30cba7db11
Add a compatibility layer for some modules.
...
OK, there are a bunch of ``import *``, but I don't see any other
reasonable way ...
2011-09-02 12:07:18 +02:00
ddf4b72109
Move project directory from scikits.learn to sklearn
2011-09-02 12:06:57 +02:00
68f27e3790
Revert "Move project directory from scikits.learn to sklearn"
...
This reverts commit fd0d3b879d
2011-09-02 12:03:18 +02:00
9fe39085dc
Revert "Add a compatibility layer for some modules."
...
This reverts commit 3be3de1fd2
2011-09-02 12:03:00 +02:00
3be3de1fd2
Add a compatibility layer for some modules.
...
OK, there are a bunch of ``import *``, but I don't see any other
reasonable way ...
2011-09-02 11:38:42 +02:00
fd0d3b879d
Move project directory from scikits.learn to sklearn
2011-09-02 11:38:24 +02:00
c6300d88aa
Merge remote-tracking branch 'JeanKossaifi/sorted_repr' into JeanKossaifi-sorted_repr
2011-08-24 15:12:08 +02:00
929c9ff6ad
FIX : docstest fail
2011-08-24 13:08:50 +02:00
c1227f0d96
Sorting parameters in BaseEstimtor.__repr__
2011-08-24 12:56:53 +02:00
ea36a26d67
FIX : ELLIPSIS in least_angle.py doctests
2011-08-24 11:05:11 +02:00
0e540d9894
ENH : more standard import of scipy.sparse
2011-08-23 17:59:18 -04:00
f2b0c614ac
API : adding proper normalize options in Lasso and ElasticNet with clean up
...
ENH : fix doctests
2011-08-23 17:52:23 -04:00
864c48c351
Merge branch 'master' of github.com:scikit-learn/scikit-learn into normalize_data
...
Conflicts:
scikits/learn/linear_model/coordinate_descent.py
scikits/learn/linear_model/least_angle.py
scikits/learn/linear_model/omp.py
scikits/learn/linear_model/sparse/coordinate_descent.py
2011-08-23 17:45:56 -04:00
2faa1099b8
FIX : Ellipsis in least_angle.py doctests
2011-08-23 19:03:12 +02:00
5dcc25a082
Merge branch 'normalize_data'
...
Conflicts:
scikits/learn/linear_model/bayes.py
scikits/learn/linear_model/least_angle.py
2011-08-23 16:00:33 +02:00
ce9814bbe8
Remove fit params from all objects.
2011-08-23 15:37:40 +02:00
e73dce95ed
Test : lasso_lars_vs_lasso_*
...
Added a same test as the existing one, with normalization
2011-08-23 12:31:25 +02:00
6c6289f2c6
API: eps -> tol in bayes
...
This is not a numerical error, but a tolerance in the optimization loop
2011-08-21 18:05:08 +02:00
0dd22e5f96
Fix : deleted a forgotten line
...
I had forgotten a none wanted line in the code, I removed it.
2011-08-12 15:04:41 +02:00
fd176ae177
Added a function as_float_array in scikits.learn.utils
...
This function converts a numpy array to dtype np.float (64 or 32, depending of the original type).
It also takes an optional argument overwrite_X.
Thus, _center_data no more takes overwrite_X as an argument, and now modify X inplace.
2011-08-12 14:56:24 +02:00
1f2ad60264
Merge remote branch 'upstream/master' into normalize_data
...
Conflicts:
scikits/learn/linear_model/omp.py
2011-08-10 17:12:17 +02:00
dd9bd4c44f
Fixed some errors (mainly docstrings)
2011-08-10 16:31:55 +02:00
f599fab517
ENH + DOC: addressing @GaelVaroquaux's comments
2011-08-10 16:12:41 +02:00
e4f4b907e7
ENH: addressing @ogrisel's comments on PR 298
2011-08-10 16:12:41 +02:00
f4ddd47cf9
MISC : adding GaelVaroquaux to the authors of least_angle.py
2011-08-10 16:12:41 +02:00
899dfcda41
ENH : new LassoLarsIC estimator
2011-08-10 16:12:41 +02:00
e274935ba7
Added parameter normalize
...
Added parameter overwrite_X
2011-08-10 14:53:43 +02:00
6c479c0a3a
BUG: normalize before center in lars_path_residue
...
This was not actually triggering a bug, as we are passing in only
centered data, but it was wrong
2011-08-10 14:40:36 +02:00
295dc3c982
Added parameter normalize to omp.py
...
Some changes in Linear_Model
2011-08-10 10:32:48 +02:00
a23365c56b
added parameter normalize to ridge.py
2011-08-09 16:31:18 +02:00
a3f7cc4cec
added parameter normalize to coordinate_descent
2011-08-09 16:22:50 +02:00
f1a7454767
Implementation of the parameter normalize in bayes.py
2011-08-09 16:10:30 +02:00
52a0158345
Completed the integration of the parameter normalize
2011-08-09 16:09:55 +02:00
4c43f8e643
LassoLARS now uses the normalize parameter
2011-08-09 13:35:53 +02:00
a6ee968848
Added parameter normalize to LinearRegression
2011-08-09 12:31:45 +02:00
bb5fe7e51a
Added normalize parameter to LinearModel
...
In fit_intercept, there is now a normalize parameter (default=False)
to normalize data.
2011-08-09 10:51:07 +02:00
93d1bbd231
FIX: broken Gram handling in OMP estimator + minor style improvements
2011-08-08 15:12:32 +02:00
4ef30ea94a
FIX: cruel bug in OMP, no more unneeded warnings now.
2011-08-07 20:39:56 +03:00
a1d8259f4a
Merge branch 'master' into omp
...
Conflicts:
doc/modules/linear_model.rst
2011-08-05 17:50:34 +03:00
2cd0ccf71e
updated samples generator according to @glouppe's refactoring
2011-08-04 16:50:47 +03:00
c496a6eed2
API: _check_cv -> check_cv
2011-08-03 20:23:18 +02:00
c8b83e77f9
MISC: mse_path in LassoLarsCV is now the mean
...
To be consistent with LassoCV
2011-08-03 20:03:22 +02:00
124c1cc44c
ENH: use _check_cv in LassoLarsCV
2011-08-03 20:03:22 +02:00
d914caccf7
MISC: LarsCV: preallocate mse_path
2011-08-03 20:03:22 +02:00
6c6e2f9a9e
pep8
2011-08-03 20:03:22 +02:00
6304b4352f
DOC: document eps in least_angle better
2011-08-03 20:03:22 +02:00
8584f98f83
TEST: Improve test coverage of LarsCV
2011-08-03 20:03:22 +02:00
8edb57488d
DOC+TEST: LarsCV
2011-08-03 20:03:22 +02:00
871491d445
ENH: Expose eps in LARS
2011-08-03 20:03:22 +02:00
3b3c5436c4
BUG: Avoid div by 0 in lars_path_residues
2011-08-03 20:03:22 +02:00
Fabian Pedregosa