It's just muscle memor, I've been using Pandas daily for over a decade, but I'll likely eventually switch over to Polars.

Part of the problem, as I said, is that I'm just spoiled by DuckDB when performance matters.