| Title: | New Graph-Based Multi-Sample Tests |
|---|---|
| Description: | New multi-sample tests for testing whether multiple samples are from the same distribution. They work well particularly for high-dimensional data. Song, H. and Chen, H. (2022) <arXiv:2205.13787>. |
| Authors: | Hoseung Song [aut, cre], Hao Chen [aut] |
| Maintainer: | Hoseung Song <[email protected]> |
| License: | GPL (>= 2) |
| Version: | 0.1.1 |
| Built: | 2024-09-15 06:27:56 UTC |
| Source: | CRAN |
This function provides graph-based multi-sample tests.
gtestsmulti(E, data_list, perm=0)gtestsmulti(E, data_list, perm=0)
E |
The edge matrix for the similarity graph. Each row contains the node indices of an edge. |
data_list |
The list of multivariate matrices corresponding to the K different classes. The length of the list is K. Each element of the list is a matrix containing observations as the rows and features as the columns. |
perm |
The number of permutations performed to calculate the p-value of the test. The default value is 0, which means the permutation is not performed and only approximated p-value based on the asymptotic theory is provided. Doing permutation could be time consuming, so be cautious if you want to set this value to be larger than 10,000. |
Returns a list teststat with each test statistic value and a list pval with p-values of the tests. See below for more details.
S |
The value of the test statistic |
S_A |
The value of the test statistic |
S_appr |
The approximated p-value of |
S_A_appr |
The approximated p-value of |
S_perm |
The permutation p-value of |
S_A_perm |
The permutation p-value of |
## Mean difference in Gaussian distribution. d = 50 mu = 0.2 sam = 50 set.seed(500) X1 = matrix(rnorm(d*sam), sam) X2 = matrix(rnorm(d*sam,mu), sam) X3 = matrix(rnorm(d*sam,2*mu), sam) data_list = list(X1, X2, X3) # We use 'mstree' in 'ade4' package to construct the minimum spanning tree. require(ade4) x = rbind(X1, X2, X3) E = mstree(dist(x)) a = gtestsmulti(E, data_list, perm = 1000) # output results based on the permutation and the asymptotic results # the test statistic values can be found in a$teststat # p-values can be found in a$pval## Mean difference in Gaussian distribution. d = 50 mu = 0.2 sam = 50 set.seed(500) X1 = matrix(rnorm(d*sam), sam) X2 = matrix(rnorm(d*sam,mu), sam) X3 = matrix(rnorm(d*sam,2*mu), sam) data_list = list(X1, X2, X3) # We use 'mstree' in 'ade4' package to construct the minimum spanning tree. require(ade4) x = rbind(X1, X2, X3) E = mstree(dist(x)) a = gtestsmulti(E, data_list, perm = 1000) # output results based on the permutation and the asymptotic results # the test statistic values can be found in a$teststat # p-values can be found in a$pval
This package can be used to determine whether multiple samples are from the same distribution.
Hoseung Song and Hao Chen
Maintainer: Hoseung Song ([email protected])
Song, H. and Chen, H. (2022). New graph-based multi-sample tests for high-dimensional and non- Euclidean data. arXiv:2205.13787
## Mean difference in Gaussian distribution. d = 50 mu = 0.2 sam = 50 set.seed(500) X1 = matrix(rnorm(d*sam), sam) X2 = matrix(rnorm(d*sam,mu), sam) X3 = matrix(rnorm(d*sam,2*mu), sam) data_list = list(X1, X2, X3) # We use 'mstree' in 'ade4' package to construct the minimum spanning tree. require(ade4) x = rbind(X1, X2, X3) E = mstree(dist(x)) a = gtestsmulti(E, data_list, perm = 1000) # output results based on the permutation and the asymptotic results # the test statistic values can be found in a$teststat # p-values can be found in a$pval## Mean difference in Gaussian distribution. d = 50 mu = 0.2 sam = 50 set.seed(500) X1 = matrix(rnorm(d*sam), sam) X2 = matrix(rnorm(d*sam,mu), sam) X3 = matrix(rnorm(d*sam,2*mu), sam) data_list = list(X1, X2, X3) # We use 'mstree' in 'ade4' package to construct the minimum spanning tree. require(ade4) x = rbind(X1, X2, X3) E = mstree(dist(x)) a = gtestsmulti(E, data_list, perm = 1000) # output results based on the permutation and the asymptotic results # the test statistic values can be found in a$teststat # p-values can be found in a$pval