Redundant cast inside expression changes result in 0.20.23

[lukeshingles]

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({"x": range(5)})
print(df.select((pl.col("x") * 1.0e9 / 10)))
print(df.select((pl.col("x") * 1.0e9 / 10).cast(pl.Float64)))

Log output

shape: (5, 1)
┌─────┐
│ x   │
│ --- │
│ f64 │
╞═════╡
│ 0.0 │
│ 1e8 │
│ 2e8 │
│ 3e8 │
│ 4e8 │
└─────┘
shape: (5, 1)
┌──────┐
│ x    │
│ ---  │
│ f64  │
╞══════╡
│ 0.0  │
│ 1e8  │
│ 2e8  │
│ null │
│ null │

Issue description

This seems to be some kind of calculation range issue that gets triggered by the integer in the expression (replacing 10 with 10.0 prevents the issue) and the cast to float as part of the expression, which should not change the result, since the column resulting from the expression is already Float64.

Expected behavior

In 0.20.22, the two DataFrames as equal, as expected.

Installed versions

--------Version info---------
Polars:               0.20.23
Index type:           UInt32
Platform:             macOS-14.4.1-arm64-arm-64bit
Python:               3.12.3 (main, Apr  9 2024, 08:09:14) [Clang 15.0.0 (clang-1500.3.9.4)]

----Optional dependencies----
adbc_driver_manager:  <not installed>
cloudpickle:          <not installed>
connectorx:           <not installed>
deltalake:            <not installed>
fastexcel:            <not installed>
fsspec:               <not installed>
gevent:               24.2.1
hvplot:               <not installed>
matplotlib:           3.8.4
nest_asyncio:         1.6.0
numpy:                1.26.4
openpyxl:             <not installed>
pandas:               2.2.2
pyarrow:              16.0.0
pydantic:             <not installed>
pyiceberg:            <not installed>
pyxlsb:               <not installed>
sqlalchemy:           <not installed>
xlsx2csv:             <not installed>
xlsxwriter:           <not installed>

[lukeshingles]

Sorry if this is a duplicate. This seems to be related to #15951, which I've confirmed also became an issue with polars 0.20.23.

[lukeshingles]

This is fixed in polars 0.20.24.