-12-24 15:03:49 +01:00
9ca8f7e60a
Test more seeds for svd sign flipping
2012-12-24 15:02:24 +01:00
cf5d34e776
Make sign flip in place
2012-12-24 15:02:24 +01:00
d5857bbb60
Sign flipping as suggested by @ogrisel, not in place
2012-12-24 15:02:24 +01:00
0df8fb20ba
Add failing test and no-op flip
2012-12-24 15:02:24 +01:00
cc63e1cbb4
ENH slight improvement of common tests.
2012-12-24 14:58:56 +01:00
966e10c252
COSMIT PEP8
2012-12-22 16:48:17 +01:00
754c2d9533
COSMIT pep8
2012-12-22 16:48:17 +01:00
a0860b114d
FIX don't use assert_allclose. It is not supported in numpy 1.3
2012-12-21 15:10:13 +01:00
db4422fcf9
ENH Simplify assert_raise_message + TST add them
2012-12-21 14:48:01 +01:00
07881645a6
COSMIT make everything use the common interface
2012-12-21 14:48:01 +01:00
7732a5f49e
pep8
2012-12-21 14:48:01 +01:00
50368c3417
COSMIT
2012-12-21 14:48:00 +01:00
a6ed58dc4d
ENH flatten random_projection module + add sklearn.utils.random
2012-12-21 14:48:00 +01:00
472902452e
ENH Add allclose to sklearn.testing
2012-12-21 14:47:57 +01:00
e69281fd9a
ENH assert_raise_message check for substring existence instead of equality
2012-12-21 14:34:23 +01:00
60a86cc248
FIX: raise ValueError when n_components is too large
2012-12-21 14:34:19 +01:00
aac12fab1c
ENH allow lists in check_arrays
2012-12-21 10:33:48 +01:00
67fccaca6a
Merge pull request #1458 from seberg/contig_strides
...
FIX: Do not rely on strides for contiguous arrays
2012-12-11 23:53:37 -08:00
210767d6d3
ENH don't import * in utils __init__.py
2012-12-11 00:04:57 +01:00
1a83c16b2b
FIX: Do not rely on strides for contiguous arrays
...
When an array is contiguous in memory but has `shape[dim] == 1`, then
its `strides[dim]` is not really used, so it can be considered arbitrary
even if the array is contiguous. This is generally true for 0-sized arrays.
If there is no chance of 0-sized (or 1-sized and 1-dimensioal) arrays, then
this code is perfectly correct in current numpy. But its a dangerous
design choice and it would be nice to be able to change numpy at some point.
2012-12-10 16:33:55 +01:00
755164f1a2
check in ``if_matplotlib`` if $DISPLAY set
2012-12-02 17:35:13 +01:00
bf752ddf54
use Agg backend with warn=False for matplotlib enabled tests
2012-12-02 17:26:17 +01:00
aff7da7834
Merge remote-tracking branch 'upstream/master' into gbrt-interactions
2012-11-30 09:29:27 +01:00
df9fc91ca2
Add missing assertion.
2012-11-29 17:57:44 +09:00
fad7143c86
COSMIT refactor input validation code; skip some issparse calls
2012-11-28 01:02:06 +01:00
7f73ce6dfb
test decorator to skip tests if matplotlib cannot be imported
2012-11-27 09:27:30 +01:00
66df36291f
Merge branch 'master' into gbrt-interactions
2012-11-25 21:27:16 +01:00
3c46ebaa6e
COSMIT pep8, removing unused imports and recommend ``toarray`` instead of ``todense``
2012-11-24 20:12:33 +00:00
dd04394039
BUG: fix murmurhash compilation with recent Cython
2012-11-22 22:36:03 +01:00
a7ebd1b72c
Merge branch 'master' into gbrt-interactions
2012-11-19 19:37:34 +01:00
2b8ab3561a
ENH feature hashing transformer
2012-11-17 17:29:20 +01:00
Andreas Mueller