mc_adjust: Multi-Collinearity Adjustment

In bradleyboehmke/anomalyDetection: Implementation of Augmented Network Log Anomaly Detection Procedures

Description

mc_adjust handles issues with multi-collinearity.

Usage

mc_adjust(data, min_var = 0.1, max_cor = 0.9, action = "exclude")

Arguments

data	named numeric data object (either data frame or matrix)
min_var	numeric value between 0-1 for the minimum acceptable variance (default = 0.1)
max_cor	numeric value between 0-1 for the maximum acceptable correlation (default = 0.9)
action	select action for handling columns causing multi-collinearity issues exclude: exclude all columns causing multi-collinearity issues (default) select: identify the columns causing multi-collinearity issues and allow the user to interactively select those columns to remove

Details

mc_adjust handles issues with multi-collinearity by first removing any columns whose variance is close to or less than min_var. Then, it removes linearly dependent columns. Finally, it removes any columns that have a high absolute correlation value equal to or greater than max_cor.

Value

mc_adjust returns the numeric data object supplied minus variables violating the minimum acceptable variance (min_var) and the maximum acceptable correlation (max_cor) levels.

Examples

## Not run: 
x <- matrix(runif(100), ncol = 10)
x %>%
  mc_adjust()

x %>%
  mc_adjust(min_var = .15, max_cor = .75, action = "select")

## End(Not run)

bradleyboehmke/anomalyDetection documentation built on Oct. 15, 2019, 12:04 a.m.