rgr: gx.lm.vif – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

gx.lm.vif

Estimate Variance Inflation Factosr (VIFs)

Description

Function estimates Variance Inflation Factors (VIFs), measures of collinearity in a linear model. The VIF provides a measure of how much the variance of an estimated regression coefficient is increased because of collinearity. Collinearity is present when there is a high correlation between the independent, predictor, variables in a model, i.e. they tell the same ‘story’. Where collinearity exists it is often best to remove predictor variables with high VIFs from the model.

Usage

gx.lm.vif(object, ...)

Arguments

`object`	a `lm` object.
`...`	any additional parameters.

Value

A (structure) table of Variable Inflation Factors for the predictor variables.

Note

VIFs >5 are indicative of collinearity, and the information conveyed in that variable is also in the subset of the remaining variables.

Author(s)

W.N. Venables,function shared on S-News, October 21, 2002

References

http://www.biostat.wustl.edu/archives/html/s-news/2001-10/msg00164.html

Examples

## Make test data available
data(sind)
attach(sind)

## Model 1
sind.1 <- lm(log(Zn) ~ Fe + log(Mn) + log(Cu) + log(Cd))
summary(sind.1)
gx.lm.vif(sind.1)

## Model 2
sind.2 <- lm(log(Zn) ~ Fe + log(Mn))
summary(sind.2)
gx.lm.vif(sind.2)
AIC(sind.1, sind.2)

## Model 3
sind.3 <- lm(log(Zn) ~ log(Mn) + log(Cu))
summary(sind.3)
gx.lm.vif(sind.3)
AIC(sind.1, sind.2, sind.3) 

## Clean-up and detach test data
rm(sind.1)
rm(sind.2)
rm(sind.3)
detach(sind)

rgr

Applied Geochemistry EDA

v1.1.15

GPL-2

Authors

Robert G. Garrett

Initial release

2018-03-05