Adds tidy-method form summary_emm-objects.

[crsh]

Currently, the tidy()-methods for various objects from emmeans/lsmeans summarize and subsequently tidy the input. This PR separates these two steps and adds a slightly extended tidier for summary_emm objects (returned by, for example, emmeans::joint_tests()).

[alexpghayes]

Hey! Thanks for the PR! I don't use emmeans enough to understand what's going on here. Can you give me the big picture motivation behind this change?

[crsh]

Hi Alex,

of course! My main motivation for this PR is that emmeans has a function named joint_tests() that provides an ANOVA-like table (rather than testing terms individually):

> fit <- lm(sales1 ~ price1 + price2 + day + store, data = oranges)
> joint_tests_summary <- joint_tests(fit)
> joint_tests_summary

 model term df1 df2 F.ratio p.value
 day          5  23   4.879  0.0035
 store        5  23   2.522  0.0583

The output of joint_tests() is currently not supported by broom because there is no method for summarized emmeans objects:

> class(joint_tests_summary)
[1] "summary_emm" "data.frame" 

Currently the tidy()-methods only apply to non-summarized objects; there is no method for objects of class summary_emm.

By separating the call to summary() from the subsequent tidying, support for summary_emm is straightforward. This PR performs this separation, adds a summary_emm-method, and adds tests to verify that joint_tests()-output is tidied correctly.

Another motivation is that I use tidy() in my package papaja to standardize the input to my function that subsequently typeset the results. For greater flexibility I like to make my functions work on summarized as well as non-summarized objects.

[alexpghayes]

Sounds good, I'm on board. Let's just make sure that the information in your comment above makes it into the documentation!

[crsh]

Thanks for the comments. I just pushed a commit that makes the requested changes.

[alexpghayes]

Can you merge in recent updates from the master branch? (Let me know if you'd like git advice!)

[crsh]

Of course! There should be no more merge conflicts now.

[alexpghayes]

Wonderful, thank you! Does this close #692?

[crsh]

I'm afraid, closing that issue will be a little more work and another PR. I've started on it but it'll be a little longer.

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.