Skip to content

v0.4.0

Choose a tag to compare

View all tags
@sipemu sipemu released this 17 Jan 20:04
· 54 commits to main since this release
v0.4.0
d954694

New Features

Robust Regression

  • quantile - Quantile regression (median and arbitrary quantiles)
  • isotonic - Isotonic (monotonic) regression using PAVA algorithm
  • Quantile and Isotonic model classes for scikit-learn-style API

Regression Diagnostics

  • condition_number - Detect multicollinearity via condition number analysis
  • check_binary_separation - Detect complete/quasi-complete separation in logistic regression
  • check_count_sparsity - Detect sparsity issues in Poisson/count regression

Conda-forge Support

  • Added conda-forge recipe for conda install polars-statistics
  • CI workflow for conda package building and testing

Documentation

  • Reorganized API documentation into modular structure
  • Added detailed descriptions for all statistical tests

Changes

  • Updated anofox-regression to v0.5.1
  • Updated pyo3 to v0.27
  • Updated faer to v0.23.2
  • Improved cross-platform test compatibility for GLM functions
  • Enhanced PyPI discoverability with 40+ keywords

Installation

pip install polars-statistics==0.4.0

Full Changelog: v0.3.0...v0.4.0

Assets 2
Loading