Casting a decimal column of a table into a float type when converting a table into a pandas dataframe

Hey All,

I was wondering what is the most elegant & efficient way of casting a decimal column of a table into a float type, when converting a table into a pandas dataframe?

In the code I'm working on, I got to a situation which a `Table` instance has a column of a `decimal128(6, 5)` data type.
I'd like to convert that into a `float` data type.

I tried the following:
(see [this](https://arrow.apache.org/docs/python/api/datatypes.html#type-checking) & [this](https://arrow.apache.org/docs/python/pandas.html#nullable-types) for reference)

```python
import pandas as pd
import pyarrow as pa

def converter(data_type):
if pa.types.is_decimal(data_type):
return pd.Float64Dtype()
return None

table = ... # contains a decimal column
df = table.to_pandas(types_mapper=converter)
```

But when I tried it I got the following error: `TypeError: Expected array of Float64 type, got decimal128(6, 5) instead`

Can anyone please suggest a better way to cast such column into a float data type?

Thanks in advance!

Start by reproducing the failure through Table.to_pandas(types_mapper=converter) with a decimal128(6, 5) column, then inspect the pandas conversion path and its handling of types_mapper. Done means decimal columns can be converted to a pandas float type without the reported TypeError, with regression coverage for this case.

Written by the indexing model from the issue text.