The implemented tests cover:

• For single-task challenges:
  • aggregate-than-rank by mean works with one algorithm for one case
  • aggregate-than-rank by mean works with two algorithms for one case (different orders in data frame, small values are better, large values are better)
  • different aggregation functions
  • different ties methods
  • invalid ties method
  • assignment of worst rank for NA
  • unused NA treatment option

• For multi-task challenges:
  • aggregate-than-rank by mean works with no missing data
  • assignment of worst rank for NA
  • unused NA treatment option