fix args to DataFrame.set_index

[Dr-Irv]

• Closes [Bug] DataFrame.set_index does not accept pd.Index anymore. #140
• Tests added:
  • test_frame.py:test_types_set_index() from original issue - added one test

[twoertwein]

I think Iterator[Hashable] includes List[Hashable], but I didn't test that.

[Dr-Irv]

I think Iterator[Hashable] includes List[Hashable], but I didn't test that.

No, it's not. Here's an example where pyright shows the incompatibility:

from typing import Iterator


def fun(a: Iterator[int]):
    k = next(a, None)
    while k is not None:
        print(k)


li = [1, 2, 3]
fun(li)

On the last line, pyright reports:

iterlist.py:11:5 - error: Argument of type "list[int]" cannot be assigned to parameter "a" of type "Iterator[int]" in function "fun"
    "list[int]" is incompatible with protocol "Iterator[int]"
      "__next__" is not present (reportGeneralTypeIssues)

[twoertwein]

Iterable[int] wouldn't require __next__ but I'm not sure whether set_index works with any iterable.

[twoertwein]

Iterable[int] wouldn't require __next__ but I'm not sure whether set_index works with any iterable.

pandas has no type annotations for keys. Let's keep it as-is.

[bashtage]

If you believe the docs, then the types that could be used are

Label
Series
Index
np.ndarray
Iterator
List[Union[Label, Series, Index, np.ndarray, Iterator]]

Clearly it also supports List everywhere Series is used.

[twoertwein]

Thanks @Dr-Irv !