Add test for pl.len() on LazyFrames

[CangyuanLi]

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

df = pl.DataFrame({
    "id1": [1, 1, 1, 2, 3],
    "id2": [100, 100, 200, 100, 300],
})

print("DataFrame")
print(df.height)
print(df.unique().height)
print(df.unique("id1").height)

print(df.unique().sort(["id1", "id2"]))

# DataFrame
# 5
# 4
# 4
# 3
# shape: (4, 2)
# ┌─────┬─────┐
# │ id1 ┆ id2 │
# │ --- ┆ --- │
# │ i64 ┆ i64 │
# ╞═════╪═════╡
# │ 1   ┆ 100 │
# │ 1   ┆ 200 │
# │ 2   ┆ 100 │
# │ 3   ┆ 300 │
# └─────┴─────┘

print("LazyFrame")
lf = df.lazy()
print(lf.select(pl.len()).collect().item())
print(lf.unique().select(pl.len()).collect().item())
print(lf.select("id1").unique().select(pl.len()).collect().item())
# print(lf.unique("id1").select(pl.len()).collect().item())

print(lf.unique().collect().sort(["id1", "id2"]))

# LazyFrame
# 5
# 3
# 3
# shape: (4, 2)
# ┌─────┬─────┐
# │ id1 ┆ id2 │
# │ --- ┆ --- │
# │ i64 ┆ i64 │
# ╞═════╪═════╡
# │ 1   ┆ 100 │
# │ 1   ┆ 200 │
# │ 2   ┆ 100 │
# │ 3   ┆ 300 │
# └─────┴─────┘

Log output

Issue description

I am not sure the extent of the issue, but I noticed that pl.len(), in combination with .unique(), reports the wrong results on LazyFrames when doing unique by more than one column. The correct results are shown on the DataFrame. Note that .unique() itself seems to be fine, as the resulting DataFrames are equivalent. Also,

print(lf.unique("id1").select(pl.len()).collect().item())

reports a ColumnNotFoundError, which seems related to #21343

Expected behavior

The expected behavior is that the DataFrame heights match up with the LazyFrame pl.len().

Installed versions

--------Version info---------
Polars:              1.22.0
Index type:          UInt32
Platform:            Linux-5.14.0-284.97.1.el9_2.x86_64-x86_64-with-glibc2.28
Python:              3.10.13 (main, Sep 11 2023, 13:21:10) [GCC 11.2.0]
LTS CPU:             False

----Optional dependencies----
Azure CLI            <not installed>
adbc_driver_manager  <not installed>
altair               5.5.0
azure.identity       <not installed>
boto3                1.24.28
cloudpickle          3.1.1
connectorx           <not installed>
deltalake            <not installed>
fastexcel            0.11.6
fsspec               2024.12.0
gevent               <not installed>
google.auth          2.38.0
great_tables         0.16.1
matplotlib           3.8.3
numpy                1.23.5
openpyxl             <not installed>
pandas               2.2.1
pyarrow              15.0.0
pydantic             <not installed>
pyiceberg            <not installed>
sqlalchemy           2.0.37
torch                <not installed>
xlsx2csv             <not installed>
xlsxwriter           <not installed>

[cmdlineluser]

I can reproduce the mismatch on 1.22.0

Testing on main, I get matching results - so it may already be fixed. (no recent commit stands out to me from a quick glance)

# DataFrame
# 5
# 4
# 3
# LazyFrame
# 5
# 4
# 3

The ColumnNotFoundError seems to have been fixed: #21154

>>> lf.unique("id1").select(pl.len()).collect().item()
3

[MarcoGorelli]

thanks for the report

so it may already be fixed. (no recent commit stands out to me from a quick glance)

in that case, it may be good to add this snippet as a test, to make sure it doesn't accidentally regress in the future