Skip to content

MRG Logistic regression preconditioning #15583

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 43 commits into from
Closed

MRG Logistic regression preconditioning #15583

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
43 commits
Select commit Hold shift + click to select a range
59f9e40
remove mean for logisticregression lbfgs
amueller Nov 10, 2019
697112a
add test that preconditioning works for offsets in X
amueller Nov 10, 2019
5a1431c
add precondition option temporarily to log_reg_scoring_path
amueller Nov 10, 2019
bf2e452
fix gradients, add test
amueller Nov 10, 2019
86f7520
remove unused grad_scale
amueller Nov 10, 2019
01c2c98
pep8
amueller Nov 10, 2019
2bfeba4
fix intercept for multinomial loss
amueller Nov 10, 2019
e72d271
fix loss for multinomial
amueller Nov 10, 2019
7d71afb
add multinomial logistic regression preconditioning with lbfgs
amueller Nov 10, 2019
f153321
pep8
amueller Nov 10, 2019
6da875d
hack around with sparse stuff, set precondition=True everywhere for c…
amueller Nov 10, 2019
3859d57
fixing warmstarting
amueller Nov 10, 2019
5dd503d
starting on sparse offset support
amueller Nov 11, 2019
090e540
Update sklearn/linear_model/_logistic.py
amueller Nov 11, 2019
2e83f7e
don't change sparse matrix inplace
amueller Nov 11, 2019
9656c2f
fix sparse offset
amueller Nov 11, 2019
fcfc54e
fix gradient error
amueller Nov 11, 2019
626ad44
stricter gradient test, add gradient test for multinomial loss
amueller Nov 12, 2019
8cc1633
use sample weights in averages
amueller Nov 12, 2019
a312bc9
fix doctest
amueller Nov 12, 2019
d8ec9d1
increase tolerance in coefficient equality test :-/
amueller Nov 12, 2019
954da5a
store loss value in logistic regression when using l-bfgs
amueller Nov 12, 2019
9c704aa
fix gradient, add gradient test
amueller Nov 12, 2019
1d61a94
offset gradient test for logistic loss
amueller Nov 12, 2019
a82edc1
cast matrix to csr for pointwise multiplication?!
amueller Nov 12, 2019
a2947ab
add docstrings and some explanation
amueller Nov 12, 2019
9ad4f93
add helper function for weighted mean and std
amueller Nov 12, 2019
150b5a3
typos
amueller Nov 12, 2019
68dfe1b
shorter docstring to placate pep8
amueller Nov 12, 2019
763b010
more pep8
amueller Nov 12, 2019
71e563a
rename loss_value_ to objective_value_
amueller Nov 13, 2019
ca44e1c
Update sklearn/linear_model/_logistic.py
amueller Nov 13, 2019
962a190
Update sklearn/linear_model/_logistic.py
amueller Nov 13, 2019
af16e99
address some of oliviers comments
amueller Nov 13, 2019
9ee255d
Merge branch 'logistic_precondition' of github.com:amueller/scikit-le…
amueller Nov 13, 2019
74ed6e9
pep8
amueller Nov 13, 2019
c7c1689
improve invariance test for logistic_loss
amueller Nov 13, 2019
1a1bcfd
fix objective value assignment
amueller Nov 13, 2019
82b0949
why do we suddenly need more dots? hum
amueller Nov 13, 2019
01a5aa2
add auto option, docstring
amueller Nov 13, 2019
77e5c99
typo
amueller Nov 13, 2019
d1109be
fix default value to 'auto'
amueller Nov 13, 2019
048e355
merge branch 'master' into logistic_precondition
amueller Oct 27, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion sklearn/feature_selection/_from_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,7 +144,7 @@ class SelectFromModel(MetaEstimatorMixin, SelectorMixin, BaseEstimator):
>>> y = [0, 1, 0, 1]
>>> selector = SelectFromModel(estimator=LogisticRegression()).fit(X, y)
>>> selector.estimator_.coef_
array([[-0.3252302 , 0.83462377, 0.49750423]])
array([[-0.32... , 0.83..., 0.49...]])
>>> selector.threshold_
0.55245...
>>> selector.get_support()
Expand Down
Loading