v3.5.0

What's Changed

Features

Splink now should work much more easily in databricks due to the following PR by roberwhiffin: Splink databricks by @robertwhiffin in #915

• Add method with a dictionary of cumulative br counts by @ThomasHepworth in #923

Other

• Add a fix match_key for match_key issue on fix_count_num_comparisons_from_blocking_rules_for_prediction_sql by @ThomasHepworth in #924
• [DOCS] Add m estimation from pairwise (clerical) labels example by @RobinL in #930
• [DOCS] Add data prep pre-requisites section to docs by @RobinL in #931
• Empty training block error by @ADBond in #934
• Add docs for estimate_m_from_pairwise_labels by @kylebutts in #938
• Update and lint docstring by @RobinL in #943
• Bump jsonschema dependency to ensure Splink works in latest jupyterlab by @RobinL in #942
• Bump to 3.5.0 by @RobinL in #944

New Contributors

• @robertwhiffin made their first contribution in #915
• @kylebutts made their first contribution in #938

Full Changelog: v3.4.5...v3.5.0