search

pola-rs / tpch

MIT License
64 stars 35 forks source link

add exists -> semi in q4 #126

Closed ritchie46 closed 1 month ago

ritchie46 commented 1 month ago

An SQL exists filters left on availability in right. Thus a semi join.

Before we did an inner join + drop-duplicates, which is incorrect.

This makes the Polars query 12% faster.