README

A new Integrated Mean Variance Correlation and Its Use in High-Dimensional Data Analysis

The goal of package newIMVC is to provide an easy way to implement the proposed methods in Xiong et al. (2024), which include a new robust correlation between continuous variables and its use in hypothesis test, feature screening and false discovery rate control.

Installation

To install newIMVC,

install.packages("newIMVC")

Example

Here are examples showing how to use main functions in package newIMVC.

library("newIMVC")
library("mvtnorm")
#> Warning: package 'mvtnorm' was built under R version 4.3.3
###The new IMVC measure###
n=200
x=rnorm(n)
y=x^2+rt(n,2)
IMVC(y,x,K=10,type="nonlinear")
#> [1] 0.2225841
###IMVC based feature screening###
n=200
p=300
pho1=0.8
mean_x=rep(0,p)
sigma_x=matrix(NA,nrow = p,ncol = p)
for (i in 1:p) {
  for (j in 1:p) {
    sigma_x[i,j]=pho1^(abs(i-j))
  }
}
x=rmvnorm(n, mean = mean_x, sigma = sigma_x,method = "chol")
x1=x[,1]
x2=x[,2]
x3=x[,12]
x4=x[,22]
y=2*x1+0.5*x2+3*x3*ifelse(x3<0,1,0)+2*x4+rnorm(n)
IMVCS(y,x,K=5,d=round(n/log(n)),type="nonlinear")
#>  [1]   1   2  22   3   4  23  12  21  13   5  14  11  15  24   6   7  25   8   9
#> [20]  20  10 104  19  16 212 224 156  39  17 168  18 175 103 226 119 128 227 218
###IMVC based hypothesis test###
n=100
x=rnorm(n)
y=2*x+rt(n,2)
IMVCT(x,y,K=5,type = "linear")
#> [1] 1.506868e-16
y=2*cos(x)+rt(n,2)
IMVCT(x,y,K=5,type = "nonlinear",num_per = 100)
#> [1] 0
###IMVC based FDR control###
n=200
p=100
pho1=0.5
mean_x=rep(0,p)
sigma_x=matrix(NA,nrow = p,ncol = p)
for (i in 1:p) {
  for (j in 1:p) {
    sigma_x[i,j]=pho1^(abs(i-j))
  }
}
x=rmvnorm(n, mean = mean_x, sigma = sigma_x,method = "chol")
x1=x[,1]
x2=x[,2]
x3=x[,3]
x4=x[,4]
x5=x[,5]
y=x1+x2+x3+x4+x5+rnorm(n)
IMVCFDR(y,x,K=5,numboot=100,timeboot=50,true_signal=c(1,2,3,4,5),null_method="hist",alpha=0.2)
#> $selected
#> [1] 3 5 4 2 6
#> 
#> $FDR
#> [1] 0.2
#> 
#> $Power
#> [1] 0.8

References

Wei Xiong, Han Pan, Hengjian Cui. (2024) “A Robust Integrated Mean Variance Correlation and Its Use in High Dimensional Data Analysis.”

mirror server hosted at Truenetwork, Russian Federation.