ENH: consistency of input args for boundaries

[attack68]

There are some methods where boundary inputs are required.

This issue proposes standardising to the keyword arg: inclusive and using values {"both", "neither", "left", "right"}.

• Styler.highlight_between(): 'inclusive' in {"both", "neither", "left", "right"}
• Series.between() (ENH: New boundary inputs #40628): 'inclusive' in {"both", "neither", "left", "right"}
• pd.date_range() ENH: consistency of input args for boundaries (pd.date_range) #43504
• DataFrame.between_time() ENH: consistency of input args for boundaries in DataFrame.between_time() #40245 #43248
• pd.Interval ENH: consistency of input args for boundaries - Interval #46522
• pd.interval_range ENH: consistency of input args for boundaries - pd.interval_range #46355
• pd.IntervalIndex ENH: consistency of input args for boundaries - Interval #46522
• pd.cut
• Series.resample
• DataFrame.rolling(closed)

More methods yet to be identified.

[hewittk]

take

[hewittk]

Does anyone have the link for the Styler.highlight_between proposed method?

[hewittk]

Does anyone have the link for the Styler.highlight_between proposed method?

Also, are there any other methods or functions that you have in mind?
@attack68

[attack68]

#39821 (comment)

so you can see part of the discussion there, that PR is not yet approved, possibly this is a small part of the reason. There is mention of some other functions (non-exhaustive) list that have boundary treatment arguments. Do you know how well numpy handles this, and what style they adopt?

[attack68]

one possible solution if it check outs might be to extend Series.between to adopt {both, neither, left, right}, whilst maintaining True is 'both' and False is 'neither' for backwards compatibility.

I also dont think anyone would object to a logical enhancement of the input argument to Series.between.

^just a quick suggestion.