automatically select a rank cutoff for given ranks and p-values
current implementation: sort the p-values according to rank. select as rank cutoff
the rank for which the worst (i.e. max) p-value in the top list is still smaller than
the best (i.e. min) p-value of the group of worst-ranking p-values of equal size as the top-list
Usage
auto.select.rank.cut(p, r, alpha = 0.1)
Arguments
p
vector of p-values
r
vector of ranks
alpha
the percentile to select the top and bottom part of the list