DataFrame.insert should accept np.float32

[tblum]

Describe the bug
Calling DataFrame.insert with a np.float32 will fail with the following error:

error: Argument 3 to "insert" of "DataFrame" has incompatible type "floating[_32Bit]"; expected "str | bytes | date | datetime | timedelta | <7 more items> | complex | Sequence[Any] | ndarray[Any, Any] | Series[Any] | Index[Any] | None" [arg-type]

[loicdiridollou]

Hey @tblum,
The issue you are facing is linked to the fact that Scalar contains float but this type is not the same as float32.
@Dr-Irv is this something that we should include in the stubs?

[Dr-Irv]

The issue you are facing is linked to the fact that Scalar contains float but this type is not the same as float32.
@Dr-Irv is this something that we should include in the stubs?

Possibly. I'd like to see a more complete example from @tblum before we address this. Would be curious to know if np.float64 also fails, so we may need to add that as well.

[loicdiridollou]

import numpy as np
import pandas as pd

df = pd.DataFrame({"A": [1, 2, 3]})

df.insert(3, "A", value=np.float32(23.3))
df.insert(3, "A", value=np.float64(23.3))

Both give out the same error:

file.py:6: error: Argument "value" to "insert" of "DataFrame" has incompatible type "floating[_32Bit]";
expected "str | bytes | date | datetime | timedelta | <7 more items> | complex | Sequence[Any] | 
ndarray[Any, Any] | Series[Any] | Index[Any] | None"  [arg-type]
file.py:7: error: Argument "value" to "insert" of "DataFrame" has incompatible type "floating[_64Bit]";
 expected "str | bytes | date | datetime | timedelta | <7 more items> | complex | Sequence[Any] |
 ndarray[Any, Any] | Series[Any] | Index[Any] | None"  [arg-type]
Found 2 errors in 1 file (checked 1 source file)

[Dr-Irv]

I think if we add np.floating to IndexIterScalar in _typing.pyi, that should address this issue.

PR with tests welcome.