Selecting a particular set of strings from a list in polars

df = pl.DataFrame({'list_column': [['a.xml', 'b.xml', 'c', 'd'], ['e.xml', 'f.xml', 'g', 'h']]})

def func(x):
    return [y for y in x if '.xml' in y]

df.with_columns(pl.col('list_column').map_elements(func, return_dtype=pl.List(pl.String)))

Is there a way to achieve the same without using map_elements?

>Solution :

There is an accepted request for list.filter()

• https://github.com/pola-rs/polars/issues/9189

You can emulate the behaviour using list.eval()

df.with_columns(
   pl.col('list_column').list.eval(
      pl.element().filter(pl.element().str.ends_with('.xml'))
   )
)

shape: (2, 1)
┌────────────────────┐
│ list_column        │
│ ---                │
│ list[str]          │
╞════════════════════╡
│ ["a.xml", "b.xml"] │
│ ["e.xml", "f.xml"] │
└────────────────────┘