a87011262a
FIX precision to float64 across the codebase
2015-10-16 11:20:29 +02:00
2f099334f9
MAINT deprecate 1d input arrays for all estimators
...
Passing 1D arrays to check_array, without setting `ensure_2d` to false now
raises a deprecation warning before reshaping it. This will later throw an
error.
All Scaler classes also throw warnings when 1D arrays are passed.
All unit tests/doctests are modified to ensure that no 1D arrays are passed,
except in explicit 1D array tests where the warnings have been silenced.
Additional tests are also included which check for different 1D array cases.
2D array tests with one samples and one features are also added and where
they failed, `check_array` call has been modified to give a more useful error
message
2015-09-09 15:49:58 +02:00
0650d5502e
DOC adding backlinks to docstrings
2015-06-03 00:24:04 -04:00
cd2ee7e454
MAINT docstring --> comments to prevent nose from using doc in verbose mode
2015-03-21 11:16:49 +05:30
270d47a87b
Fix gibbs sampling behavior in RBM with integral random_state.
2015-02-10 15:07:17 -05:00
cd63accc05
Add test that score takes y, fix KMeans, FIX pipeline compatibility of clustering algorithms!
2015-02-06 13:26:42 +01:00
abd31d2cad
MAINT Make uniform the error raised for not fitted condition
2015-01-12 00:33:40 +05:30
24c0efd12f
PEP8 Fix E112 and E113 errors
2015-01-08 04:00:28 +05:30
b3bdb08964
DOC fix formatting of attributes etc. in docstrings
2014-07-28 19:04:59 +10:00
c235c3b566
ENH add allowed_sparse named argument for @ogrisel
2014-07-20 15:31:28 +02:00
6e2a83b4e1
remove check_arrays stuff and old input validation
2014-07-20 13:31:45 +02:00
c7db70ac9f
removed unused variables and imports, add used imports where missing.
2014-06-08 21:44:01 +02:00
d12d3132c4
Fix some fun column span alignment errors.
2014-06-07 16:48:48 +02:00
91a45af295
ENH micro-optimize a few tests
2014-04-04 11:27:07 +02:00
3c753a30c5
Merge conflict in whats_new.rst
2014-03-26 11:37:09 +01:00
7ed792f0a8
adding test for csc and csr sparse matrix
2014-03-24 20:14:51 -04:00
696cf55792
removing tests for multiple sparse matrix formats
2014-03-22 09:01:13 -04:00
874235c5b9
removing copy() and adding tests for multiple forms of sparse matrices
2014-03-21 21:29:01 -04:00
eb037dde8a
pep8-ing
2014-03-20 19:22:17 -04:00
5c120f585c
adding test for csr data for partial fit
2014-03-20 19:16:37 -04:00
d576ec8e79
pep8 corrections
2014-03-10 09:25:58 -04:00
0afa59673d
changing computation of batch slices to np.array_split
2014-03-09 22:17:12 -04:00
c9e4cd079d
ENH more optimizations for RBM
...
Shaved off another ~12% on the digits example; the biggest part of the
savings is in _sample_visibles.
2014-03-03 00:47:24 +01:00
a83fddf01a
Merge remote-tracking branch 'upstream/master' into rbmbranch
2014-03-01 14:42:37 -05:00
b3a8759222
changing formatting of documentation
2014-03-01 14:34:25 -05:00
0ba1e28cd5
ENH speed up RBM training with scipy.special.expit
...
SciPy 0.10 already has an implementation of the logistic function called
scipy.special.expit.
Using this makes RBM training 12% faster, as measured by observing the time
per iteration as reported by the plot_rbm_logistic_classification.py example,
disregarding the first iteration as an outlier (it's consistently faster than
the rest, not sure why). The speedup is in the expit function, not the
inplace operations; those are there to make benchmarking easier, but they
certainly won't hurt.
2014-02-23 17:46:53 +01:00
0c22cbb272
FIX numerical stability issue in BernoulliRBM
...
Fixes #2785 by replacing log1p(exp(x)) with logaddexp(0, x).
2014-02-18 19:41:43 +01:00
016469b30e
changing xrange to range
2014-02-05 20:12:01 -05:00
b928cdb3e0
making adjustments to variable names and account for sparse matrices
2014-02-05 19:47:15 -05:00
dbc26ce996
COSMIT remove unused imports and variables
2014-02-03 19:43:56 +11:00
c997cdee9f
COSMIT pep8 + full stop police
...
metrics/tests/test_score_objects.py made stricter by replacing . with \.
in regular expressions.
2014-02-01 15:28:53 +01:00
0413b55063
saving the random state
2013-12-14 12:04:48 -05:00
8b6b770f91
DOC fix BernoulliRBM._fit docstring
...
Return value is gone now. Thanks @jnothman.
2013-12-07 23:57:06 +01:00
470643502b
ENH speed up progress reporting in RBM
...
Previously took up to 45% (!) of the time in training, not the 10% promised
in the docstring, for the digits example. Now down to at most 3%.
Pseudo-likelihood is now reported for the entire X, rather than batch-by-batch,
which will take more memory.
2013-12-07 00:02:32 +01:00
b0485c47e2
BUG don't densify sparse matrix in BernoulliRBM.score_samples
...
Also documented the non-determinism of this method.
2013-12-06 23:38:21 +01:00
bced6739f0
changing test
2013-11-13 20:59:30 -05:00
3a2ef7902b
changing partial fit to be seperate from _fit
2013-11-03 16:47:16 -05:00
8fc06223d1
adding test to rbm
2013-11-03 14:35:47 -05:00
13a83155eb
FIX sparse matrix indexing in BernoulliRBM
...
Fixes #2545 , reported by @jfelectron.
2013-10-27 17:51:33 +01:00
89225014e4
DOC sparse matrix support in BernoulliRBM
2013-10-27 17:50:36 +01:00
605d34f5bb
Check that verbose output is sound.
2013-09-24 13:27:09 +02:00
f7026b04f5
Expanded docstring of the verbose parameter.
2013-09-23 15:33:53 +02:00
0ac17a5f2f
Cosmetic changes.
2013-09-19 12:15:42 +02:00
eaa8a5dfd5
Add issparse import.
2013-09-19 08:44:42 +02:00
4a2acd3f17
Cosmetic changes.
2013-09-19 08:42:02 +02:00
b3d9a30b0a
Change RBM sparse format from CSC to CSR.
2013-09-18 17:02:47 +02:00
24216b54be
Issue #2455 : Make RBM work with sparse input when verbose=True.
2013-09-18 16:38:23 +02:00
b154499cc6
changing private _fit to public partial_fit
2013-09-08 09:56:06 -04:00
32eb8b57f8
ENH rename eval / pseudolikelihood to score_samples
2013-07-26 23:32:23 +02:00
3abd51bcfb
:
2013-07-26 13:23:05 +02:00
Raghav R V