Filter with expression expansion

Question

Is it possible to convert the following filter, which uses two conditions, to something that uses expression expansion or a custom function in order to apply the DRY priciple (avoid the repetition)?
Here is the example:

import polars as pl

df = pl.DataFrame(
    {
        "a": [1, 2, 3, 4, 5],
        "val1": [1, None, 0, 0, None],
        "val2": [1, None, None, 0, 1],
    }
)

df.filter((~pl.col("val1").is_in([None, 0])) | (~pl.col("val2").is_in([None, 0])))

Results in:

┌─────┬──────┬──────┐
│ a   ┆ val1 ┆ val2 │
│ --- ┆ ---  ┆ ---  │
│ i64 ┆ i64  ┆ i64  │
╞═════╪══════╪══════╡
│ 1   ┆ 1    ┆ 1    │
│ 5   ┆ null ┆ 1    │
└─────┴──────┴──────┘

Answer 1

.any_horizontal() and .all_horizontal() can be used to build | and & chains.

.not_() can also be used instead of ~ if you prefer.

df.filter(
    pl.any_horizontal(
        pl.col("val1", "val2").is_in([None, 0]).not_()
    )
)

shape: (2, 3)
┌─────┬──────┬──────┐
│ a   ┆ val1 ┆ val2 │
│ --- ┆ ---  ┆ ---  │
│ i64 ┆ i64  ┆ i64  │
╞═════╪══════╪══════╡
│ 1   ┆ 1    ┆ 1    │
│ 5   ┆ null ┆ 1    │
└─────┴──────┴──────┘