agg.ADAGES: Aggregation by ADAGES

In cKarypidis/multiknockoffs: Multiple knockoff procedures

Description

This function aggregates a list of selection sets by ADAGES.

Usage

agg.ADAGES(Shat.list, p)

Arguments

Shat.list	list of K elements containing the selection sets.
p	number of original variables of the model.

Details

The function can be used in combination with multi.knockoffs and multi.knockfilter.

The function does not have to be used in the context of multiple knockoffs. It can also be used to aggregate the selection sets of any FDR controlling procedure in general.

Applies the minimization of the complexity ratio as a criterion to determine the optimal threshold.

Value

A list containing following components:

Shat	aggregated selection set.
c	optimal threshold.
K	number of aggregated sets.

References

Gui (2020). ADAGES: adaptive aggregation with stability for distributed feature selection. Proceedings of the 2020 ACM-IMS on Foundations of Data Science Conference. https://arxiv.org/pdf/2007.10776.pdf

Examples

#Multiple knockoff example
n <- 400; p <- 200; s_0 <- 30
amplitude <- 1; mu <- rep(0,p); rho <- 0.25
Sigma <- toeplitz(rho^(0:(p-1)))
X <- MASS::mvrnorm(n, mu, Sigma)
nonzero <- sample(p, s_0)
beta <- amplitude * (1:p %in% nonzero)
y <- X %*% beta + rnorm(n)

Xk <- multi.knockoffs(X, K = 5)
multi.res <- multi.knockfilter(X, Xk, y)
agg.ADAGES(multi.res$Shat.list, p = p)


#General example (selection sets with indices between 1 and 30)
Shat.list <- list(s1 = c(2,4,3,1,20,30), s2 = c(3,30,23,1,4,8),
                 s3 = c(3,4,5,13,15,12, 23:29, 30), s4 = c(1:10, 13:15, 17),
                 s5 = c(15:20, 23))
agg.ADAGES(Shat.list, p = 30)

cKarypidis/multiknockoffs documentation built on Dec. 19, 2021, 12:53 p.m.