pearson.pvalue: fast p-value approximation

In multivariance: Measuring Multivariate Dependence Using Distance Multivariance

Description

Computes the p-value of a sample using Pearson's approximation of Gaussian quadratic forms with the estimators developed by Berschneider and Böttcher in [4].

Usage

pearson.pvalue(x, vec = NA, type = "multi", ...)

Arguments

x	matrix, the rows should be iid samples
vec	vector, which indicates which columns of x are treated together as one sample. The default case treats each column as a separate sample.
type	one of "multi","total","m.multi.2","m.multi.3","all"
...	these are passed to cdms

Details

This is the method recommended in [4], i.e., using Pearson's quadratic form estimate with the unbiased finite sample estimators for the mean and variance of normalized multivariance together with the unbiased estimator for the limit skewness.

References

For the theoretic background see the reference [4] given on the main help page of this package: multivariance-package.

multivariance documentation built on Oct. 6, 2021, 5:08 p.m.