Panic when calling fill_null on Date column with float value

[mkleinbort-ic]

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 numpy as np 

# This code used to work (at least as of version 0.20.21)
(pl
 .DataFrame({
     'date1': ['2020-01-01', '2020-01-02'],
     'date2': ['2020-01-01', None]
  })
 .cast({'date1':pl.Date, 'date2':pl.Date})
 .filter(
     pl.col('date1') > pl.col('date2').fill_null(-np.inf)
    )
 )

Now raises:

PanicException: implementation error, cannot get ref Int32 from Float64

Log output

No response

Issue description

There has been a minor change in behavior which I could not find documented in the release notes.

Expected behavior

I agree with the error, but a better error message would be better.

Installed versions

--------Version info---------
Polars:               0.20.25
Index type:           UInt32
Platform:             Windows-10-10.0.22621-SP0
Python:               3.11.0 (main, Oct 24 2022, 18:26:48) [MSC v.1933 64 bit (AMD64)]

----Optional dependencies----
adbc_driver_manager:  0.11.0
cloudpickle:          3.0.0
connectorx:           0.3.3
deltalake:            <not installed>
fastexcel:            0.10.4
fsspec:               2023.12.2
gevent:               <not installed>
hvplot:               <not installed>
matplotlib:           3.8.4
nest_asyncio:         1.6.0
numpy:                1.26.4
openpyxl:             <not installed>
pandas:               2.2.2
pyarrow:              15.0.0
pydantic:             2.6.1
pyiceberg:            0.6.1
pyxlsb:               <not installed>
sqlalchemy:           2.0.30
torch:                2.3.0+cpu
xlsx2csv:             0.8.2
xlsxwriter:           3.2.0

[stinodego]

For clarity: this should not work, but the error message could be better. More minimal repro:

from datetime import date

import polars as pl

df = pl.DataFrame({"date": [date(2022, 1, 1), None]})
result = df.select(pl.col("date").fill_null(1.0))
print(result)

thread '<unnamed>' panicked at crates/polars-core/src/series/mod.rs:938:13:
implementation error, cannot get ref Int32 from Float64
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Traceback (most recent call last):
  File "/home/stijn/code/polars/py-polars/repro.py", line 6, in <module>
    result = df.select(pl.col("date").fill_null(1.0))
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/stijn/code/polars/py-polars/polars/dataframe/frame.py", line 8132, in select
    return self.lazy().select(*exprs, **named_exprs).collect(_eager=True)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/stijn/code/polars/py-polars/polars/lazyframe/frame.py", line 1816, in collect
    return wrap_df(ldf.collect(callback))
                   ^^^^^^^^^^^^^^^^^^^^^
pyo3_runtime.PanicException: implementation error, cannot get ref Int32 from Float64

For your use case, I would recommend filling with date.min instead.