mdd_test_perm: Conditional mean independence test using permutation

In jeffjyan/MDD: Conditional Mean Independence Testing via Martingale Difference Divergence

Description

Conditional mean independence test using permutation

Usage

mdd_test_perm(X, Y, R = 999)

Arguments

X	distances or data of first sample
Y	distances or data of second sample
R	number of replicates

Value

mdd_test_perm returns a list containing

• test_statistic: observed value of the test statistic

• MDD: sample martingale difference divergence

• replicates: replicates of the test statistic

• p_value: approximate p-value of the test

References

Shao, Xiaofeng, and Jingsi Zhang. "Martingale difference correlation and its use in high-dimensional variable screening." Journal of the American Statistical Association 109.507 (2014): 1302-1318.

Park, Trevor, Xiaofeng Shao, and Shun Yao. "Partial martingale difference correlation." Electronic Journal of Statistics 9.1 (2015): 1492-1517.

Examples

A <- iris[1:50, 1:2]
B <- iris[1:50, 3:4]
set.seed(1)
mdd_test_perm(A, B)

jeffjyan/MDD documentation built on Dec. 9, 2019, 12:16 a.m.