Releases | Open Source Science

v0.7.1 - 2025-07-25

Changed

improved memory requirements estimate

- C++
Published by paulbkoch 7 months ago

v0.7.0 - 2025-07-13

Breaking Changes

the shape of the bags parameter has been changed from (n_outer_bags, n_samples) to (n_samples, n_outer_bags) in order to better match the format of the X parameter which has shape (n_samples, n_features) In this release if bags is passed in as (n_outer_bags, n_samples), then a warning is issued and the shape is accepted using the old format.

- C++
Published by paulbkoch 8 months ago

v0.6.16 - 2025-07-09

Fixed

removed the use of large blocks of shared memory since it is not available in docker containers

- C++
Published by paulbkoch 8 months ago

v0.6.15 - 2025-07-08

Added

added estimate_mem function to estimate the memory usage of an EBM model ### Changed
small decrease in required memory usage for EBM models

- C++
Published by paulbkoch 8 months ago

v0.6.14 - 2025-07-04

Fixed

removed potential large delay introduced in v0.6.13 while fitting EBMs on some datasets

- C++
Published by paulbkoch 8 months ago

v0.6.13 - 2025-06-28

Added

support for early termination of EBM training using a callback mechanism

- C++
Published by paulbkoch 8 months ago

v0.6.12 - 2025-06-17

Changed

support for numpy 2.3.x
increased default number of interaction terms

- C++
Published by paulbkoch 8 months ago

v0.6.11 - 2025-06-05

Changed

increased max_rounds to 50,000
possibly faster prediction in some scenarios (unverified)
remove obsolete dash components from requirements

- C++
Published by paulbkoch 9 months ago

v0.6.10 - 2025-03-26

Added

reorder_classes function which allows reordering of the classes after fitting
support for ARM based Linux ### Changed
changed default to max_leaves=2 for classification
changed default to n_jobs=2
changed default to outer_bags=14 ### Fixed
restrict to dash 2.x since visualizations are not working on dash 3.x

- C++
Published by paulbkoch 11 months ago

v0.6.9 - 2025-01-06

Added

refitting of the intercept term after fitting the rest of the model to improve the intercept value
new options for handling missing values: "low", "high", "separate", and "gain"
use Fischer (1958) for handling categorical values. This is the same method employed by LightGBM.
added new parameters to control overfitting of nominal categoricals: gain_scale, min_cat_samples, cat_smooth ### Changed
enable AVX-512 by default
modified default EBM parameters: outer_bags=16, n_jobs=-1 ### Fixed
fixed memory leak in the purification function

- C++
Published by paulbkoch about 1 year ago

v0.6.7 - 2024-11-27

Changed

minimum python version increased to 3.9
minimum numpy version increased to 1.25 ### Fixed
removed scipy dependency to resolve Issue #588

- C++
Published by paulbkoch over 1 year ago

v0.6.6 - 2024-10-20

Changed

added predict_with_uncertainty function by @degenfabian in PR #584
handle mono-classification in SHAP by @degenfabian in PR #582
improvements to tree building in C++ ### Fixed
issue that develop/debug options were not being honored in Windows when 1<n_jobs in joblib
fix several bugs in C++ from negative hessians or negative gain values caused by floating point noise

- C++
Published by paulbkoch over 1 year ago

v0.6.5 - 2024-10-23

Changed

default EBM parameters changed to improve model performance
switch to using exact versions of exp/log instead of the previously used approximate versions ### Fixed
fix issue where very large feature values fail in the UI PR #581 by @degenfabian

- C++
Published by paulbkoch over 1 year ago

v0.6.4 - 2024-09-28

Added

support for regularization parameters reg_alpha, and reg_lambda in EBMs
support for the parameter max_delta_step in EBMs
improved fitting speed for most of the alternative objectives

- C++
Published by paulbkoch over 1 year ago

v0.6.3 - 2024-08-07

Added

visualizations for the APRL (Automatic Piecewise Linear Regression) package by @mathias-von-ottenbreit ### Changed
early_stopping_tolerance default changed to 1e-5 to reduce EBMs fitting time slightly
shuffle initial feature order within each bag and during greedy boosting ### Fixed
fixed numpy 2.0 issue in the Marginal class

- C++
Published by paulbkoch over 1 year ago

v0.6.2 - 2024-06-22

Added

pass optional kwargs to DecisionTreeClassifier in PR #537 by @busFred
support for multiclass purification
support for higher dimensional purification
allow higher levels of purification than would be supported via the tolerance parameter ### Changed
numpy 2.0 support for EBMs
update documentation regarding monotonicity in PR #531 by @Krzys25
moved purification utility from "interpret/glassbox/_ebm/_research" to "interpret.utils" ### Fixed
possible fix for issue #543 where merge_ebms was creating unexpected NaN values

- C++
Published by paulbkoch over 1 year ago

v0.6.1 - 2024-04-14

Fixed

added compatibility with numpy 2.0 thanks to @DerWeh in PR #525
fixed bug that was preventing SIMD from being used in python
removed approximate division in SIMD since the approximation was too inaccurate ### Changed
EBM fitting time reduced

- C++
Published by paulbkoch almost 2 years ago

v0.6.0 - 2024-03-16

Added

Documentation on recommended hyperparameters to help users optimize their models.
Support for monotone_constraints during model fitting, although post-processed monotonization is still suggested/preferred.
The EBMModel class now includes _more_tags for better integration with the scikit-learn API, thanks to contributions from @DerWeh. ### Changed
Default max_rounds parameter increased from 5,000 to 25,000, for improved model accuracy.
Numerous code simplifications, additional tests, and enhancements for scikit-learn compatibility, thanks to @DerWeh.
The greedy boosting algorithm has been updated to support variable-length greedy sections, offering more flexibility during model training.
Full compatibility with Python 3.12.
Removal of the DecisionListClassifier from our documentation, as the skope-rules package seems to no longer be actively maintained. ### Fixed
The sweep function now properly returns self, correcting an oversight identified by @alvanli.
Default exclude parameter set to None, aligning with scikit-learn's expected defaults, fixed by @DerWeh.
A potential bug when converting features from categorical to continuous values has been addressed.
Updated to handle the new return format for TreeShap in the SHAP 0.45.0 release. ### Breaking Changes
replaced the greediness __init__ parameter with greedy_ratio and cyclic_progress parameters for better control of the boosting process (see documentation for notes on greedy_ratio and cyclic_progress)
replaced breakpoint_iteration_ with best_iteration_, which now contains the number of boosting steps rather than the number of boosting rounds

- C++
Published by paulbkoch almost 2 years ago

v0.5.1 - 2024-02-08

Added

Added new __init__ parameter: interactionsmoothingrounds
Added new __init__ parameter: min_hessian
synthetic dataset generator (make_synthetic) for testing GAMs and for documentation ### Changed
default parameters have been modified to improve the accuracy of EBMs
changed boosting internals to use LogitBoost to improve accuracy
changed interaction detection to use hessians to improve interaction selection
enabled smoothing_rounds by default to improve the smoothness of EBMs
added the ability to specify interactions via feature names or negative indexing
improved the speed of Morris sensitivity and partial dependence
python 3.12 support for core EBMs. Some of our optional dependencies do not yet support python 3.12 though
made early stopping more consistent and changed the earlystoppingtolerance to be a percentage ### Fixed
avoid displaying a scroll bar by default in jupyter notebook cells
removed the dependency on deprecated distutils ### Breaking Changes
changed the internal representation for classifiers that have just 1 class

- C++
Published by paulbkoch about 2 years ago

v0.5.0 - 2023-12-13

Added

added support for AVX-512 in PyPI installations to improve fitting speed
introduced an option to disable SIMD optimizations through the debug_mode function in python
exposed public utils.link_func and utils.inv_link functions ### Changed
the interpret-core package now installs the dependencies required to build and predict EBMs by default without needing to specify the [required] pip install flag
experimental/private support for OVR multiclass EBMs
added bagged_intercept_ attribute to store the intercepts for the bagged models ### Fixed
resolved an issue in merge_ebms where the merge would fail if all EBMs in the merge contained features with only one bin (issue #485)
resolved multiple future warnings from other packages ### Breaking Changes
changed how monoclassification (degenerate classification with 1 class) is expressed
replaced predict_and_contrib function with simpler eval_terms function that returns only the per-term contribution values. If you need both the contributions and predictions use: interpret.utils.inv_link(ebm.eval_terms(X).sum(axis=1) + ebm.intercept_, ebm.link_)
separate to_json into to_jsonable (for python objects) and to_json (for files) functions
create a new link function string for multiclass that is separate from binary classification
for better scikit-learn compliance, removed the decision_function from the ExplainableBoostingRegressor

- C++
Published by paulbkoch about 2 years ago

v0.4.4 - 2023-08-26

Added

added the following model editing functions: copy, removeterms, removefeatures, sweep, scale
added experimental support for a JSON exporter function: to_json

- C++
Published by paulbkoch over 2 years ago

v0.4.3 - 2023-08-04

Changed

Training speed improvements due to the use of SIMD on Intel processors in PyPI release Results may vary, but expect approx 2.75x faster for classification and 1.3x faster for RMSE regression
Changed from using 64-bit floats to using 32-bit floats internally. Regression performed on datasets with large targets that sum to greater than 3.4E+38 will overflow. ### Fixed
Fixed an issue with the monotonize function that would occur when monotonizing a feature with missing values
Resolved issue where excluding the 1st feature would cause an exception

- C++
Published by paulbkoch over 2 years ago

v0.4.2 - 2023-05-31

Added

support for specifying outer bags ### Changed
exceptions raised in the joblib child processes will be re-raised in the main process rather than be expressed as a TerminatedWorkerError
small additional improvements in memory compression
small improvements in maximizing the benefit of the privacy budget for Differentially Private EBMs ### Fixed
fixed segfault that was occurring in the Anaconda build
fixed a bug that would prevent Differentially Private EBMs from using the exclude parameter

- C++
Published by paulbkoch over 2 years ago

[v0.4.1] - 2023-05-16

Added

support for visualizations in streamlit ### Fixed
fixed dangling pointer issue in call to CalcInteractionStrength

- C++
Published by paulbkoch almost 3 years ago

[v0.4.0] - 2023-05-11

Added

alternative objective functions: poissondeviance, tweediedeviance, gammadeviance, pseudohuber, rmse_log (log link)
greediness __init__ parameter that allows selecting a behavior between cyclic boosting and greedy boosting
smoothing_rounds __init__ parameter
added type hints to the EBM __init__ parameters and class attributes
init_score parameter to allow boosting and prediction on top of a previous model
multiclass support in merge_ebms
ability to monotonize features using post process model editing ### Changed
default BaseLinear regressor is changed from Lasso to LinearRegression class
placed limits on the amount of memory used to find interactions with high cardinality categoricals ### Fixed
validationsize of 0 is now handled by disabling earlystopping and using the final model ### Breaking Changes
replaced the __init__ param "mains" with "exclude"
removed the binning __init__ param as this functionality was already fully supported in feature_types
removed the unused zerovalcount attribute and n_samples attribute
renamed the noisescale attribute to noisescaleboosting_ and added noisescalebinning_ to DPEBMs

- C++
Published by paulbkoch almost 3 years ago

[v0.3.2] - 2023-03-14

Fixed

fix the issue that the shared library would only work on newer linux versions

- C++
Published by paulbkoch almost 3 years ago

[v0.3.1] - 2023-03-13

Added

Mac m1 support in conda-forge
SPOTGreedy prototype selection (PR #392) ### Fixed
fix visualization when both cloud and non-cloud environments are detected (PR #210)
fix ShapTree bug where it was treating classifiers as regressors
resolve scikit-learn warnings occurring when models were trained using Pandas DataFrames
change the defaults to prefer 'continuous' over 'nominal' when a feature has 1 or 2 unique float64 values ### Breaking Changes
in the blackbox and greybox explainers, change from accepting a predictfn to accepting either a model or a predictfn
feature type 'categorical' has been renamed to 'nominal' for the remaining feature_type parameters in the package (EBMs were already using 'nominal')
removed the unused sampler parameters to the Explainer classes

- C++
Published by paulbkoch almost 3 years ago

[v0.3.0] - 2022-11-16

Added

Full Complexity EBMs with higher order interactions supported: GA3M, GA4M, GA5M, etc... 3-way and higher-level interactions lose exact global interpretability, but retain exact local explanations Higher level interactions need to be explicitly specified. No automatic FAST detection yet
Mac m1 support
support for ordinals
merge_ebms now supports merging models with interactions, including higher-level interactions
added classic composition option during Differentially Private binning
support for different kinds of feature importances (avgweight, minmax)
exposed interaction detection API (FAST algorithm)
API to calculate and show the importances of groups of features and terms. ### Changed
memory efficiency: About 20x less memory is required during fitting
predict time speed improvements. About 50x faster for Pandas CategoricalDType, and varying levels of improvements for other data types
handling of the differential privacy DPOther bin, and non-DP unknowns has been unified by having a universal unknown bin
bin weights have been changed from per-feature to per-term and are now multi-dimensional
improved scikit-learn compliance: We now conform to the scikit-learn 1.0 feature names API by using self.featurenamesin_ for the X column names and self.nfeaturesin. We use the matching self.featuretypesin for feature types, and self.termnames for the additive term names. ### Fixed
merge_ebms now distributes bin weights proportionally according to volume when splitting bins
DP-EBMs now use sample weights instead of bin counts, which preserves privacy budget
improved scikit-learn compliance: The following init attributes are no longer overwritten during calls to fit: self.interactions, self.featurenames, self.featuretypes
better handling of floating point overflows when calculating gain and validation metrics ### Breaking Changes
EBMUtils.mergemodels function has been renamed to mergeebms
renamed binning type 'quantilehumanized' to 'roundedquantile'
feature type 'categorical' has been specialized into separate 'nominal' and 'ordinal' types
EBM models have changed public attributes:
- feature_groups_ -> term_features_ global_selector -> n_samples_, unique_val_counts_, and zero_val_counts_ domain_size_ -> min_target_, max_target_ additive_terms_ -> term_scores_ bagged_models_ -> BaseCoreEBM has been depricated and the only useful attribute has been moved into the main EBM class (bagged_models_.model_ -> bagged_scores_) feature_importances_ -> has been changed into the function term_importances(), which can now also generate different types of importances preprocessor_ & pair_preprocessor_ -> attributes have been moved into the main EBM model class (details below)
EBMPreprocessor attributes have been moved to the main EBM model class
- col_names_ -> feature_names_in_ col_types_ -> feature_types_in_ col_min_ -> feature_bounds_ col_max_ -> feature_bounds_ col_bin_edges_ -> bins_ col_mapping_ -> bins_ hist_counts_ -> histogram_counts_ hist_edges_ -> histogram_edges_ col_bin_counts_ -> bin_weights_ (and is now a per-term tensor)

- C++
Published by paulbkoch over 3 years ago

v0.2.7 - 2021-09-23

Added

Synapse cloud support for visualizations. ### Fixed
All category names in bar charts now visible for inline rendering (used in cloud environments).
Joblib preference was previously being overriden. This has been reverted to honor the user's preference.
Bug in categorical binning for differentially privatized EBMs has been fixed.

- C++
Published by interpret-ml over 4 years ago

v0.2.6 - 2021-07-20

Adde6

Differential-privacy augmented EBMs now available as interpret.privacy.{DPExplainableBoostingClassifier,DPExplainableBoostingRegressor}.
Packages interpret and interpret-core now distributed via docker. ### Changed
Sampling code including stratification within EBM now performed in native code. ### Fixed
Computer provider with joblib can now support multiple engines with serialization support.
Labels are now all shown for inline rendering of horizontal bar charts.
JS dependencies updated.

- C++
Published by interpret-ml over 4 years ago

v0.2.5 - 2021-06-21

Added

Sample weight support added for EBM.
Joint predict_and_contrib added to EBM where both predictions and feature contributions are generated in one call.
EBM predictions now substantially faster with categorical featured predictions.
Preliminary documentation for all of interpret now public at https://interpret.ml/docs.
Decision trees now work in cloud environments (InlineRenderer support).
Packages interpret and interpret-core now distributed via sdist. ### Fixed
EBM uniform binning bug fixed where empty bins can raise exceptions.
Users can no longer include duplicate interaction terms for EBM.
CSS adjusted for inline rendering such that it does not interfere with its hosting environment.
JS dependencies updated. ### Experimental
Ability to merge multiple EBM models into one. Found in interpret.glassbox.ebm.utils.

- C++
Published by interpret-ml over 4 years ago

v0.2.4 - 2021-01-19

Fixed

Bug fix on global EBM plots.
Rendering fix for AzureML notebooks. ### Changed
JavaScript dependencies for inline renderers updated.

- C++
Published by interpret-ml about 5 years ago

v0.2.3 - 2021-01-13

Major upgrades to EBM in this release. Automatic interaction detection is now included by default. This will increase accuracy substantially in most cases. Numerous optimizations to support this, especially around binary classification. Expect similar or slightly slower training times due to interactions.

Fixed

Automated interaction detection uses low-resolution binning for both FAST and pairwise training. ### Changed
EBM argument has been reduced from outer_bags=16 to outer_bags=8.
EBM now includes interactions by default from interactions=0 to interactions=10.
Algorithm treeinterpreter is now unstable due to upstream dependencies.
Automated interaction detection now operates from two-pass to one-pass.
Numeric approximations used in boosting (i.e. approx log / exp).
Some arguments have been re-ordered for EBM initialization.

- C++
Published by interpret-ml about 5 years ago

interpret - Version 0.2.2

v0.2.2 - 2020-10-19

Fixed

Fixed bug on predicting unknown categories with EBM.
Fixed bug on max value being placed in its own bin for EBM pre-processing.
Numerous native fixes and optimizations. ### Added
Added max_interaction_bins as argument to EBM learners for different sized bins on interactions, separate to mains.
New binning method 'quantile_humanized' for EBM. ### Changed
Interactions in EBM now use their own pre-processing, separate to mains.
Python 3.5 no longer supported.
Switched from Python to native code for binning.
Switched from Python to native code for PRNG in EBM.

- C++
Published by interpret-ml over 5 years ago

interpret - Version 0.2.1

v0.2.1 - 2020-08-07

Added

Python 3.8 support. ### Changed
Dash based visualizations will always default to listen port 7001 on first attempt; if the first attempt fails it will try a random port between 7002-7999. ### Experimental (WIP)
Further cloud environment support.
Improvements for multiclass EBM global graphs.

- C++
Published by interpret-ml over 5 years ago

interpret - Version 0.2.0

v0.2.0 - 2020-07-21

Breaking Changes

With warning, EBM classifier adapts internal validation size when there are too few instances relative to number of unique classes. This ensures that there is at least one instance of each class in the validation set.
Cloud Jupyter environments now use a CDN to fix major rendering bugs and performance.
- CDN currently used is https://unpkg.com
- If you want to specify your own CDN, add the following as the top cell ```python from interpret import setvisualizeprovider from interpret.provider import InlineProvider from interpret.version import version
Change this to your custom CDN.

JSURL = "https://unpkg.com/@interpretml/interpret-inline@{}/dist/interpret-inline.js".format(version) setvisualizeprovider(InlineProvider(jsurl=JS_URL)) ```
EBM has changed initialization parameters:
- schema -> DROPPED n_estimators -> outer_bags holdout_size -> validation_size scoring -> DROPPED holdout_split -> DROPPED main_attr -> mains data_n_episodes -> max_rounds early_stopping_run_length -> early_stopping_rounds feature_step_n_inner_bags -> inner_bags training_step_epsiodes -> DROPPED max_tree_splits -> max_leaves min_cases_for_splits -> DROPPED min_samples_leaf -> ADDED (Minimum number of samples that are in a leaf) binning_strategy -> binning max_n_bins -> max_bins
EBM has changed public attributes:
- ``` nestimators -> outerbags holdoutsize -> validationsize scoring -> DROPPED holdoutsplit -> DROPPED mainattr -> mains datanepisodes -> maxrounds earlystoppingrunlength -> earlystoppingrounds featurestepninnerbags -> innerbags trainingstepepsiodes -> DROPPED maxtreesplits -> maxleaves mincasesforsplits -> DROPPED minsamplesleaf -> ADDED (Minimum number of samples that are in a leaf) binningstrategy -> binning maxnbins -> max_bins
attributesets -> featuregroups attributesetmodels_ -> additiveterms (Pairs are now transposed) modelerrors -> termstandarddeviations_

mainepisodeidxs_ -> breakpointiteration[0] interepisodeidxs_ -> breakpointiteration[1]

meanabsscores_ -> featureimportances ```

Fixed
Internal fixes and refactor for native code.
Updated dependencies for JavaScript layer.
Fixed rendering bugs and performance issues around cloud Jupyter notebooks.
Logging flushing bug fixed.
Labels that are shaped as nx1 matrices now automatically transform to vectors for training.

Experimental (WIP)
Added support for AzureML notebook VM.
Added local explanation visualizations for multiclass EBM.

- C++
Published by interpret-ml over 5 years ago

interpret - Version 0.1.22

v0.1.22 - 2020-04-27

Upcoming Breaking Changes

EBM initialization arguments and public attributes will change in a near-future release.
There is a chance Explanation API will change in a near-future release. ### Added
Docstrings for top-level API including for glassbox and blackbox. ### Fixed
Minor fix for linear models where class wasn't propagating for logistic. ### Experimental
For research use, exposed optionaltempparams for EBM's Python / native layer.

- C++
Published by interpret-ml almost 6 years ago

interpret - Version 0.1.21

v0.1.21 - 2020-04-02

Added

Module "glassbox.ebm.research" now has purification utilities.
EBM now exposes "maxnbins" argument for its preprocessing stage. ### Fixed
Fix intercept not showing for local EBM binary classification.
Stack trace information exposed for extension system failures.
Better handling of sparse to dense conversions for all explainers.
Internal fixes for native code.
Better NaN / infinity handling within EBM. ### Changed
Binning strategy for EBM now defaulted to 'quantile' instead of 'uniform'.

- C++
Published by interpret-ml almost 6 years ago

interpret - Version 0.1.20

v0.1.20 - 2019-12-11

Fixed

Major bug fix around EBM interactions. If you use interactions, please upgrade immediately. Part of the pairwise selection was not operating as expected and has been corrected.
Fix for handling dataframes when no named columns are specified.
Various EBM fixes around corner-case datasets. ### Changed
All top-level methods relating to show's backing web server now use visualize provider directly. In theory this shouldn't affect top-level API usage, but please raise an issue in the event of failure.
Memory footprint heavily reduced for EBM at around 2-3 times.

- C++
Published by interpret-ml about 6 years ago