Series.mean() does not return an float

[alkatar21]

Describe the bug
Series.mean() is typed as float, but actually returns a numpy.float64.

This errors because I need to call .item() to get the actual float.

To Reproduce

# pyright: strict
import pandas as pd

mean = pd.Series([1, 2, 3, 4]).mean()
print(type(mean)) # numpy.float64 instead of float
mean.item() # Cannot access member "item" for type "float"  Member "item" is unknown Pylance (reportGeneralTypeIssues)

Please complete the following information:

• OS: Linux
• OS Version 20.04
• python 3.8.10
• version of type checker: 1.1.293
• version of installed pandas-stubs: Pylance v2023.2.31

Additional context
I must have the actual float for yaml.safe_dump.

[Dr-Irv]

I couldn't find docs in numpy for numpy.float64.item(), but it is in the type stubs for numpy.

Just needs a change here:

pandas-stubs/pandas-stubs/core/series.pyi

Line 1540 in 300adaf

) -> float: ...

But I'm also guessing that some tests will need to be updated as well, e.g.,

pandas-stubs/tests/test_series.py

Line 294 in 300adaf

def test_types_mean() -> None:

which should be converted to the check(assert_type(... pattern used elsewhere

[ramvikrams]

I think we would also have to change it in frame.pyi because test_frame.py fails mypy checks

[Dr-Irv]

I think we would also have to change it in frame.pyi because test_frame.py fails mypy checks

Not sure that I see what you mean. In frame.pyi, mean() returns a Series, so the np.float64 versus float doesn't apply.