Package 'sdafilter'

Title: Symmetrized Data Aggregation
Description: We develop a new class of distribution free multiple testing rules for false discovery rate (FDR) control under general dependence. A key element in our proposal is a symmetrized data aggregation (SDA) approach to incorporating the dependence structure via sample splitting, data screening and information pooling. The proposed SDA filter first constructs a sequence of ranking statistics that fulfill global symmetry properties, and then chooses a data driven threshold along the ranking to control the FDR. For more information, see the website below and the accompanying paper: Du et al. (2020), "False Discovery Rate Control Under General Dependence By Symmetrized Data Aggregation", <arXiv:2002.11992>.
Authors: Lilun Du [aut, cre], Xu Guo [ctb], Wenguang Sun [ctb], Changliang Zou [ctb]
Maintainer: Lilun Du <[email protected]>
License: GPL (>= 2)
Version: 1.0.0
Built: 2024-12-05 07:02:14 UTC
Source: CRAN

Help Index


Symmetrized Data Aggregation

Description

This is the core function for the paper posted in arXiv preprint arXiv:2002.11992

Usage

SDA_M(dat, alpha, Omega, nonsparse = FALSE, stable = TRUE)

Arguments

dat

a n by p data matrix

alpha

the FDR level

Omega

the inverse covariance matrix; if missing, it will be estimated by the glasso package

nonsparse

if TRUE, the covariance matrix will be estimated by the POET package

stable

if TRUE, the sample will be randomly splitted B=10 times for stability performance; otherwise, only single sample splitting is used.

Value

the indices of the hypotheses rejected

Examples

n = 50
p = 100
dat = matrix(rnorm(n*p), nrow=n)
mu = rep(0, p)
mu[1:as.integer(0.1*p)]=0.3
dat = dat+rep(1, n)%*%t(mu)
alpha = 0.2
out = SDA_M(dat, alpha, diag(p))
print(out)