Merge pull request #915 from labdmitriy/group-time-series

Group time series split
rasbt · May 27, 2022 · 9e08841 · 9e08841
2 parents 2d4ca32 + fc17856
commit 9e08841
Show file tree

Hide file tree

Showing 7 changed files with 1,541 additions and 0 deletions.
diff --git a/docs/mkdocs.yml b/docs/mkdocs.yml
@@ -70,6 +70,7 @@ nav:
     - user_guide/evaluate/create_counterfactual.md
     - user_guide/evaluate/feature_importance_permutation.md
     - user_guide/evaluate/ftest.md
+    - user_guide/evaluate/GroupTimeSeriesSplit.md
     - user_guide/evaluate/lift_score.md
     - user_guide/evaluate/mcnemar_table.md
     - user_guide/evaluate/mcnemar_tables.md

diff --git a/docs/sources/CHANGELOG.md b/docs/sources/CHANGELOG.md
@@ -24,6 +24,7 @@ The CHANGELOG for the current development version is available at
 - The `mlxtend.evaluate.bootstrap_point632_score` now supports `fit_params`. ([#861](https://github.com/rasbt/mlxtend/pull/861))
 - The `mlxtend/plotting/decision_regions.py` function now has a `contourf_kwargs` for matplotlib to change the look of the decision boundaries if desired. ([#881](https://github.com/rasbt/mlxtend/pull/881) via [[pbloem](https://github.com/pbloem)])
 - Add a `norm_colormap` parameter to `mlxtend.plotting.plot_confusion_matrix`, to allow normalizing the colormap, e.g., using `matplotlib.colors.LogNorm()` ([#895](https://github.com/rasbt/mlxtend/pull/895))
+- Add new `GroupTimeSeriesSplit` class for evaluation in time series tasks with support of custom groups and additional parameters in comparison with scikit-learn's `TimeSeriesSplit`. ([#915](https://github.com/rasbt/mlxtend/pull/915) via [Dmitry Labazkin](https://github.com/labdmitriy))
 
 ##### Changes
 

diff --git a/docs/sources/USER_GUIDE_INDEX.md b/docs/sources/USER_GUIDE_INDEX.md
@@ -36,6 +36,7 @@
 - [create_counterfactual](user_guide/evaluate/create_counterfactual.md)
 - [feature_importance_permutation](user_guide/evaluate/feature_importance_permutation.md)
 - [ftest](user_guide/evaluate/ftest.md)
+- [GroupTimeSeriesSplit](user_guide/evaluate/GroupTimeSeriesSplit.md)
 - [lift_score](user_guide/evaluate/lift_score.md)
 - [mcnemar_table](user_guide/evaluate/mcnemar_table.md)
 - [mcnemar_tables](user_guide/evaluate/mcnemar_tables.md)

diff --git a/docs/sources/user_guide/evaluate/GroupTimeSeriesSplit.ipynb b/docs/sources/user_guide/evaluate/GroupTimeSeriesSplit.ipynb
diff --git a/mlxtend/evaluate/__init__.py b/mlxtend/evaluate/__init__.py
@@ -21,6 +21,7 @@
 from .permutation import permutation_test
 from .proportion_difference import proportion_difference
 from .scoring import scoring
+from .time_series import GroupTimeSeriesSplit
 from .ttest import paired_ttest_5x2cv, paired_ttest_kfold_cv, paired_ttest_resampled
 
 __all__ = [
@@ -47,4 +48,5 @@
     "bias_variance_decomp",
     "accuracy_score",
     "create_counterfactual",
+    "GroupTimeSeriesSplit",
 ]