| Type: | Package |
| Title: | Multicovariance and Multicorrelation for p-Variables |
| Version: | 0.1.0 |
| Description: | Implements the multicorrelation coefficient for p-variables as described in Cankaya (2023). The package provides a numerically stable implementation using logarithmic transformations and a log-sum-exp approach to reduce numerical overflow and underflow when calculations involve a large number of variables. |
| License: | GPL-3 |
| Encoding: | UTF-8 |
| Config/roxygen2/version: | 8.1.0 |
| NeedsCompilation: | no |
| Packaged: | 2026-09-07 08:16:08 UTC; mehmet.cankaya |
| Author: | Mehmet Niyazi Cankaya [aut, cre] |
| Maintainer: | Mehmet Niyazi Cankaya <mehmet.cankaya@usak.edu.tr> |
| Repository: | CRAN |
| Date/Publication: | 2026-09-15 11:50:08 UTC |
Calculate Multicorrelation for p-Variables
Description
Computes the multicorrelation coefficient for p variables.
Usage
multicorrelation(x, r = 1)
Arguments
x |
A numeric matrix or data frame. |
r |
A positive numeric power parameter. Default is 1. |
Details
Each variable is centered by its arithmetic mean and scaled by its mean absolute deviation from the mean. Logarithmic transformations and a log-sum-exp approach are used to improve numerical stability.
Value
A numeric multicorrelation coefficient.
References
Cankaya, M. N. (2023). Multicovariance and Multicorrelation for p-variables. In Mathematical Methods for Engineering Applications (ICMASE 2022), pp. 273-284. Cham: Springer International Publishing.
Examples
set.seed(123)
X <- matrix(rnorm(1000), nrow = 100, ncol = 10)
multicorrelation(X)
multicorrelation(X, r = 2)