mhtboot (version 1.3.3)

qelbow: Finding corner of a quantile of ordered transformed p values

Description

Given a matrix of empirical distribution of ordered transformed p values, this function finds the corner point for a particular quantile.

Usage

qelbow(porder, rbuff = 25, h = 30, qi = 0.9)

Arguments

porder

matrix, usually feed from pboot functions. Bxm matrix of ordered p values, where B is the replication size and m is dimension.

rbuff

right buffer, scaler, control for elbow()

h

window size, default 30.

qi

number between 0 and 1, quantile of the distribution. default 0.9.

Value

vector with two elements. estimates of the corner point by two methods.

Details

In the distribution of the transformed ordered p values, we choose a particular quantile given by the user. We estimate the change point, which will be an estimate of the number of true alternatives corresponding to that quantile of the p values. As the values of the quantile increases, the estimates can only increasing, because we are dealing with ordered p values.

Examples

    
## Not run:     
# n = 50;m = 250;m0 = 20;
# sigeff = 1;
# Sigma <- 0.25*diag(m)
# X <- datgen(n,m,m0,sigeff,Sigma = Sigma)
# porder <- pboot.1sample(X=X,B=100,ncpus = 1)
# out <- qelbow(porder = porder)
# out
# ## End(Not run)