Calculate percentage values of groups vs total of values per column

Consider the following dataframe:

import polars as pl

df = pl.DataFrame({
    "id": [1, 1, 2, 2, 2],
    "sku1": [4, 2, 3, None, 1],
    "sku2": [None, 3, None, 3, None],
})

I want to calculate the percentage of each sum per identifier over the total per sku.
So, the end result should like this:

Any ideas? Please note that the number of sku columns might vary over time.

>Solution :

import polars as pl

df = pl.DataFrame({
    "id": [f'identifier {x}' for x in [1, 1, 2, 2, 2]],
    "sku1": [4, 2, 3, None, 1],
    "sku2": [None, 3, None, 3, None],
})

kcols = ['id']
vcols = pl.all().exclude(kcols)


print(df.group_by(kcols).agg(vcols.sum()).with_columns(vcols / vcols.sum()))

shape: (2, 3)
┌──────────────┬──────┬──────┐
│ id           ┆ sku1 ┆ sku2 │
│ ---          ┆ ---  ┆ ---  │
│ str          ┆ f64  ┆ f64  │
╞══════════════╪══════╪══════╡
│ identifier 1 ┆ 0.6  ┆ 0.5  │
│ identifier 2 ┆ 0.4  ┆ 0.5  │
└──────────────┴──────┴──────┘