feature request by user:
in case of multiple tasks
subset.bootstrap.list(ranking_bootstrapped, top = 10)
could provide top 10 performing algorithms according to consensus ranking. Currently an error is given
"Subset of algorithms only sensible for single-task challenges. Otherwise no consensus ranking is possible."
the complication arises because in the task specific plots, the selected algorithms should provide the task specific rank, such that for a specific task there may be no rank 1 for example because the best algorithm for that task has been omitted because it is not among the best 10 according to consensus.