Argument 1 to "DatetimeIndex" has incompatible type "dict[str, str]"

[randolf-scholz]

import pandas as pd

anomalies = pd.DatetimeIndex({
    "Jan. 1, 2008": "New Year’s Day",
    "Jan. 21, 2008": "Martin Luther King Jr. Day",
    "Feb. 18, 2008": "Washington’s Birthday",
    "Mar. 9, 2008": "Anomaly day",
    "May 26, 2008": "Memorial Day",
    "Jul. 4, 2008": "Independence Day",
    "Sep. 1, 2008": "Labor Day",
    "Oct. 13, 2008": "Columbus Day",
    "Nov. 11, 2008": "Veterans Day",
    "Nov. 27, 2008": "Thanksgiving",
    "Dec. 25, 2008": "Christmas Day",
    "Jan. 1, 2009": "New Year’s Day",
    "Jan. 19, 2009": "Martin Luther King Jr. Day",
    "Feb. 16, 2009": "Washington’s Birthday",
    "Mar. 8, 2009": "Anomaly day",
})

[twoertwein]

dict[str, str] is not officially documented but Iterable[str] (excluding str itself) might be "array-like" and would allow dict[str, Any].

[randolf-scholz]

Btw. I have a potential workaround for the Iterable[str] vs str, but it's not sure if it will be accepted: python/cpython#112319

[twoertwein]

Btw. I have a potential workaround for the Iterable[str] vs str, but it's not sure if it will be accepted: python/cpython#112319

In pandas we use python/typing#256 (comment) (but that is for a Sequence, not an Iterable - but Sequence might be more array-like than any iterable).

[Dr-Irv]

While this does work, it isn't documented. Related issue in pandas here: pandas-dev/pandas#55425

[Dr-Irv]

In pandas dev meeting on 2/12/25, we agreed to accept "ordered iterables", so that would include what is now defined as Axes in the stubs.

Need to modify the arguments in various Index.__new__() and subclasses of Index in either __new__() and/or __init__() to accept Axes rather than Iterable (or whatever else is there)