API: In Series.map, change the default value for na_action to "ignore""

[topper-123]

In Pandas 2.1, Series.map with the parameter na_ignore set to na_ignore="ignore" will work for all array types. (see #51809 + the various PRs that add na_action="ignore" to the various ExtensionArrays). Previously the support for na_action="ignore" was quite spotty.

IMO na_action="ignore" is a better default value than na_action=None (the current default), because 99 % of the time when I use .map, I want it to ignore Nans and assume it’s the same for others. E.g.:

>>> import numpy as np 
>>> import pandas as pd
>>> ser = pd.Series(["a", np.nan, "b"])
>>> ser.map(str.upper)
TypeError: descriptor 'upper' for 'str' objects doesn't apply to a 'float' object
>>> ser.map(str.upper, na_action="ignore")
0      A
1    NaN
2      B
dtype: object

So I propose deprecating None as the default for na_action in favor of "ignore".