VIF computes multicolinearity-related diagnostic statistics for a linear model with an absolute term.

VIF(m)

Arguments

m

linear model fitted by 'lm'

Details

The implementation was taken from Zvara2008.

Value

a matrix. In rows: diagnostic statistics related to individual regressors in the original model. In columns: regression coefficient estimate in a standardized model, variance inflation factor, the coefficient of determination R^2 in model explaining the current regressor using all the others, and tolerance (1-R^2 = 1/VIF).

References

Karel Zvara: Regrese, Matfyzpress Praha 2008

See also

Examples

require(stats)
#> Loading required package: stats
require(datasets)
#> Loading required package: datasets
m<-lm(Sepal.Width~Sepal.Length+Petal.Width+Petal.Length,iris) VIF(m)
#> b.star VIF R2 tol #> Sepal.Length 1.1533143 4.278282 0.7662613 0.23373870 #> Petal.Width 0.9758767 14.089441 0.9290249 0.07097514 #> Petal.Length -2.3734841 19.426391 0.9485236 0.05147637