Nondeterministic execution of value counts + replace

[Ilykuleshov]

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({"foo": [0, 0, 0, 1, 1]})
vc = pl.col("foo").value_counts()
classes = vc.struct[0]
counts = vc.struct[1]
df.select(pl.col("foo").replace(classes, counts))

Log output

No response

Issue description

The reproducible example gives different results each run.

Expected behavior

I expected it to output [3, 3, 3, 2, 2]. It sometimes outputs [2, 2, 2, 3, 3].

Installed versions

--------Version info---------
Polars:               1.5.0
Index type:           UInt32
Platform:             Linux-5.15.0-94-generic-x86_64-with-glibc2.35
Python:               3.10.12 (main, Mar 22 2024, 16:50:05) [GCC 11.4.0]

----Optional dependencies----
adbc_driver_manager:  <not installed>
cloudpickle:          3.0.0
connectorx:           <not installed>
deltalake:            <not installed>
fastexcel:            <not installed>
fsspec:               2024.2.0
gevent:               <not installed>
great_tables:         <not installed>
hvplot:               <not installed>
matplotlib:           3.8.3
nest_asyncio:         1.6.0
numpy:                1.26.4
openpyxl:             <not installed>
pandas:               2.2.1
pyarrow:              15.0.0
pydantic:             <not installed>
pyiceberg:            <not installed>
sqlalchemy:           <not installed>
torch:                2.2.1+cu121
xlsx2csv:             <not installed>
xlsxwriter:           <not installed>```

</details>

[cmdlineluser]

You can make it deterministic with .value_counts(sort=True)

It's a sneaky one - as it doesn't happen with .lazy() because CSE caches the result and value_counts() is executed just once.

[ritchie46]

Right.. Also a tough one to fix. I think we should enable cse on eager from now on.