Add group_mc_cv()

[mikemahoney218]

This PR addresses #207 by adding a new function, group_mc_cv(), which can be used to create grouped Monte Carlo cross-validation resamples. Note that the user-facing function doesn't have a balance argument, because it doesn't make sense to balance anything other than the proportion of data assigned to the assessment fold.

This PR also re-organizes code so that grouped functions now live in the same file as their ungrouped variants. I believe we were agreed that would make the most sense going forward, but happy to undo if not.

library(rsample)
data(ames, package = "modeldata")
set.seed(123)
group_mc_cv(ames, group = Neighborhood, times = 5)
#> # Grouped Monte Carlo cross-validation (0.75/0.25) with 5 resamples  
#> # A tibble: 5 × 2
#>   splits             id       
#>   <list>             <chr>    
#> 1 <split [2182/748]> Resample1
#> 2 <split [2196/734]> Resample2
#> 3 <split [1980/950]> Resample3
#> 4 <split [2056/874]> Resample4
#> 5 <split [2178/752]> Resample5

^{Created on 2022-06-27 by the reprex package (v2.0.1)}

[mikemahoney218]

I think that if we like how this is implemented, then we can also quickly add group_validation_split(), group_initial_split(), and group_bootstraps() (slightly harder, but only slightly). That's everything mentioned in #207, although I think we also want to implement stratification before considering that "done".

[hfrick]

nice!

[juliasilge]

This is looking great! 🙌 I have just a couple of comments, one of which is about an existing example.

[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.