TomDLT
ceeef70037
ENH refactor NMF and add CD solver
2015-09-21 11:42:07 +02:00
TomDLT
94eb61960a
ENH add sag solver in LogisticRegression and Ridge
2015-09-10 13:28:02 -04:00
Lars Buitinck
e4df618956
ENH micro-optimize LatentDirichletAllocation + cosmetics
...
Main changes:
* Instead of checking the dimensionality of _log_dirichlet_expectation
in a loop, just call the appropriate 1-d or 2-d version directly.
* Moved exp() call to Cython level.
* Reshuffled _update_doc_distribution inner loop to do fewer
computations when stopping due to early convergence.
2015-08-12 09:28:16 +02:00
Andreas Mueller
6beacc3fb6
remove deprecated stuff from 0.17
2015-03-18 14:49:04 -04:00
Lars Buitinck
27d4376dcf
ENH optimize DBSCAN by rewriting in Cython
...
~30% off the running time for a 3.7e5 set of 3-d points.
2015-03-17 16:21:10 +01:00
Lars Buitinck
e90e4c91d3
MAINT ignore sparsefuncs_fast in diffs
2014-07-13 18:35:21 +02:00
Lars Buitinck
cdd2279c09
ENH fix astype usage to prevent copying
...
Removed where possible; using copy=False (from utils.fixes) where needed.
Also some C integer type fixes to gradient boosting.
2014-06-18 11:01:28 +02:00
Lars Buitinck
70011287cc
MAINT remove last trail of the ArrayBuilder
2014-01-11 16:02:41 +01:00
Lars Buitinck
f309023d7f
MAINT ignore sklearn/tree/_utils.c in diff
2014-01-11 15:44:43 +01:00
Lars Buitinck
24cb1f0102
PY3 fix RBM test
2013-07-24 18:07:57 +02:00
Lars Buitinck
e3d2ecf86b
MAINT _logistic_sigmoid.c is "binary"
2013-07-24 17:36:06 +02:00
Lars Buitinck
32325ebbbf
DOC IsotonicRegression wasn't in the changelog at all
2013-03-11 12:04:24 +01:00
Lars Buitinck
a2e37ebba3
COSMIT add lgamma to gitattributes
2013-01-31 11:29:13 +01:00
Lars Buitinck
305c304d02
COSMIT wrong path in .gitattributes
2013-01-10 18:11:36 +01:00
Lars Buitinck
68c7b6a13b
COSMIT no diff for pairwise_fast.c
2012-11-27 01:46:07 +01:00
Lars Buitinck
2b8ab3561a
ENH feature hashing transformer
2012-11-17 17:29:20 +01:00
Kenneth C. Arnold
13fd25e0a0
Mark Cython outputs as binary so their changes don't clutter diffs.
2012-10-06 17:02:36 -04:00