group_by with map_elements result incorrectly gets wrapped in a list for lazyframes

[heshamdar]

Checks

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of Polars.

Reproducible example

import polars as pl
import os
os.environ['POLARS_VERBOSE'] = '1'


def simple_func(x):
    return 1


df = pl.DataFrame({"a": [1, 2, 3, 4, 5]})
out = df.group_by(pl.lit(None)).agg(pl.col("a").map_elements(
        lambda x: simple_func(x), return_dtype=pl.Int64)
    )
print(out.collect_schema())

df = df.lazy()
out = df.group_by(pl.lit(None)).agg(pl.col("a").map_elements(
        lambda x: simple_func(x), return_dtype=pl.Int64)
    )
print(out.collect_schema())

Log output

Schema([('literal', Null), ('a', Int64)])
Schema([('literal', Null), ('a', List(Int64))])
keys/aggregates are not partitionable: running default HASH AGGREGATION

Issue description

When running a group_by aggregation with map_elements in lazy mode, even if the return datatype is specified it will return a schema with that datatype wrapped in a list. This does not happen when running in eager mode.

Expected behavior

The return datatype specified by the user should be honoured, e.g. in the example above, the function is set to return just a single value, and not assume that a list of values is being returned.

Installed versions

--------Version info---------
Polars:               1.0.0-beta.1
Index type:           UInt32
Platform:             macOS-14.2.1-arm64-arm-64bit
Python:               3.10.4 (main, May 28 2022, 12:23:44) [Clang 13.1.6 (clang-1316.0.21.2.5)]

----Optional dependencies----
adbc_driver_manager:  <not installed>
cloudpickle:          3.0.0
connectorx:           <not installed>
deltalake:            0.15.3
fastexcel:            0.9.1
fsspec:               2024.3.1
gevent:               <not installed>
great_tables:         0.6.0
hvplot:               <not installed>
matplotlib:           3.8.3
nest_asyncio:         1.6.0
numpy:                1.26.4
openpyxl:             3.1.2
pandas:               2.0.3
pyarrow:              15.0.0
pydantic:             2.7.0
pyiceberg:            <not installed>
sqlalchemy:           2.0.29
torch:                <not installed>
xlsx2csv:             0.8.2
xlsxwriter:           <not installed>