Mlxtend

Changes

- Updated dependency on distutils for python 3.12 and above ([[1072](https://github.com/rasbt/mlxtend/issues/1072)](https://github.com/rasbt/mlxtend/issues/1072) via [[peanutsee](https://github.com/peanutsee)](https://github.com/peanutsee))

Downloads

- [[Source code (zip)](https://github.com/rasbt/mlxtend/archive/v0.21.1.zip)](https://github.com/rasbt/mlxtend/archive/v0.21.1.zip)

- [[Source code (tar.gz)](https://github.com/rasbt/mlxtend/archive/v0.22.1.tar.gz)](https://github.com/rasbt/mlxtend/archive/v0.22.1.tar.gz)

Changes

- Address NumPy deprecations to make mlxtend compatible to NumPy 1.24
- Changed the signature of the `LinearRegression` model of sklearn in the test removing the `normalize` parameter as it is deprecated. ([[1036](https://github.com/rasbt/mlxtend/issues/1036)](https://github.com/rasbt/mlxtend/issues/1036))
- Add `pyproject.toml` to support PEP 518 builds ([[1065](https://github.com/rasbt/mlxtend/issues/1065)](https://github.com/rasbt/mlxtend/issues/1065) via [[jmahlik](https://github.com/jmahlik)](https://github.com/jmahlik))
- Fixed installation from sdist failing ([[1065](https://github.com/rasbt/mlxtend/issues/1065)](https://github.com/rasbt/mlxtend/issues/1065) via [[jmahlik](https://github.com/jmahlik)](https://github.com/jmahlik))
- Converted configuration to `pyproject.toml` ([[1065](https://github.com/rasbt/mlxtend/issues/1065)](https://github.com/rasbt/mlxtend/issues/1065) via [[jmahlik](https://github.com/jmahlik)](https://github.com/jmahlik))
- Remove `mlxtend.image` submodule with face recognition functions due to poor `dlib` support in modern environments.

New Features and Enhancements

- Document how to use `SequentialFeatureSelector` and multiclass ROC AUC.

Changes

- When [`ExhaustiveFeatureSelector`](https://rasbt.github.io/mlxtend/user_guide/feature_selection/ExhaustiveFeatureSelector/) is run with `n_jobs == 1`, joblib is now disabled, which enables more immediate (live) feedback when the `verbose` mode is enabled. ([#985](https://github.com/rasbt/mlxtend/pull/985) via [Nima Sarajpoor](https://github.com/NimaSarajpoor))
- Disabled unnecessary warning in [`EnsembleVoteClassifier`](https://rasbt.github.io/mlxtend/user_guide/classifier/EnsembleVoteClassifier/) ([#941](https://github.com/rasbt/mlxtend/issues/941))
- Fixed various documentation issues ([849](https://github.com/rasbt/mlxtend/issues/849) and [#951](https://github.com/rasbt/mlxtend/issues/951) via [Lekshmanan Natarajan](https://github.com/zuari1993))
- Fixed "Edit on GitHub" button ([1024](https://github.com/rasbt/mlxtend/issues/1024))

New Features and Enhancements

- The [`mlxtend.frequent_patterns.association_rules`](https://rasbt.github.io/mlxtend/user_guide/frequent_patterns/association_rules/) function has a new metric - Zhang's Metric, which measures both association and dissociation. ([#980](https://github.com/rasbt/mlxtend/pull/980))
- Internal [`mlxtend.frequent_patterns.fpmax`](https://rasbt.github.io/mlxtend/user_guide/frequent_patterns/association_rules/) code improvement that avoids casting a sparse DataFrame into a dense NumPy array. ([#1000](https://github.com/rasbt/mlxtend/pull/1000) via [Tim Kellogg](https://github.com/tkellogg))
- The [`plot_decision_regions`](https://rasbt.github.io/mlxtend/user_guide/plotting/plot_decision_regions/) function now has a `n_jobs` parameter to parallelize the computation. (In a particular use case, on a small dataset, there was a 21x speed-up (449 seconds vs 21 seconds on local HPC instance of 36 cores). ([#998](https://github.com/rasbt/mlxtend/pull/998) via [Khalid ElHaj](https://github.com/Ne-oL))
- Added [`mlxtend.frequent_patterns.hmine`](https://rasbt.github.io/mlxtend/user_guide/frequent_patterns/hmine/) algorithm and documentation for mining frequent itemsets using the H-Mine algorithm. ([#1020](https://github.com/rasbt/mlxtend/pull/1020) via [Fatih Sen](https://github.com/fatihsen20))

New Features and Enhancements

- The `mlxtend.evaluate.feature_importance_permutation` function has a new `feature_groups` argument to treat user-specified feature groups as single features, which is useful for one-hot encoded features. ([955](https://github.com/rasbt/mlxtend/pull/955))
- The `mlxtend.feature_selection.ExhaustiveFeatureSelector` and `SequentialFeatureSelector` also gained support for `feature_groups` with a behavior similar to the one described above. ([957](https://github.com/rasbt/mlxtend/pull/957) and [#965](https://github.com/rasbt/mlxtend/pull/965) via [Nima Sarajpoor](https://github.com/NimaSarajpoor))

Changes

- The `custom_feature_names` parameter was removed from the `ExhaustiveFeatureSelector` due to redundancy and to simplify the code base. The [`ExhaustiveFeatureSelector` documentation](http://rasbt.github.io/mlxtend/user_guide/feature_selection/ExhaustiveFeatureSelector/) illustrates how the same behavior and outcome can be achieved using pandas DataFrames. ([#957](https://github.com/rasbt/mlxtend/pull/957))

Bug Fixes

- None

New Features and Enhancements

Downloads

- [Source code (zip)](https://github.com/rasbt/mlxtend/archive/v0.20.0.zip)

- [Source code (tar.gz)](https://github.com/rasbt/mlxtend/archive/v0.20.0.tar.gz)



New Features and Enhancements

- 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

- Due to compatibility issues with newer package versions, certain functions from six.py have been removed so that mlxtend may not work anymore with Python 2.7.
- As an internal change to speed up unit testing, unit testing is now faciliated by GitHub workflows, and Travis CI and Appveyor hooks have been removed.
- Improved axis label rotation in `mlxtend.plotting.heatmap` and `mlxtend.plotting.plot_confusion_matrix` ([872](https://github.com/rasbt/mlxtend/pull/872))
- Fix various typos in McNemar guides.
- Raises a warning if non-bool arrays are used in the frequent pattern functions `apriori`, `fpmax`, and `fpgrowth`. ([934](https://github.com/rasbt/mlxtend/pull/934) via [NimaSarajpoor](https://github.com/rasbt/mlxtend/issues?q=is%3Apr+is%3Aopen+author%3ANimaSarajpoor))

Bug Fixes

- Fix unreadable labels in `heatmap` for certain colormaps. ([852](https://github.com/rasbt/mlxtend/pull/852))
- Fix an issue in `mlxtend.plotting.plot_confusion_matrix` when string class names are passed ([894](https://github.com/rasbt/mlxtend/pull/894))

New Features

- Adds a second "balanced accuracy" interpretation ("balanced") to `evaluate.accuracy_score` in addition to the existing "average" option to compute the scikit-learn-style balanced accuracy. ([764](https://github.com/rasbt/mlxtend/pull/764))
- Adds new `scatter_hist` function to `mlxtend.plotting` for generating a scattered histogram. ([757](https://github.com/rasbt/mlxtend/issues/757) via [Maitreyee Mhasaka](https://github.com/Maitreyee1))
- The `evaluate.permutation_test` function now accepts a `paired` argument to specify to support paired permutation/randomization tests. ([768](https://github.com/rasbt/mlxtend/pull/768))
- The `StackingCVRegressor` now also supports multi-dimensional targets similar to `StackingRegressor` via `StackingCVRegressor(..., multi_output=True)`. ([802](https://github.com/rasbt/mlxtend/pull/802) via [Marco Tiraboschi](ChromaticIsobar))

Changes

- Updates unit tests for scikit-learn 0.24.1 compatibility. ([774](https://github.com/rasbt/mlxtend/pull/774))
- `StackingRegressor` now requires setting `StackingRegressor(..., multi_output=True)` if the target is multi-dimensional; this allows for better input validation. ([802](https://github.com/rasbt/mlxtend/pull/802))
- Removes deprecated `res` argument from `plot_decision_regions`. ([803](https://github.com/rasbt/mlxtend/pull/803))
- Adds a `title_fontsize` parameter to `plot_learning_curves` for controlling the title font size; also the plot style is now the matplotlib default. ([818](https://github.com/rasbt/mlxtend/pull/818))
- Internal change using `'c': 'none'` instead of `'c': ''` in `mlxtend.plotting.plot_decision_regions`'s scatterplot highlights to stay compatible with Matplotlib 3.4 and newer. ([822](https://github.com/rasbt/mlxtend/pull/822))
- Adds a `fontcolor_threshold` parameter to the `mlxtend.plotting.plot_confusion_matrix` function as an additional option for determining the font color cut-off manually. ([827](https://github.com/rasbt/mlxtend/pull/827))
- The `frequent_patterns.association_rules` now raises a `ValueError` if an empty frequent itemset DataFrame is passed. ([843](https://github.com/rasbt/mlxtend/pull/843))
- The .632 and .632+ bootstrap method implemented in the `mlxtend.evaluate.bootstrap_point632_score` function now use the whole training set for the resubstitution weighting term instead of the internal training set that is a new bootstrap sample in each round. ([844](https://github.com/rasbt/mlxtend/pull/844))

Bug Fixes

- Fixes a typo in the SequentialFeatureSelector documentation ([835](https://github.com/rasbt/mlxtend/issues/835) via [João Pedro Zanlorensi Cardoso](https://github.com/joaozanlorensi))