v0.1.7

Quick fix on numpy compatibility.

v0.1.6

• Improvement on package compatibility | @xuyxu
• Add support for pandas dataframe and list | @IncubatorShokuhou
• Minor bug fix | @xuyxu

v0.1.5

This release includes:

• Official support on Python 3.9
• Bug fix for deep forest in the customized mode

v0.1.4

Added

• Add support on customized estimators (#48) @xuyxu
• Add official support for ManyLinux-aarch64 (#47) @xuyxu

Fixed

• Fix the prediction workflow with only one cascade layer (#56) @xuyxu
• Fix inconsistency on predictor name (#52) @xuyxu
• Fix accepted types of target for CascadeForestRegressor (#44) @xuyxu

Improved

• Improve target checks for CascadeForestRegressor (#53) @chendingyan

v0.1.3

Added

• Add multi-output support for CascadeForestRegressor (#40) @Alex-Medium
• Add layer-wise feature importances (#39) @xuyxu
• Add scikit-learn backend (#36) @xuyxu
• Add official support for Mac-OS (#34) @T-Allen-sudo
• Add support on configurable criterion (#28) @tczhao

v0.1.2

Added

• Add CascadeForestRegressor for regression problem (#25) @tczhao

v0.1.1

Added

• Implement the get_forest() method for efficient indexing (#22) @xuyxu
• Support class label encoding (#18) @NiMaZi
• Support sample weight in fit() (#7) @tczhao
• Add configurable predictor parameter (#9) @tczhao
• Add base class BaseEstimator and ClassifierMixin (#8) @pjgao

Fixed

• Fix accepted data types on the binner (#23) @xuyxu