Refactor datatype logic

[mayer79]

This is a large PR that refactors the way missRanger() deals with variables that cannot be directly modeled by ranger(). The new implementation is slightly more picky, but also more safe.

It is an important step towards out-of-sample application (#58).

Here a summary:

• Columns of special type like date/time can't be imputed anymore.
• pmm() is more picky: xtrain and xtest must both be either numeric, logical, or factor (with identical levels).
• Now requires ranger >= 0.16.0.
• More compact vignettes.
• Many relevant ranger() arguments are now explicit arguments in missRanger() to improve tab-completion experience:
  • num.trees = 500
  • mtry = NULL
  • min.node.size = NULL
  • min.bucket = NULL
  • max.depth = NULL
  • replace = TRUE
  • sample.fraction = if (replace) 1 else 0.632
  • case.weights = NULL
  • num.threads = NULL
  • save.memory = FALSE
• Slightly more info before fitting.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

Attention: Patch coverage is 87.17949% with 15 lines in your changes missing coverage. Please review.

Project coverage is 87.83%. Comparing base (20247e3) to head (77037c1).
Report is 11 commits behind head on main.

Files	Patch %	Lines
R/missRanger.R	84.94%	14 Missing ⚠️
R/pmm.R	94.11%	1 Missing ⚠️

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

Additional details and impacted files

@@             Coverage Diff             @@
##             main      #68       +/-   ##
===========================================
- Coverage   97.95%   87.83%   -10.13%     
===========================================
  Files           5        5               
  Lines         245      263       +18     
===========================================
- Hits          240      231        -9     
- Misses          5       32       +27     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.