WZ2026.GLHTBF.2cNRT(), an F-type normal-reference
test for heteroscedastic high-dimensional general linear hypothesis
testing (GLHT) problems.WZ2026.GLHTBF.2cNRT() supports the grouped-data GLHTBF
interface WZ2026.GLHTBF.2cNRT(Y, G, n, p), where
Y is a list of groupwise data matrices, G is
the raw full-row-rank contrast matrix, n is the vector of
group sample sizes, and p is the common data
dimension.NRtest object reports the test statistic,
p-value, fitted numerator and denominator degrees of freedom, and the
corresponding approximation method.CCXH2024.GLHTBF.2cNRT() for the rank-one
scale-invariant GLHTBF normal-reference procedure of Cao et
al. (2024).LHNB2025.GLHTBF.NABT() for the rank-one
random-integration GLHTBF normal-approximation procedure of Li et
al. (2025).G is the raw contrast matrix supplied by the user;
the normalized contrast matrix and the induced contrast operator are
constructed internally.WZ2026.GLHTBF.2cNRT() for omnibus and rank-one GLHTBF
analyses using grouped data.TNEW so
that its callable routine is consistently recorded as
WZ2026.GLHTBF.2cNRT().Y, contrast matrix
G, sample-size vector n, and common dimension
p.BS1996.TS.NART to
BS1996.TS.NABT for consistency.save-always with
actions/cache@v3 in GitHub Actions workflow.The function name has been changed. The title, example, output format of the function have been changed, and some typos have also been corrected.
Added a helper function to ‘HDNRA.cpp’ file in order to improve computation speed; updated the code of all functions to facilitate faster computation.
Added an ‘NRtest.object’ to output an S3 class ‘NRtest’ for our package, and also constructed a corresponding print function to output the appropriate format.
Added a ‘zzz.R’ file to manage package startup messages and initialization.