a87011262a
FIX precision to float64 across the codebase
2015-10-16 11:20:29 +02:00
cac1d281a4
FIX AdaBoostRegressort test failure with numpy 1.10
2015-10-15 18:21:52 +02:00
4f4c3598e6
FIX dtypes to conform to the stricter type cast rules of numpy 1.10
...
FIX set copy to (copy & whiten).
FIX/DOC Use float outputs for doctest
2015-10-15 18:21:51 +02:00
bb2acedac9
AdaBoostRegressor should not raise errors if the base_estimator
...
does not support sample_weights
2015-09-06 23:30:37 -04:00
0156dcf114
BUG Use epsilon threshold in `_samme_proba` and `_boost_real`
...
Instead of thresholding <0 probabilities to 1e-5, threshold <epsilon to epsilon. This avoids the issue of, e.g., probability values of 0 becoming larger than values of 1e-7.
Add a unit test for `_samme_proba` which checks that probability ordering is unchanged.
Resolves issue #4944 .
2015-07-12 13:34:40 -05:00
0650d5502e
DOC adding backlinks to docstrings
2015-06-03 00:24:04 -04:00
abd31d2cad
MAINT Make uniform the error raised for not fitted condition
2015-01-12 00:33:40 +05:30
8621f00f6f
Merge pull request #3173 from arjoly/sparse-tree
...
[MRG] Sparse input support for decision tree and forest
2014-11-21 10:10:33 +01:00
248a5037f6
Fixed issue 3815. Discrete AdaBoostClassifier now fails early if the base classifier if worse than random.
2014-11-07 13:42:19 +00:00
c6af5c6e94
ENH adaboost should accept c and fortran array
2014-11-06 13:58:31 +01:00
c03c01a504
ENH Bring sparse input support to tree-based methods
...
Author: Arnaud Joly <arnaud.v.joly@gmail.com>
Fares Hedayati <fares.hedayati@gmail.com>
2014-11-06 13:58:28 +01:00
4860126bfa
+ Fix a typo.
2014-10-02 15:09:30 -04:00
367845cacd
Shifted the _check_sample_weight into BaseWeightBoosting class
2014-09-28 20:03:58 +05:30
4d94392b49
Added _check_sample_weight
2014-09-28 19:23:36 +05:30
16bc265ed8
dded tests to test_boosting
2014-09-28 14:58:55 +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
1ddc225080
Trivial doc alteration, travis timeout
2014-06-02 21:01:54 -04:00
0234a57100
Fixed typo toCSR -> tocsr
2014-06-02 20:44:53 -04:00
bb752bd70e
Remove exectuable permision bit
2014-06-02 20:30:38 -04:00
e7eb17c3d4
Assert sample weight negative sum value error, add authors entry
2014-06-02 20:13:48 -04:00
7d664ca7c0
Updated documenation for weighted_boosting fit() to indicate behavior of forced dtype
2014-06-02 20:13:48 -04:00
df31dbe808
Resolved python3 forest import error in weighted_boosting.py
2014-06-02 20:13:48 -04:00
d55959a3eb
Inserted base classifier check for BaseDesicionTree and and BaseForest before enforcing DTYPE in wieghted boosting fit
2014-06-02 20:13:48 -04:00
372c0edb91
Re inserted the check for X.ndim == 2 in initialization, updated a doc detail for X: coo is converted to csr only in the initialization not in the predict functions
2014-06-02 20:13:48 -04:00
fd1e316a19
Enforced dtype for X, and clarified documentation by indcating the sparse formats supported and how they are treated
2014-06-02 20:13:48 -04:00
bffadaa3f3
Updated documention to reflect sparse support on input data
2014-06-02 20:13:48 -04:00
63934fbd86
Revised recuring import statement to respect conventions
2014-06-02 20:13:48 -04:00
bcdc26d7fd
Removed densification calls on X '.asarry()' from all predict methods, removed 'dense' parameter from check arrays call, removed contiguous cast of the input array X, detected X as list or array to find the number of examples accordinly using len() or shape[0]
2014-06-02 20:13:48 -04:00
2cdf2835cb
weight_boosting: unneeded np.copy
2014-04-21 10:52:31 +02:00
e9d660d80c
add sample_weight to base score and weight_boosting staged_score
2014-04-21 10:52:31 +02:00
cc7f7d603d
DOC trees/ensembles: class labels need not be integers
2014-03-21 10:54:20 +01:00
2988a23089
AdaBoostRegressor: fix redundant recalculation of error_vect.max()
2013-10-08 20:15:41 -07:00
94bd0e9a31
BUG: do not force base estimators to inherit from sklearn in AdaBoost
2013-09-11 13:51:12 +02:00
74d395215b
ENH bagging meta-estimator
2013-09-11 13:51:09 +02:00
0292f14699
Try to raise and test warnings.
2013-07-28 12:10:45 +02:00
3009a85a64
more sphinx fixes
2013-07-27 19:51:12 +02:00
d8f90d137d
use column_or_1d, move it to utils
2013-07-26 17:16:27 +02:00
1c133e8eb5
ENH add "make_y_1d" to utils, use it in estimators where needed.
2013-07-26 17:16:26 +02:00
e488704849
COSMIT pep8
2013-07-23 15:17:32 +02:00
ed9d50317a
FIX: broken test
2013-07-22 12:20:02 +02:00
f03095a978
WIP: use np.ascontiguousarray instead of array2d
2013-07-22 12:20:02 +02:00
e0dcedd4b2
WIP: avoid data conversion in AdaBoost
2013-07-22 12:20:02 +02:00
10a4155c30
WIP: address @ogrisel comments (2)
2013-07-22 12:20:00 +02:00
9b5cafe696
WIP: catch ValueError only
2013-07-22 12:20:00 +02:00
563f907fae
Some PEP8 / Flake8
2013-07-22 12:19:58 +02:00
c806cdf7a1
FIX: set random_state in adaboost
2013-07-22 12:19:58 +02:00
5e5dd90425
WIP: fix adaboost
2013-07-22 12:19:57 +02:00
8861833ec9
DOC spellfixes
2013-06-29 16:18:23 +03:00
Raghav R V