DateTime type lost after unique() inside GroupBy Context

[0xbe7a]

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
from datetime import date, timedelta

df = pl.DataFrame(
    {
        "key": ["A", "A", "A", "A"],
        "dates": [date(1970, 1, 1), date(1970, 1, 1), date(1970, 1, 2), date(1970, 1, 3)],
    }
)

# BROKEN
# output [86400000000, 86400000001, 86400000002]
df.groupby("key").agg(pl.col("dates").unique() + timedelta(days=1))



# WORKS
# output [1970-01-02, 1970-01-03, 1970-01-04]
df.groupby("key").agg((pl.col("dates") + timedelta(days=1)).unique())

Issue description

When adding a timedelta to a deduplicated column inside a "GroupBy" context the type is lost and the output has type 'Int64'

Expected behavior

The output type should be 'Date'

Installed versions

--------Version info---------
Polars:              0.18.7
Index type:          UInt32
Platform:            macOS-13.3.1-arm64-arm-64bit
Python:              3.11.4 | packaged by conda-forge | (main, Jun 10 2023, 18:08:41) [Clang 15.0.7 ]

----Optional dependencies----
adbc_driver_sqlite:  <not installed>
connectorx:          <not installed>
deltalake:           <not installed>
fsspec:              <not installed>
matplotlib:          <not installed>
numpy:               1.25.0
pandas:              <not installed>
pyarrow:             <not installed>
pydantic:            <not installed>
sqlalchemy:          <not installed>
xlsx2csv:            <not installed>
xlsxwriter:          <not installed>