Add median() to ndframe

[mesejo]

Includes:

• Add implementation of median
• Add docstring of std, var and, median
• Run black

[elasticmachine]

Since this is a community submitted pull request, a Jenkins build has not been kicked off automatically. Can an Elastic organization member please verify the contents of this patch and then kick off a build manually?

[sethmlarson]

Jenkins test this please

[sethmlarson]

This change looks great, one tiny comment!

[sethmlarson]

Should be .var()?

[mesejo]

Yes it should. Done, :)

[mesejo]

@sethmlarson Regarding the refactoring, here are my thoughts:

At least aggs, _metric_aggs, and describe share functionality. All of the functions built aggregation queries, perform some operations, and extract the information from the aggregation. My suggestion is to extract the built aggregation and extraction functionalities from those functions into separate functions (some candidates): built_agg_query and extract_agg_response and refactor the above mentioned functions.

[sethmlarson]

LGTM

[sethmlarson]

@mesejo Would be good to keep terms_aggs() and hist_aggs() in mind as well but don't have to really worry about them yet.

I like the idea of build_X_query() and extract_X_response(). Would be good on the extract_X_response() to have the return type be something that can be returned into a pd.DataFrame() or pd.Series() so it can be used within agg() which returns DataFrame and the individual aggs like median() that return a Series.

Would you be willing to create a quick mockup and submit as a PR? Doesn't need to be mergeable or tested, just something to review and think about how this might look. :) Thanks for your hard work btw, it's really appreciated.

[mesejo]

@sethmlarson I'm willing to do it, but I cannot guarantee that I can deliver it fast. Thank you, :)

[sethmlarson]

@mesejo No worries! Don't break your back over it, the work and design input you're doing now is already more than expected. 🎉