Model Tuner 0.0.34b2

What's Changed

• (+) GroupShuffleSplit by @lshpaner in #217

Full Changelog: 0.0.34b1...0.0.34b2

Model Tuner 0.0.34b1

What's Changed

• Fixed nested lists for binary precision and specificity by correcting average assignment by @lshpaner in #215

Full Changelog: 0.0.33b...0.0.34b

Version 0.0.34b

What's Changed

• Fixed nested lists for binary precision and specificity by correcting average assignment by @lshpaner in #215

Full Changelog: 0.0.33b...0.0.34b

Version 0.0.33b

What's Changed

• Bug in F1 Score metric by @elemets in #213

Full Changelog: 0.0.32b...0.0.33b

0.0.32b

What's Changed

• Multi-Threshold Support to Bootstrapper by @elemets in #212

Full Changelog: 0.0.31b...0.0.32b

Model Tuner 0.0.31b

What's Changed

• Use model’s optimal threshold when predicting on bootstrap resamples by @lshpaner in #209

Full Changelog: 0.0.30b...0.0.31b

Model Tuner 0.0.30b

What's Changed

• Threshold Ranges in Threshold Optimisation by @elemets in #203
• Upd. requirements/versions by @lshpaner in #204

Full Changelog: 0.0.29b1...0.0.30b

Model Tuner 0.0.29b1

Release 0.0.29b1 — Packaging Fix

This release addresses a packaging issue introduced in 0.0.29b related to an invalid dependency format in pyproject.toml. The fix ensures full compliance with PEP 508 and allows installation to succeed across supported environments for Python 3.7.

What changed

• Corrected malformed environment marker for scikit-optimize in project.dependencies
• No functional code changes were made — this is strictly a build fix for PyPI compatibility

Why this matters

• 0.0.29b could not be installed via pip due to a syntax error in the build configuration.
• 0.0.29b1 replaces it as the correct beta release for downstream testing and usage.

Model Tuner 0.0.29b

What's Changed

• Upd. versioning for 3.7 by @lshpaner in #200
• Add addtl' pytests; improved cov; rmv unused code by @lshpaner in #199
• Upd. thresh logic from >= to >; chng nomenclature of y_pred; rmv unus… by @lshpaner in #198

Full Changelog: 0.0.28b...0.0.29b

Model Tuner 0.0.28b

What's Changed

• upd. CalibratedClassifier to use validation set by @lshpaner in #193
• Update docstring for stratify_y in train_val_test_split by @lshpaner in #194

Full Changelog: 0.0.27b...0.0.28b