| Type: | Package |
| Title: | Test for Equality of Covariance Matrices |
| Version: | 0.1.0 |
| Description: | Computes p-values using the largest root test using an approximation to the null distribution by Johnstone (2008) <doi:10.1214/08-AOS605>. |
| Depends: | R (≥ 3.0.0) |
| Imports: | RMTstat, stats, corpcor |
| License: | MIT + file LICENSE |
| LazyData: | true |
| URL: | http://github.com/turgeonmaxime/covequal |
| BugReports: | http://github.com/turgeonmaxime/covequal/issues |
| Suggests: | testthat, covr |
| RoxygenNote: | 6.0.1 |
| NeedsCompilation: | no |
| Packaged: | 2017-10-12 12:57:44 UTC; mturgeon |
| Author: | Maxime Turgeon [aut, cre] |
| Maintainer: | Maxime Turgeon <maxime.turgeon@mail.mcgill.ca> |
| Repository: | CRAN |
| Date/Publication: | 2017-10-14 13:11:11 UTC |
Test for equality of covariance matrices
Description
Uses Roy's union-intersection principle for testing for equality of covariance matrices between two samples. Also provides p-values.
Usage
test_covequal(X, Y, inference = c("TW", "permutation"), nperm)
Arguments
X |
matrix of size n1 x p |
Y |
matrix of size n2 x p |
inference |
Method for computing p-value. |
nperm |
Number of permutations. See details. |
Value
A list containing the test statistic and the p-value.
Examples
X <- matrix(rnorm(50*100), ncol = 100)
Y <- matrix(rnorm(40*100), ncol = 100)
test_covequal(X, Y, inference = "TW", nperm = 10)