ARHT: Adaptable Regularized Hotelling's T^2 Test for High-Dimensional
Data
Perform the Adaptable Regularized Hotelling's T^2 test (ARHT) proposed by
Li et al., (2016) <doi:10.48550/arXiv.1609.08725>. Both one-sample and two-sample mean test are available with
various probabilistic alternative prior models. It contains a function to consistently
estimate higher order moments of the population covariance spectral distribution using
the spectral of the sample covariance matrix (Bai et al. (2010) <doi:10.1111/j.1467-842X.2010.00590.x>).
In addition, it contains a function to sample from 3-variate chi-squared random vectors approximately
with a given correlation matrix when the degrees of freedom are large.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=ARHT
to link to this page.