AZtrees data

[tdhock]

Closes #4

there are two different group concepts:

• train on one group, predict on another: SameOther group, maybe call it group_generalization?
• keep rows together when resampling: usual mlr3 group, maybe call it group_atomic? we need two different names. current mlr3 names are

  > str(task.obj$col_roles)
  List of 7
  $ feature: chr [1:256] "V2" "V3" "V4" "V5" ...
  $ target : chr "y"
  $ name   : chr(0) 
  $ order  : chr(0) 
  $ stratum: chr [1:2] "predefined.set" "y"
  $ group  : chr "predefined.set"
  $ weight : chr(0) 

  we can create new roles:

  > task.obj$col_roles$group_atomic <- "predefined.set"
  > str(task.obj$col_roles)
  List of 8
  $ feature     : chr [1:256] "V2" "V3" "V4" "V5" ...
  $ target      : chr "y"
  $ name        : chr(0) 
  $ order       : chr(0) 
  $ stratum     : chr [1:2] "predefined.set" "y"
  $ group       : chr "predefined.set"
  $ weight      : chr(0) 
  $ group_atomic: chr "predefined.set"

  TODO example and vignette using AZtrees data.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 83.56%. Comparing base (6024174) to head (55993c5).

:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #12 +/- ## ======================================= Coverage 83.56% 83.56% ======================================= Files 4 4 Lines 213 213 ======================================= Hits 178 178 Misses 35 35 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.