README.md

In leaffur/mcvis: Multi-Collinearity Visualization

mcvis: visualisation of multicollinearity in data

Introduction

mcvis is a R package for visualising multicollinearity in a data design matrix. The underlying methodology uses resampling techniques to identify groups of variables that causes multicollinearity.

You can learn more about mcvis from this vignette.

Installation

mcvis can be installed using the devtools package.

devtools::install_github("kevinwang09/mcvis")

A quick example

Using a mcvis bipartite plot, variables (bottom row) that cause strong collinearity are visualised as bolded lines connecting with our “tau” statistics (top row).

library(mcvis)
library(ggplot2)

set.seed(1)
p = 10
n = 100

X = matrix(rnorm(n*p), ncol = p)
## Inducing collinearity into the design matrix
X[,1] = X[,2] + rnorm(n, 0, 0.1) 


mcvis_result = mcvis(X)
plot(mcvis_result)

Reference

• Lin, C., Wang, K. Y. X., & Mueller, S. (2021). mcvis: A new framework for collinearity discovery, diagnostic and visualization. Journal of Computational and Graphical Statistics, 30:1, 125-132.

leaffur/mcvis documentation built on Aug. 28, 2023, 9:54 a.m.