You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+1-1
Original file line number
Diff line number
Diff line change
@@ -4,7 +4,7 @@
4
4
5
5
> A unified interface for machine learning with time series
6
6
7
-
:rocket:**Version 0.15.1 out now!**[Check out the release notes here](https://www.sktime.org/en/latest/changelog.html).
7
+
:rocket:**Version 0.16.0 out now!**[Check out the release notes here](https://www.sktime.org/en/latest/changelog.html).
8
8
9
9
sktime is a library for time series analysis in Python. It provides a unified interface for multiple time series learning tasks. Currently, this includes time series classification, regression, clustering, annotation and forecasting. It comes with [time series algorithms](https://www.sktime.org/en/stable/estimator_overview.html) and [scikit-learn] compatible tools to build, tune and validate time series models.
Copy file name to clipboardExpand all lines: docs/source/changelog.rst
+180-1
Original file line number
Diff line number
Diff line change
@@ -13,12 +13,191 @@ All notable changes to this project will be documented in this file. We keep tra
13
13
For upcoming changes and next releases, see our `milestones <https://github.com/sktime/sktime/milestones?direction=asc&sort=due_date&state=open>`_.
14
14
For our long-term plan, see our :ref:`roadmap`.
15
15
16
-
Version 0.15.1 - 2023-01-12
16
+
Version 0.16.0 - 2023-01-30
17
17
---------------------------
18
18
19
19
Highlights
20
+
~~~~~~~~~~
21
+
22
+
* ``HierarchyEnsembleForecaster`` for level- or node-wise application of forecasters on panel/hierarchical data (:pr:`3905`) :user:`VyomkeshVyas`
23
+
* new transformer: ``BKFilter``, Baxter-King filter, interfaced from ``statsmodels`` (:pr:`4127`) :user:`klam-data`, :user:`pyyim``
24
+
* ``get_fitted_params`` of pipelines and other heterogenous meta-estimators now supports parameter nesting (:pr:`4110`) :user:`fkiraly`
25
+
26
+
Dependency changes
27
+
~~~~~~~~~~~~~~~~~~
28
+
29
+
* ``statsmodels`` is now a soft dependency. Estimators dependent on ``statsmodels``
30
+
can be used exactly as before if ``statsmodels`` is present in the python environment.
31
+
32
+
Core interface changes
33
+
~~~~~~~~~~~~~~~~~~~~~~
34
+
35
+
BaseEstimator
36
+
^^^^^^^^^^^^^
37
+
38
+
* The method ``get_fitted_params``, of all ``BaseEstimator`` descendants
39
+
(any estimator with ``fit``), has a new boolean argument ``deep``, default ``True``.
40
+
Similar to the argument of the same name of ``get_params``, this allows to control
41
+
for composite estimators, whether to return fitted parameters with or
42
+
without estimator nesting.
43
+
44
+
Forecasting
45
+
^^^^^^^^^^^
46
+
47
+
* all forecasters: the public ``cutoff`` attribute of forecasters has changed
48
+
to ``pd.Index`` subtype, from index element. To update previously
49
+
functional code, replace references to ``cutoff`` by ``cutoff[0]``.
50
+
51
+
52
+
Deprecations and removals
53
+
~~~~~~~~~~~~~~~~~~~~~~~~~
54
+
55
+
Dependencies
56
+
^^^^^^^^^^^^
57
+
58
+
* ``statsmodels`` has changed from core dependency to soft dependency in ``sktime 0.16.0``.
59
+
To ensure functioning of setups of ``sktime`` code dependent on ``statsmodels`` based estimators
60
+
going forward, ensure to install ``statsmodels`` in the environment explicitly,
61
+
or install the ``all_extras`` soft dependency set which will continue to contain ``statsmodels``.
62
+
63
+
Data types, checks, conversions
64
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
65
+
66
+
* ``check_is_scitype``: the ``msg_legacy_interface`` argument has now been removed.
67
+
Future behaviour is as per the default of the argument, ``msg_legacy_interface=False``.
68
+
69
+
Forecasting
70
+
^^^^^^^^^^^
71
+
72
+
* all forecasters: the public ``cutoff`` attribute of forecasters has changed
73
+
to ``pd.Index`` subtype, from index element. To update previously
74
+
functional code, replace references to ``cutoff`` by ``cutoff[0]``.
75
+
76
+
Transformations
77
+
^^^^^^^^^^^^^^^
78
+
79
+
* ``Catch22``: the ``transform_single_feature`` method has been removed from the ``Catch22``
80
+
transformer
81
+
* ``FourierFeatures``: in 0.17.0, the default value of the ``keep_original_columns``
82
+
parameter will change to ``False``
83
+
84
+
Enhancements
20
85
~~~~~~~~~~~~
21
86
87
+
BaseEstimator
88
+
^^^^^^^^^^^^^
89
+
90
+
* [ENH] ``get_fitted_params`` for pipelines and other heterogenous meta-estimators (:pr:`4110`) :user:`fkiraly`
91
+
* [ENH] ``deep`` argument for ``get_fitted_params`` (:pr:`4113`) :user:`fkiraly`
92
+
93
+
Data types, checks, conversions
94
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
95
+
96
+
* [ENH] significantly speed up ``nested_univ`` (nested dataframe) check for non-nested data (:pr:`4130`) :user:`danbartl`
97
+
* [ENH] refactor - localize broadcasting in ``VectorizedDF`` (:pr:`4132`) :user:`fkiraly`
98
+
* [ENH] ``get_time_index`` rework to get faster run times for grouped data (:pr:`4141`) :user:`danbartl`
99
+
100
+
Forecasting
101
+
^^^^^^^^^^^
102
+
103
+
* [ENH] ``HierarchyEnsembleForecaster`` for level- or node-wise application of forecasters on panel/hierarchical data (:pr:`3905`) :user:`VyomkeshVyas`
104
+
* [ENH] second set of test parameters for ``ARIMA`` (:pr:`4099`) :user:`fkiraly`
0 commit comments