2bd87f6ed8
Remove python < 3.5 from CI ( #12746 )
2018-12-14 10:53:12 +01:00
fa98a72dcc
MNT Replaced all occurrences of assert_true and assert_false with assert ( #12588 )
2018-11-28 09:16:26 +08:00
362cb3bcab
TST autoreplace assert_true(...==...) with plain assert ( #12547 )
2018-11-11 09:05:34 +08:00
eec7649236
MAINT Complete 0.20 deprecations ( #9570 )
2018-06-24 23:06:26 +10:00
cb1dbc25a7
MNT: reduce the number of warnings in test_common.py ( #11151 )
...
Improve handling of warning filters in sklearn.utils.testing by avoiding to reset warnings.filters.
2018-06-11 08:59:12 +02:00
f049ec72eb
TST Pytest parametrization part3 - feature_extraction, gaussian_process modules ( #11143 )
2018-06-04 22:23:29 +08:00
d9a0774afc
MAINT Always import signature from sklearn.utils.fixes ( #11019 )
2018-04-24 09:27:45 +10:00
6e01feff14
OPTIM make GaussianProcessRegressor faster with return_std=True
2017-09-01 13:13:12 +02:00
846313b570
[MRG+1] Deprecate sklearn.utils.testing.raises and remove it from tests ( #9660 )
2017-09-01 12:53:59 +02:00
ec0dbf0399
[MRG+1] Fixes predicting std and cov without fitting in GPR by default. ( #9177 )
...
* If self.kernel_ is None then self.kernel_ is set to RBFKernel in predict() as fit(). Fixes #6573
* xxx_ attributes can not be altered outside fit. Removing self.kernel_ from predict() and using previously implemented self.kernel attribute.
* Cleanup and add non-regression-test
2017-06-25 15:32:16 -07:00
8dde4096bd
[MRG] ENH make rng seed thread safe everywhere it is possible ( #9184 )
...
* ENH make rng seed thread safe everywhere it is possible
* ENH remove unneeded use of np.random.seed
2017-06-23 07:40:33 +02:00
fa82eee82f
[MRG+1] Better error message for GPR ( #8386 )
...
* FIX: better error message for GPR
* TST: added tests for error message
* FIX: correct error message and simplified tests
* DOC: updated docstring for GPR
* FIX: pass float as value for alpha
* Raise original error after modification to preserve traceback
Amend test accordingly
2017-06-08 15:06:48 +10:00
39076ff385
fix deprecated comparison to string in GP ( #8518 )
2017-03-04 13:37:11 -08:00
0ec6664ff1
[MRG] Make tests runnable with pytest without error ( #8246 )
...
* Make tests runnable with pytest without error.
Errors were due to pytest quirks with (deprecated) yield support.
* Add pytest build on Travis
and tweak pytest settings in setup.cfg
* Tweak comment
2017-02-20 13:08:09 +01:00
9f6b849c35
TST Change rstrip() to truncation in test function ( #8237 )
2017-01-27 08:41:03 +11:00
4da44c8541
[MRG+1] replaced some assert_true(np.allclose(x, y)) with assert_almost_equal ( #7742 )
...
* replaced some assert_true(np.allclose(x, y)) with assert_almost_equal for better error messages.
also some pep8.
* typo fixes
2016-10-25 09:15:58 -04:00
9b2aac9e5c
[MRG + 1] [TST] (half-cosmetic) use less nose.tools import to simplify future transition to py.test ( #7384 )
...
* use less nose.tools import to simplify future transition to activly developing test suites/runners
* assert_equal -> assert_array_equal in test_feature_hasher_pairs_with_string_values
and one missed ImportError that should be replaced with AttributeError
* test for py2.6 compat with except AttributeError
* fix importing of SkipTest
* force using nose in python2.6 for now
* there was no assert_dict_equal in py2.6. but we can use assert_equal
although failed test will look a little bit ugly
* remove nose imports from doc/datasets
2016-10-07 12:46:52 -04:00
e17d5c9337
[MRG] Fix GaussianProcess batch predict in Py3k #7329 ( #7330 )
...
* fix batch
* add test for eval_MSE
2016-10-06 14:07:40 +11:00
19aebe2431
use == instead of 'is' for string comparison ( #7422 )
2016-09-15 09:42:34 +10:00
3d1a2fdb5c
COSMIT don't use docstrings in tests (gp kernel tests) ( #7399 )
2016-09-13 13:17:38 +02:00
49fb295561
[MRG+2] Fix repr on isotropic kernels when a 1-D length scale is given ( #7259 )
2016-09-03 22:02:07 +10:00
c31ad7ada3
[MRG + 1] [MAINT] Remove non-trivial logic during kernel initialization ( #6963 )
...
* Make sure set_params does not break clone on kernels
For example, in kernels such as RBF, using set_params with a
length_scale of different length, used to break clone because of
the non-trivial modification of length_scale in the constructor.
This PR resolves this by moving this logic to __call__ and
making hyperparameter attributes properties.
* Fix tests by using get_params and implement __eq__ for Hyperparameter
* Explain XXX
* Check that hyperparameters are equal after non-trivial set_params
* Refactor into _check_length-scale
2016-08-16 23:43:58 -07:00
d2aaf0f4bb
doc: fix spellings of 'license'
2016-03-31 17:25:31 -07:00
1e5820b0ad
Semantics of ``assert_almost_equal`` now consistent on floats and arrays. see https://github.com/numpy/numpy/issues/5200
2016-02-15 14:09:18 +01:00
58c859e7ba
Limit the number of thread and force install libgfortran
2016-02-06 18:56:03 +01:00
6470a83f99
MAINT speed up test_kernels
2016-02-06 18:55:23 +01:00
dd6ab4e896
MAINT speed up test_gpr
2016-02-06 18:55:23 +01:00
8d270fb40c
MAINT speed up test_gpc
2016-02-06 18:55:22 +01:00
8878088b19
FIX inspect.argspect is deprecated; use signature
2015-10-24 10:27:26 +02:00
cea3f29e0f
TST test_lml_precomputed() checks only for equality in first 7 digits
2015-10-19 18:07:15 +02:00
3957fc3b34
ENH GPR.log_marginal_likelihood() returns the current log-likelihood if no theta vector is provided
2015-10-19 18:07:14 +02:00
e2a340a4ee
REF More meaningful names for hyperparameters of GP kernels
2015-10-19 18:07:14 +02:00
2dd1f285b9
PEP8 Making pep8 and pyflakes happy
2015-10-19 18:07:14 +02:00
98e8093c93
FIX Further python3 related fixes
2015-10-19 18:07:14 +02:00
55d26aa9a6
FIX Fixing test_kernel_theta
2015-10-19 18:07:14 +02:00
f76e64c9c7
FIX Fix failing tests (doctest, order of hyperparameters, pairwise_kernels)
2015-10-19 18:07:14 +02:00
eb54b496d2
REF Removing theta_vars attribute from kernels and using Hyperparameter instances instead
2015-10-19 18:07:14 +02:00
3f1a0f1a5a
TST Adding two unittests for testing GP kernels
...
test_kernel_versus_pairwise() checks that GP kernels can also be used as pairwise kernels.
test_set_get_params() checks that set_params()/get_params() is consistent with kernel.theta.
2015-10-19 18:07:14 +02:00
65555b300b
REF Renaming sigma_squared_n to alpha (consistency with Ridge)
2015-10-19 18:07:13 +02:00
c6db24e56f
REF Several small changes based on @eickenberg's suggestions
2015-10-19 18:07:13 +02:00
a6d564b4bb
TST Testing n_jobs option of GaussianProcessClassifier
2015-10-19 18:07:13 +02:00
adc7451201
PEP8 Fixing PEP8 issues in gaussian_process package
2015-10-19 18:07:13 +02:00
768c81fd1b
TST Further GP-related fixes
2015-10-19 18:07:13 +02:00
1f7deb7f6a
TST Extending testing of multi-class GP
2015-10-19 18:07:12 +02:00
9929b37270
FIX Excluding WhiteKernel from one unit-test
2015-10-19 18:07:12 +02:00
0fa050c72c
TST Testing GPC for multi-class classification problems
2015-10-19 18:07:12 +02:00
588cabfc46
TST Added test checking that GPR can deal with different outputs for same input
2015-10-19 18:07:12 +02:00
daa90c22d3
ENH Adding Matern kernel for Gaussian Processes
2015-10-19 18:07:12 +02:00
2839f66933
REFACTOR normalize_y in GPR does not modify standard-deviation
2015-10-19 18:07:12 +02:00
78caa0b5d0
TST Testing GPC with custom external optimizer
2015-10-19 18:07:12 +02:00
Adrin Jalali