This is the first part of a series of PRs to improve conditional_join, with the final aim being to support aggregations within conditional_join. Supporting aggregations within conditional_join means we can avoid materializing the dataframe before aggregating, and should generally improve performance for such operations.

PR Description

Please describe the changes proposed in the pull request:

more control on how the final dataframe is created
special focus on non inner join operations ('left','right','outer')

This PR relates to #1269 .

Please tag maintainers to review.

@ericmjl

pyjanitor-devs / pyjanitor

[ENH] More control on creating the output dataframe for `conditional_join` - PR no 1 #1396

PR Description