Bootstrap 2-sample mean test for (hyper-)spherical data
Bootstrap 2-sample mean test for (hyper-)spherical data.
hcf.boot(x1, x2, fc = TRUE, B = 999) lr.boot(x1, x2, B = 999) hclr.boot(x1, x2, B = 999) embed.boot(x1, x2, B = 999) het.boot(x1, x2, B = 999)
x1 |
A matrix with the data in Euclidean coordinates, i.e. unit vectors. |
x2 |
A matrix with the data in Euclidean coordinates, i.e. unit vectors. |
fc |
A boolean that indicates whether a corrected F test should be used or not. |
B |
The number of bootstraps to perform. |
The high concentration (hcf.boot), log-likelihood ratio (lr.boot), high concentration log-likelihood ratio (hclr.boot), embedding approach (embed.boot) or the non equal concentration parameters approach (het.boot) is used.
A vector including two or three numbers, the test statistic value, the bootstrap p-value of the test and the common concentration parameter kappa based on all the data.
Michail Tsagris.
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr.
Mardia, K. V. and Jupp, P. E. (2000). Directional statistics. Chicester: John Wiley & Sons.
Rumcheva P. and Presnell B. (2017). An improved test of equality of mean directions for the Langevin-von Mises-Fisher distribution. Australian & New Zealand Journal of Statistics, 59(1), 119-135.
x <- rvmf(60, rnorm(3), 15) ina <- rep(1:2, each = 30) x1 <- x[ina == 1, ] x2 <- x[ina == 2, ] hcf.boot(x1, x2) lr.boot(x1, x2) het.boot(x1, x2)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.