DOC: add note to docstring about dtype keyword when creating Series from abother Series

[taranarmo]

When creating Series from another Series dtype keyword is not not used. Add note on dtype keyword to Series and DataFrame constructor. See GH #59060.

• closes BUG: dtype keyword in Series constructor ignored for Series input in certain cases #59060
• Tests added and passed if fixing a bug or adding a new feature
• All code checks passed.
• Added type annotations to new arguments/methods/functions.
• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

[mroeschke]

A warning would be too noisy. A note in the docs would be sufficient.

[taranarmo]

Undo the previous changes and add a note to the dtype keyword on pd.Series and pd.DataFrame that says that the keyword is ignored if data is Series and DataFrame correspondingly.

[mroeschke]

Thanks @taranarmo

[jorisvandenbossche]

I was going to comment that I find this a bit surprising behaviour, and so whether we shouldn't rather consider this as a bug instead of documenting it.

But from a quick check, it seems that the dtype argument in fact is not ignored, at least not in all cases (#59060 clearly shows there are also cases where it is indeed ignored):

In [18]: pd.Series(pd.Series([1.0], dtype=float), dtype=int) Out[18]: 0 1 dtype: int64 

Here it does cast the float series to an int64 dtype, so the dtype keyword was not ignored.

Will also comment on the issue, but so it seems the simple "If data is Series then is ignored." message added in this PR is not fully correct.

[jbrockmendel]

IIUC I agree with Joris. The solution is not to change the doc, but to make the behavior consistent. The dtype kwarg should never be silently ignored.