Upsampling a polars dataframe with groupby

Question

I'm trying to upsample a Polars dataframe while grouping by a particular column. In the following example, I wish to group by 'fruit' and then upsample by date.

fruit	date	count
apple	2022-06-01	5
apple	2022-06-03	6
apple	2022-06-04	2
apple	2022-06-07	1
pear	2022-06-01	9
pear	2022-06-07	12

This is what the output should look like:

fruit	date	count
apple	2022-06-01	5
apple	2022-06-02	5
apple	2022-06-03	6
apple	2022-06-04	2
apple	2022-06-05	2
apple	2022-06-06	2
apple	2022-06-07	1
pear	2022-06-01	9
pear	2022-06-02	9
pear	2022-06-03	9
pear	2022-06-04	9
pear	2022-06-05	9
pear	2022-06-06	9
pear	2022-06-07	12

For a non group-by scenario, the following command gets me the result I need:

df.upsample('date', every='1d').fill_null("forward")

However, I've not been able to get it working when a groupby is involved

ps: here is a similar question, but using pandas - Pandas: resample timeseries with groupby

Answer 1

I realized that the upsample function has a 'by' parameter that gives me the results that I need. Here is a link to API docs for the .upsample() method.