Addressed issues 192-194

[rachlobay]

Since Issues #192, #193, and #194 are closely related, addressed all three here.

After discussion with Logan, made the following changes to [.epi_df code for the written issue:

• Issue Adjust other_keys or decay to tibble when selecting, renaming columns #192: Any key columns that have been selected out from an epi_df are removed from other_keys in the metadata (refer to the ex in that issue).
• Issue When decaying from epi_df to tibble, remove metadata that doesn't make sense #193: Dropped all metadata when decay to tibble.
• Issue Abort or decay to tibble if epi_df column duplication or renaming invalidates key #194: Abort if column duplication in an epi_df (to be consistent with tibble behaviour).

[brookslogan]

Couple a minor things + a request to add any relevant tests for the new checks. Are all the rename-related things already split out into other issues?

[brookslogan]

Also, misc thing: it looks like git might be using attaching your computer username to the commits here, which might be a bit confusing looking at git logs. Could you please try running either:
(1)

git config user.name "<Your Name>"
git config user.email "<email-addr>@<.....>.edu"

in each cloned repo, or
(2)

git config --global user.name "<Your Name>"
git config --global user.email "<your-email-address>@<.....>.edu"

from anywhere?

You can check for success using

git config --list

in each repo, or

git config --global --list

to view just the global settings.

[rachlobay]

Since the rename issues that were mentioned in #192 and #194 were related, I put them into one issue here, where I added a little more info based on our last discussion.

[dajmcdon]

LGTM once all @brookslogan comments are addressed.

[brookslogan]

Addressed a minor check comment. If checks pass, think this is ready to merge.