87aaabc06a
DOC backticks in attribute docstrings unnecessaru since #3489
2015-06-12 05:13:02 +10:00
0650d5502e
DOC adding backlinks to docstrings
2015-06-03 00:24:04 -04:00
011ad8d8a4
ENH add a prior strategy for the dummy estimator
2015-04-14 09:44:52 +02:00
6beacc3fb6
remove deprecated stuff from 0.17
2015-03-18 14:49:04 -04:00
d16e9eeb1b
FIX stricter validation for non-empty input data
2015-02-14 00:39:26 +01:00
88555987cc
PEP8 Fix E101, E111 errors and W191, W293, W293 and W391 warnings.
2015-01-08 04:28:24 +05:30
e0f4cd92f2
Use np.average instead of np.mean
2014-10-16 17:29:31 +02:00
da31344751
Add sample_weight support to Dummy Regressor
2014-10-16 16:41:49 +02:00
b698d9f55a
Revert unwanted modificatoin
2014-10-16 16:38:13 +02:00
6b0b773c72
COSMIT less nested constant check + use comprehension
2014-10-16 15:43:07 +02:00
2810f5a28d
COSMIT reshape for all regression strategy + avoid xrange
2014-10-16 15:38:44 +02:00
ea1548ad48
Add quantile strategy to DummyRegressor. Fixes #3421 .
2014-10-16 09:49:09 +02:00
39da06f567
Merge pull request #3438 from hamsal/sprs-out-dmy
...
[MRG+1] Sparse Target Dummy Classifier
2014-08-17 16:22:24 +02:00
2a3f504f1c
Make cosmetic adjustments, Fix random.py header, use message with assert_warns
2014-08-08 18:48:23 -04:00
7b7784e0aa
Make cosmetic changes
...
Break list comprehenesions differently in dummy precit
Drop uncesseray float() in multiclass class_distrribtuion
Combine two lines in random_choice_csc
2014-08-06 23:01:42 -04:00
222cb87af1
Remove outer for loop in dense case predict
2014-08-05 15:31:07 -04:00
f69dcefa9b
Use UserWarning in place of SparseEfficiencyWarning
2014-08-05 14:16:55 -04:00
214b761d2f
Clarify unfiorm sparse warning message, Reword sparse_output_ doc
2014-08-05 14:13:22 -04:00
22ef22e0fe
Clarify unfiorm sparse warning message, Reword sparse_output_ doc
2014-08-05 14:11:26 -04:00
5ba7df8b6c
Include dense case support in class_distribution, pep8 revisions
...
Make The function previoulsy sparse_class_distribution support both the sparse and
dense case, and make some cosmetic updates for imporved readability
2014-08-05 09:45:20 -04:00
7e694fa765
Uodate imports to correct function loactions
...
For the functions random_choice_csc and sparse_class_distribution
2014-08-04 18:29:22 -04:00
7bb1c01b6d
Reverse conditionals in fit and predict, positive case first
2014-08-04 10:55:06 -04:00
2ec8e978c2
Correct spell 'Predicting', doument (sparse_target_input_ -> sparse_output)
2014-08-04 10:47:44 -04:00
cfad359978
Rever formating of warning, Reword warning for sparse uniform case
2014-08-04 09:54:14 -04:00
44499fafad
Move array intiialiaztions into dense conditional, nz_indices -> col_nonzero
2014-08-01 16:19:27 -04:00
65f4f953db
Absract class prior construction to function sparse_class_distribution
2014-07-31 18:27:38 -04:00
91a5b92882
Support matrix output random_choice_csc, Update usage in dummy predict
...
Use this general function for each case int the sparse predict
of the dummy classfier
2014-07-31 14:00:52 -04:00
d0f52b6eb0
Raise warning unifrom fit sparse target, Raise error uniform precit sparse target
2014-07-31 14:00:52 -04:00
72885ef7f8
Use array for sparse matrix construction, fix constant check
...
-Replace the numpy array construction of data,indicies, and inptr
with regular arrays.
-loop over all output columns when checking for the prescence of
the user provided constant.
2014-07-31 14:00:52 -04:00
3d692a9661
Implement uniform and stratified w/ sparse random choices, optimize concats
...
-Write a sparse version of numpy.random.choice in utils/sparsefuncs
-Remove concats from inner loop of predict and use only one finaly concat call
2014-07-31 14:00:52 -04:00
05a8dd7cd0
Select nonzero elements in predict stratified case
2014-07-31 14:00:52 -04:00
5c5ef1ad7c
Update y with y.tocsc (Not an inplace op)
2014-07-31 14:00:52 -04:00
69adaa43f4
Combine dense column check in fit
2014-07-31 14:00:52 -04:00
8c3a986500
Remove redundant indices, data, indptr appends in predict
...
Take them out of the four branches and do them once at the end of
all of the conditionals
2014-07-31 14:00:52 -04:00
7d9114cbf7
Correct the class priors computation in the sparse target case
...
Consider the wieghts of the zero elements and hendle the insertion
of the count of zeros in the bincount manaully
2014-07-31 14:00:52 -04:00
a120827002
Emulate sample weight support from dense (Rebase)
2014-07-31 14:00:52 -04:00
388ac61484
remove zero from random uniform return data
2014-07-31 14:00:52 -04:00
c0981d24e6
Reorder fit and share denes sprs, Scafold sprs pred
...
Put the code for managing the constant case in fit
outside of the sparse conditional so it is shared
by both branches of execution.
Scaffold the four cases for sparse prediciton,
use a new class attribute sparse_target_input_
to carry over issparse from fit. Cases for
stratified and unifrom need improvement.
2014-07-31 14:00:52 -04:00
73f2f0ca4d
Share array intialization sparse and dense fit
...
The arrays n_outputs_, classes_, n_classes_, class_prior_
2014-07-31 14:00:51 -04:00
f7efdc9115
Op directly on y.data in fit, xrange -> range, y.tocsc
2014-07-31 14:00:51 -04:00
c04ba19055
Fit dummy classifier with sparse target data
2014-07-31 14:00:51 -04:00
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
9a898daaa6
DOC sample_weight attribute
2014-07-18 10:43:11 +02:00
f9773ba8e2
ENH add sample_weight support to dummy classifier
2014-07-18 09:53:13 +02:00
67b8601314
Minor fixes in the docs
2014-06-07 16:38:53 +02:00
eb10c4c29a
FIX: numpy mtrand does not accept Python long instances under Windows
2014-03-27 15:35:16 +01:00
302a3c0eb0
Merge pull request #2886 from maheshakya/dummy_regressor
...
[MRG+1] ENH Implemented median and constant strategies in DummyRegressor
2014-03-10 16:37:46 +01:00
3412f29c5a
Removed one of 'is a regressor' part in the DummyRegressor documentation
2014-03-10 18:36:04 +05:30
Joel Nothman