Point taken but most data wrangling these days — especially at scale — is of the long and thin variety (what is also known as 3rd normal form or tidy format — which actually allows for more flexibility if you think in terms of coordinatized data theory) where aggregations and joins dominate column operations (Pandas’ also allows array like column operations due to its index but there are other ways to achieve the same thing).
I typical do the type of column operation in your example only on subsets of data, and typically I do it in SQL using DuckDB. Interop between Polars and DuckDB is virtually zero cost so I seamlessly move between the two. And to be honest I don’t remember the last time I needed to do this but that’s just the nature of my work and not a generalized statement.
But yes if you are still in a world where you need to perform Excel like operations then I agree.
I typical do the type of column operation in your example only on subsets of data, and typically I do it in SQL using DuckDB. Interop between Polars and DuckDB is virtually zero cost so I seamlessly move between the two. And to be honest I don’t remember the last time I needed to do this but that’s just the nature of my work and not a generalized statement.
But yes if you are still in a world where you need to perform Excel like operations then I agree.