version 0.6.0 release

CHANGELOG.md

@@ -4,6 +4,27 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and the versioning is mostly derived from [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [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
+
 ## [v0.5.1] - 2024-02-08
 ### Added
 - Added new init parameter: interaction_smoothing_rounds

docs/interpret/hyperparameters.md

@@ -67,7 +67,7 @@ guidance: greedy_ratio is a good candidate for hyperparameter tuning as the best
 ## cyclic_progress
 default: 1.0
 
-hyperparameters: [0.0, 0.25, 0.5, 1.0]
+hyperparameters: [0.0, 0.5, 1.0]
 
 guidance: cyclic_progress is a good candidate for hyperparameter tuning as the best value is dataset dependent.
 

python/interpret-core/interpret/_version.py

@@ -2,4 +2,4 @@
 # Distributed under the MIT software license
 
 # NOTE: Version is replaced by a regex script.
-__version__ = "0.5.1"
+__version__ = "0.6.0"

python/interpret-core/interpret/glassbox/_ebm/_ebm.py

@@ -2421,18 +2421,18 @@ class ExplainableBoostingClassifier(EBMModel, ClassifierMixin, ExplainerMixin):
         Maximum number of leaves allowed in each tree.
     monotone_constraints: list of int, default=None
 
-        This parameter allows you to specify monotonic constraints for each feature's 
-        relationship with the target variable during model fitting. However, it is 
-        generally recommended to apply monotonic constraints post-fit using the 
-        `monotonize` function rather than setting them during the fitting process. 
-        This recommendation is based on the observation that, during fitting, 
-        the boosting algorithm may compensate for a monotone constraint on one 
-        feature by utilizing another correlated feature, potentially obscuring 
+        This parameter allows you to specify monotonic constraints for each feature's
+        relationship with the target variable during model fitting. However, it is
+        generally recommended to apply monotonic constraints post-fit using the
+        `monotonize` function rather than setting them during the fitting process.
+        This recommendation is based on the observation that, during fitting,
+        the boosting algorithm may compensate for a monotone constraint on one
+        feature by utilizing another correlated feature, potentially obscuring
         any monotonic violations.
 
-        If you choose to define monotone constraints, `monotone_constraints` 
-        should be a list with a length equal to the number of features. 
-        Each element in the list corresponds to a feature and should take 
+        If you choose to define monotone constraints, `monotone_constraints`
+        should be a list with a length equal to the number of features.
+        Each element in the list corresponds to a feature and should take
         one of the following values:
 
             - 0: No monotonic constraint is imposed on the corresponding feature's partial response.
@@ -2742,18 +2742,18 @@ class ExplainableBoostingRegressor(EBMModel, RegressorMixin, ExplainerMixin):
         Maximum number of leaves allowed in each tree.
     monotone_constraints: list of int, default=None
 
-        This parameter allows you to specify monotonic constraints for each feature's 
-        relationship with the target variable during model fitting. However, it is 
-        generally recommended to apply monotonic constraints post-fit using the 
-        `monotonize` function rather than setting them during the fitting process. 
-        This recommendation is based on the observation that, during fitting, 
-        the boosting algorithm may compensate for a monotone constraint on one 
-        feature by utilizing another correlated feature, potentially obscuring 
+        This parameter allows you to specify monotonic constraints for each feature's
+        relationship with the target variable during model fitting. However, it is
+        generally recommended to apply monotonic constraints post-fit using the
+        `monotonize` function rather than setting them during the fitting process.
+        This recommendation is based on the observation that, during fitting,
+        the boosting algorithm may compensate for a monotone constraint on one
+        feature by utilizing another correlated feature, potentially obscuring
         any monotonic violations.
 
-        If you choose to define monotone constraints, `monotone_constraints` 
-        should be a list with a length equal to the number of features. 
-        Each element in the list corresponds to a feature and should take 
+        If you choose to define monotone constraints, `monotone_constraints`
+        should be a list with a length equal to the number of features.
+        Each element in the list corresponds to a feature and should take
         one of the following values:
 
             - 0: No monotonic constraint is imposed on the corresponding feature's partial response.

python/interpret-core/setup.py

@@ -11,7 +11,7 @@
 from setuptools.command.sdist import sdist
 
 # NOTE: Version is replaced by a regex script.
-version = "0.5.1"
+version = "0.6.0"
 
 
 def _copy_native_code_to_setup():

python/interpret/setup.py

@@ -5,7 +5,7 @@
 
 name = "interpret"
 # NOTE: Version is replaced by a regex script.
-version = "0.5.1"
+version = "0.6.0"
 long_description = """
 In the beginning machines learned in darkness, and data scientists struggled in the void to explain them.
 

scripts/release_process.txt

@@ -33,6 +33,16 @@
   - bdist: interpret-*-py3-none-any.whl
   - bdist: interpret_core-*-py3-none-any.whl
 
+- check the docs
+  - cd <PACKAGE_DOWNLOAD_DIRECTORY>
+  - unzip the docs.zip file
+  - open one of the html files and go to the first document in the list
+  - do a side by side browser comparison to the existing documentation at: https://interpret.ml/docs
+  - clone the repo: https://github.com/interpretml/docs
+  - delete all the files, except possibly for ".gitignore" (TODO: can we remove .gitignore even since all the files are uploaded?)
+  - copy the new files into that repo
+  - commit changes to the docs repo, BUT DO NOT PUSH YET
+
 - test the bdist:
   - open anaconda console window
   - cd <PACKAGE_DOWNLOAD_DIRECTORY>
@@ -76,15 +86,6 @@ set_visualize_provider(InlineProvider())
   - upload the R package at <PACKAGE_DOWNLOAD_DIRECTORY> to test on multiple platforms in: https://builder.r-hub.io
   - In particular, the "Oracle Developer Studio 12.6" is worth testing as that C++ compiler is picky, and CRAN tests it
 
-- check the docs
-  - cd <PACKAGE_DOWNLOAD_DIRECTORY>
-  - unzip the docs.zip file
-  - open one of the html files and go to the first document in the list
-  - do a side by side browser comparison to the existing documentation at: https://interpret.ml/docs
-  - clone the repo: https://github.com/interpretml/docs
-  - delete all the files, except possibly for ".gitignore" (TODO: can we remove .gitignore even since all the files are uploaded?)
-  - copy the new files into that repo, BUT DO NOT PUSH YET
-
 - publish on NPM
   - cd <PACKAGE_DOWNLOAD_DIRECTORY>
   - we can't re-use previously published versions, but the visualization code is likely to remain unchanged, and our 

shared/vis/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@interpretml/interpret-inline",
-  "version": "0.5.1",
+  "version": "0.6.0",
   "description": "Interpret inline library for rendering visualizations across all notebook environments.",
   "main": "index.js",
   "keywords": [],