Calculate the beta scores for normalized rank vector.
vector of values in [0, 1]
Takes in a vector with values in [0, 1]. It sorts the values to get the order statistics and calculates p-values for each of the order statistics. These are based on their expected distribution under the null hypothesis of uniform distribution.
In RRA algorithm context the inputs are supposed to be normalized ranks. However, p-values in general follow the uniform distribution, therefore it can be used with any kind of p-value vectors, to see if there are more small values than expected.
The NA values are removed before calculation and all results take into account only existing values.
The functions returns a vector of p-values, that correspond to the sorted input vector. The NA-s are pushed to the end.
Raivo Kolde <email@example.com>
Kolde et al "Robust Rank Aggregation for gene list integration and meta-analysis" (in preparation)