Changelog

Change tags (adopted from Scikit-learn and mvlearn):

  • [Major Feature] : something big that you couldn’t do before.

  • [Feature] : something that you couldn’t do before.

  • [Efficiency] : an existing feature now may not require as much computation or memory.

  • [Enhancement] : a miscellaneous minor improvement.

  • [Fix] : something that previously didn’t work as documentated – or according to reasonable expectations – should now work.

  • [API] : you will need to change your code to have the same effect in the future; or a feature will be removed in the future.

Version 0.3.1

May 28, 2025

Updates in this release:

imml.classify

  • [Fix] Fixed device when processing images in M3Care.

  • [Fix] Better alignment with the original implementation of RAGPT.

imml.cluster

  • [Fix] Fixed eigensolver to use eigsh instead of eigs for symmetric kernel matrices in EEIMVC, ensuring equivalence with the original MATLAB implementation.

  • [Fix] Fixed error in original Octave implementation SIMCADC.

  • [Fix] Revised Python translation in LFIMVC.

  • [Fix] Revised Python translation in DAIMC.

  • [Fix] Revised Python translation in NEMO.

  • [API] Block_size parameter changed to batch_size in OPIMC.

imml.model_selection

  • [Fix] Fixed default return_type parameter in MMSplitter.

imml.preprocessing

  • [Fix] Fixed select_complete_samples and select_incomplete_samples when working with string and object dtypes.

  • [Fix] Removed redundant mask calculation in select_incomplete_samples.

Version 0.3.0

March 17, 2025

Updates in this release:

  • [Major Feature] We have created a new module imml.model_selection.

  • [Enhancement] We have added a "See also" section to related classes.

  • [Enhancement] We have added a section to link tutorials with classes.

  • [Enhancement] iMML supports now Python 3.14.

  • [API] Matlab module and arguments were replaced by Octave to better reflect their usage.

  • [Efficiency] Imports of optional modules have been centralized.

  • [API] SimpleModImputer and simple_mod_imputer was removed. You can use MMTransformer(transformer = SimpleImputer()) instead.

  • [Efficiency] snfpy package was removed from the requirements.

imml.impute

  • [API] SimpleModImputer and simple_mod_imputer was removed. You can use MMTransformer(transformer = SimpleImputer()) instead.

imml.model_selection

  • [Feature] MMSplitter was added.

  • [Feature] train_test_mm_split was added.

imml.preprocessing

  • [API] Multi_Mod_Transformer was renamed to MMTransformer.

imml.statistics

  • [API] pid now returns also the total information.

imml.utils

  • [API] check_Xs was converted to check_Xs_y.

imml.visualize

  • [Feature] plot_summary was added.

  • [Feature] plot_combinations was added.

Version 0.2.0

November 3, 2025

Updates in this release:

  • [Fix] Corrected inheritance hierarchy in clustering algorithms by replacing ClassifierMixin with the appropriate ClusterMixin base class from Scikit-learn.

  • [Enhancement] Improved code readability by updating references to Lightning package base classes to use their explicit class names instead of generic references.

  • [Enhancement] Enhanced navigation in the algorithm selection guide by adding direct hyperlinks from each algorithm to its corresponding detailed documentation page, making it easier for users to explore specific implementations.

  • [Efficiency] numba package was removed from the requirements.

imml.ampute

imml.classify

  • [Fix] MUSE Fixed text extractor load when using text modality.

  • [Fix] M3Care Fixed error when working with multiple data modalities.

imml.impute

  • [Enhancement] MissingModIndicator Now support lists and pytorch tensors.

  • [Enhancement] get_missing_mod_indicator Now support lists and pytorch tensors.

  • [Enhancement] ObservedModIndicator Now support lists and pytorch tensors.

  • [Enhancement] get_observed_mod_indicator Now support lists and pytorch tensors.

imml.load

  • [API] M3CareDataset observed_mod_indicator argument was removed.

  • [API] MUSEDataset observed_mod_indicator and y_indicator arguments were removed.

imml.utils

Version 0.1.1

October 17, 2025

Updates in this release:

  • [Enhancement] Improving documentation for several methods.

  • [Enhancement] Improved documentation for installation and extra dependencies.

  • [Enhancement] Adding guidelines on how to choose an algorithm.

  • [Enhancement] Added license headers to all files.

  • [Fix] Fixed iPython dependency issue. Oct2Py depends on iPython but returned an error when importing ipython>=9.0.0.

.github/workflows/ci_test.yml

  • [Fix] Fixing actions/missing-workflow-permissions security.

Version 0.1.0

October 03, 2025

We are happy to announce the first major public version of iMML!