The thing with Polars is it's really hard for me to get past the annoyance of having to do `pl.col("blah")` instead of `df.blah`. I find pandas easier for quick interactive work which is basically everything I do with it.