Skip to content

v1.4.2 — Geo-Compound Auto-Config Blocking

Choose a tag to compare

View all tags
@benzsevern benzsevern released this 06 Apr 18:21
v1.4.2
658aa47

Geo-compound blocking in auto-config

Auto-config now compounds geo columns (state, city, county) with name columns in blocking keys. This prevents cross-region false positives where entities with the same name in different states were incorrectly matched.

What changed

  • When auto-config builds name-based blocking keys, it checks for available geo columns and compounds them with the name column
  • Picks the geo column that reduces max block size the most while staying under the safe block limit
  • Guards against None from empty group_by().max() results

Impact

On the CMS Hospital General Information dataset (5,426 records), this eliminates cross-state false positives where "MEMORIAL HOSPITAL" in IL was being matched with "MEMORIAL HOSPITAL" in IN.

Install

pip install goldenmatch==1.4.2

Full Changelog: v1.4.1...v1.4.2

Assets 2
Loading