8000 MNT fix changelog 1.0.1 (#21416) · scikit-learn/scikit-learn@5d75547 · GitHub
[go: up one dir, main page]
Skip to content

Commit 5d75547

Browse files
glemaitreglemaitre
committed
MNT fix changelog 1.0.1 (#21416)
1 parent 316c364 commit 5d75547

File tree

2 files changed

+69
-1
lines changed

2 files changed

+69
-1
lines changed

‎doc/whats_new/v1.0.rst

Lines changed: 30 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@
77
Version 1.0.1
88
=============
99

10-
**In Development**
10+
**October 2021**
1111

1212
Changelog
1313
---------
@@ -43,6 +43,14 @@ Fixed models
4343
between sparse and dense input. :pr:`21195`
4444
by :user:`Jérémie du Boisberranger <jeremiedbb>`.
4545

46+
:mod:`sklearn.ensemble`
47+
.......................
48+
49+
- |Fix| Fixed a bug that could produce a segfault in rare cases for
50+
:class:`ensemble.HistGradientBoostingClassifier` and
51+
:class:`ensemble.HistGradientBoostingRegressor`.
52+
:pr:`21130` :user:`Christian Lorentzen <lorentzenchr>`.
53+
4654
:mod:`sklearn.gaussian_process`
4755
...............................
4856

@@ -59,12 +67,21 @@ Fixed models
5967
longer checks for uppercase characters in the provided vocabulary. :pr:`21251`
6068
by :user:`Jérémie du Boisberranger <jeremiedbb>`.
6169

70+
- |Fix| Fixed a bug in :class:`feature_extraction.CountVectorizer` and
71+
:class:`feature_extraction.TfidfVectorizer` by raising an
72+
error when 'min_idf' or 'max_idf' are floating-point numbers greater than 1.
73+
:pr:`20752` by :user:`Alek Lefebvre <AlekLefebvre>`.
74+
6275
:mod:`sklearn.linear_model`
6376
...........................
6477

6578
- |Fix| Improves stability of :class:`linear_model.LassoLars` for different
6679
versions of openblas. :pr:`21340` by `Thomas Fan`_.
6780

81+
- |Fix| :class:`linear_model.LogisticRegression` now raises a better error
82+
message when the solver does not support sparse matrices with int64 indices.
83+
:pr:`21093` by `Tom Dupre la Tour`_.
84+
6885
:mod:`sklearn.neighbors`
6986
........................
7087

@@ -95,6 +112,18 @@ Fixed models
95112
:mod:`sklearn.utils`
96113
....................
97114

115+
- |Enhancement| :func:`utils.validation._check_sample_weight` can perform a
116+
non-negativity check on the sample weights. It can be turned on
117+
using the only_non_negative bool parameter.
118+
Estimators that check for non-negative weights are updated:
119+
:func:`linear_model.LinearRegression` (here the previous
120+
error message was misleading),
121+
:func:`ensemble.AdaBoostClassifier`,
122+
:func:`ensemble.AdaBoostRegressor`,
123+
:func:`neighbors.KernelDensity`.
124+
:pr:`20880` by :user:`Guillaume Lemaitre <glemaitre>`
125+
and :user:`András Simon <simonandras>`.
126+
98127
- |Fix| Solve a bug in :func:`~sklearn.utils.metaestimators.if_delegate_has_method`
99128
where the underlying check for an attribute did not work with NumPy arrays.
100129
:pr:`21145` by :user:`Zahlii <Zahlii>`.

‎doc/whats_new/v1.1.rst

Lines changed: 39 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -67,6 +67,42 @@ Changelog
6767
- |Fix| :class:`linear_model.LogisticRegression` now raises a better error
6868
message when the solver does not support sparse matrices with int64 indices.
6969
:pr:`21093` by `Tom Dupre la Tour`_.
70+
:mod:`sklearn.cross_decomposition`
71+
..................................
72+
73+
- |Enhancement| :func:`cross_decomposition._PLS.inverse_transform` now allows
74+
reconstruction of a `X` target when a `Y` parameter is given. :pr:`19680` by
75+
:user:`Robin Thibaut <robinthibaut>`.
76+
77+
:mod:`sklearn.impute`
78+
.....................
79+
80+
- |API| Adds :meth:`get_feature_names_out` to :class:`impute.SimpleImputer`,
81+
:class:`impute.KNNImputer`, :class:`impute.IterativeImputer`, and
82+
:class:`impute.MissingIndicator`. :pr:`21078` by `Thomas Fan`_.
83+
84+
:mod:`sklearn.metrics`
85+
......................
86+
87+
- |API| :class:`metrics.DistanceMetric` has been moved from
88+
:mod:`sklearn.neighbors` to :mod:`sklearn.metric`.
89+
Using `neighbors.DistanceMetric` for imports is still valid for
90+
backward compatibility, but this alias will be removed in 1.3.
91+
:pr:`21177` by :user:`Julien Jerphanion <jjerphan>`.
92+
93+
:mod:`sklearn.model_selection`
94+
..............................
95+
96+
- |Enhancement| raise an error during cross-validation when the fits for all the
97+
splits failed. Similarly raise an error during grid-search when the fits for
98+
all the models and all the splits failed. :pr:`21026` by :user:`Loïc Estève <lesteve>`.
99+
100+
:mod:`sklearn.pipeline`
101+
.......................
102+
103+
- |Enhancement| Added support for "passthrough" in :class:`FeatureUnion`.
104+
Setting a transformer to "passthrough" will pass the features unchanged.
105+
:pr:`20860` by :user:`Shubhraneel Pal <shubhraneel>`.
70106

71107
:mod:`sklearn.utils`
72108
....................
@@ -89,6 +125,9 @@ Changelog
89125
- |Enhancement| Added support for "passthrough" in :class:`FeatureUnion`.
90126
Setting a transformer to "passthrough" will pass the features unchanged.
91127
:pr:`20860` by :user:`Shubhraneel Pal <shubhraneel>`.
128+
- |Enhancement| :func:`utils.estimator_html_repr` shows a more helpful error
129+
message when running in a jupyter notebook that is not trusted. :pr:`21316`
130+
by `Thomas Fan`_.
92131

93132
Code and Documentation Contributors
94133
-----------------------------------

0 commit comments

Comments
 (0)
0