gforce.FDR_control: FDR Control Procedure.

In GFORCE: Clustering and Inference Procedures for High-Dimensional Latent Variable Models

Description

Performs by default the Benjamini-Yekutieli FDR control procedure. Optionally, the Benjamini-Hochberg thresholding rule can be used instead. As input it takes a symmetric matrix of test statistics with standard normal null distributions. The number of hypotheses tested is d(d-1)/2.

Usage

gforce.FDR_control(test_stats, alpha, procedure = "BY")

Arguments

test_stats	d x d symmetric matrix of test statistics.
alpha	alpha level for the FDR control procedure.
procedure	a string. procedure == 'BY' indicates to use the Benjamini-Yekutieli thresholding rule. procedure == 'BH' indicates to use the Benjamini-Hochberg thresholding rule.

Value

An object with following components

reject_null

d x d upper triangular matrix. TRUE entries indicate the null hypothesis should be rejected.

R_tau_hat

an integer. Indicates the number of hypotheses rejected.

tau_hat

a real number. Indicates a threshold above which the null hypothesis is rejected.

num_hypotheses

an integer. Indicates the number of hypotheses tested.

References

Y. Benjamini and Y. Hochberg. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society, 1995.

Y. Benjamini and D. Yekutieli. The Control of the False Discovery Rate in Multiple Testing Under Dependency. The Annals of Statistics, 2001.

GFORCE documentation built on May 2, 2019, 3:44 a.m.