Python - Rolling Indexing in Polars library?

I'd like to ask around if anyone knows how to do rolling indexing in polars?I have personally tried a few solutions which did not work for me (I'll show them below):

What I'd like to do: Indexing the number of occurrences within the past X days by NameExample: Let's say I'd like to index occurrences within the past 2 days:

In this case, the counter resets to "1" starting from the past X days (e.g. for 3 Jan 23, it starts "1" from 2 Jan 23), or if a new name is detected

What I've tried:

df.groupby_rolling(index_column='Date', period='2d', by='Name', check_sorted=False).agg((pl.col("Date").rank(method='ordinal')).alias("Counter"))

The above does not work because it outputs:

df.with_columns( Counter=pl.col("mask").rolling_sum(window_size='2d', by="Date") )

Where I made a column "Mask" which is just a column of "1"s, and tried to sum them, but it outputs:

And it also cannot handle "New Guy" correctly because rolling_sum cannot do by="Name", "Date"

df.with_columns(Counter = pl.col("Date").rank(method='ordinal').over(["Name", "Date"]) )

The above code works correctly, but can only be used for indexing within the same day (i.e. period="1d")

Additional Notes: I also did this in Excel, and also using a brute/raw method of using a "for"-loop. Both worked perfectly, however they struggled with huge amounts of data.

What I read:Some references to help in answers: (Most didn't work because they have fixed rolling window instead of a dynamic window by "Date")

How to implement rolling rank in Polars version 0.19

https://github.com/pola-rs/polars/issues/4808

How to do group_by_rolling grouped by day by hour in polars in Python?

How to groupby and rolling in polars？

https://docs.pola.rs/py-polars/html/reference/series/api/polars.Series.rank.html

https://docs.pola.rs/py-polars/html/reference/dataframe/api/polars.DataFrame.groupby_rolling.html