| Type: | Package |
| Date: | 2021-09-07 |
| Title: | Robust Test for Complete Independence in High-Dimensions |
| Version: | 1.3 |
| Author: | Hasan BULUT [aut,cre] |
| Maintainer: | Hasan BULUT <hasan.bulut@omu.edu.tr> |
| Description: | Test Statistics for Independence in High-Dimensional Datasets. This package consists of two functions to perform the complete independence test based on test statistics proposed by Bulut (unpublished yet) and suggested by Najarzadeh (2021) <doi:10.1080/03610926.2019.1702699>. The Bulut's statistic is not sensitive to outliers in high-dimensional data, unlike one of Najarzadeh (2021) <doi:10.1080/03610926.2019.1702699>. So, the Bulut's statistic can be performed robustly by using RDnp function. |
| License: | GPL-2 |
| Depends: | R (≥ 4.0) |
| Imports: | cellWise, MASS |
| Encoding: | UTF-8 |
| RoxygenNote: | 7.1.1 |
| NeedsCompilation: | no |
| Packaged: | 2021-09-07 21:46:55 UTC; user |
| Repository: | CRAN |
| Date/Publication: | 2021-09-09 09:00:16 UTC |
Test for Independence in High-Dimensional Data
Description
A Test Statistic for Independence in High-Dimensional Datasets
Usage
Dnp_Test(X)
Arguments
X |
the data. It must be matrix. |
Details
Dnp_Test function tests the complete independence in high-dimensional
data sets. This statistic was proposed by Najarzadeh (2021).
Value
a list with 2 elements:
TestValue |
The value of test statistic |
pval |
The p value |
robust |
Logical. Indicates whether the results are based on robust
statistic. Here, it returns |
Author(s)
Hasan BULUT <hasan.bulut@omu.edu.tr>
References
Najarzadeg, D (2021). Testing independece in high-dimensional multivariate normal data, Communication in Statistics: Theory and Methods. 50 (14): 3421-3435.
Examples
# Under H0
library(MASS)
data_H0<-mvrnorm(n = 20,mu = rep(0,30),Sigma = diag(30))
Dnp_Test(data_H0)
# Under H1
library(MASS)
data_H1<-mvrnorm(n = 20,mu = rep(0,30),Sigma = (diag(30)+1))
Dnp_Test(data_H1)
Robust Test for Independence in High-Dimensional Data
Description
A Robust Test Statistic for Independence in High-Dimensional Datasets
Usage
RDnp_Test(X, alpha = 0.75)
Arguments
X |
the data. It must be matrix. |
alpha |
numeric parameter. It gives the rate of uncontaminated observations. Allowed values are between 0.5 and 1 and the default is 0.75. |
Details
RDnp_Test function tests the complete independence in high-dimensional
data sets without being affected by outliers.
Value
a list with 2 elements:
TestValue |
The value of test statistic |
pval |
The p value |
robust |
Logical. Indicates whether the results are based on robust
statistic. Here, it returns |
Author(s)
Hasan BULUT <hasan.bulut@omu.edu.tr>
References
Bulut, H (Unpublished). A Robust Test Statistic for Independence in High Dimensional Data
Examples
# Under H0
library(MASS)
data_H0<-mvrnorm(n = 20,mu = rep(0,30),Sigma = diag(30))
RDnp_Test(data_H0)
# Under H1
library(MASS)
data_H1<-mvrnorm(n = 20,mu = rep(0,30),Sigma = (diag(30)+1))
RDnp_Test(data_H1)