Given a matrix of observations, computes how many features are out of range.

See ecdf_pvals() for details. It will use Holm's method to correct p-values for every observation and return number of significantly out-of-range features.

Usage

get_n_out_of_range(obs, reference, thr = 0.01)

Arguments

obs

a matrix of observation, where every column is a variable and every row is an observation

reference

either a data.frame with reference variable distributions or a list of stats::ecdf functions computed based on reference distribution. If NULL, cd8ip_features_distribution is used.

thr

p-value threshold

Value

a vector with number of out-of-range features for every sample