Currently, when you pass a series into daft.lit, it actually uses that series as essentially a column instead of a singular list value, which I find to be kind of strange compared to other literal types. Is this intended behavior? For example, when I run

>>> import daft
>>> df = daft.from_pydict({"foo": [1, 2, 3], "bar": ["a", "b", "c"]})
>>> s = daft.Series.from_pylist(["x", "y", "z"])
>>> df = df.with_column("baz", daft.lit(s))
>>> df.show()

I get

╭───────┬──────┬──────╮
│ foo   ┆ bar  ┆ baz  │
│ ---   ┆ ---  ┆ ---  │
│ Int64 ┆ Utf8 ┆ Utf8 │
╞═══════╪══════╪══════╡
│ 1     ┆ a    ┆ x    │
├╌╌╌╌╌╌╌┼╌╌╌╌╌╌┼╌╌╌╌╌╌┤
│ 2     ┆ b    ┆ y    │
├╌╌╌╌╌╌╌┼╌╌╌╌╌╌┼╌╌╌╌╌╌┤
│ 3     ┆ c    ┆ z    │
╰───────┴──────┴──────╯

when instead I would expect

╭───────┬──────┬─────────────────╮
│ foo   ┆ bar  ┆ baz             │
│ ---   ┆ ---  ┆ ---             │
│ Int64 ┆ Utf8 ┆ List[Utf8]      │
╞═══════╪══════╪═════════════════╡
│ 1     ┆ a    ┆ ['x', 'y', 'z'] │
├╌╌╌╌╌╌╌┼╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤
│ 2     ┆ b    ┆ ['x', 'y', 'z'] │
├╌╌╌╌╌╌╌┼╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤
│ 3     ┆ c    ┆ ['x', 'y', 'z'] │
╰───────┴──────┴─────────────────╯

i think i came across this a few months ago as well and found the behavior odd.

It's cuz

eval_expression

on a

lit(series)

just returns the series.

Is ur intention to create a list column from a literal list?

We could turn it into a listarray with one element and it would be broadcasted properly.

'

We could turn it into a listarray with one element and it would be broadcasted properly.

Oh sweet! Yeah that’ll work

I guess the question I really have is if doing that would be a breaking change or if this is actually just a bug

It would be a bug, and would fail for example if u try and append a series lit to a concated dataframe.

Sounds good, will make a PR to fix this then

Awesome, just a heads up I believe the ‘is_in’ and ‘count_matches’ expressions depend on series lits

gotcha, thanks for the heads up