Return the scale estimate for a regression model

This function provides a consistent method to return the estimated scale from a linear, generalized linear, nonlinear, or other model.

sigmaHat(object)

Arguments

object

A regression object of type lm, glm or nls

Details

For an lm or nls object, the returned quantity is the square root of the estimate of \(\sigma^2\). For a glm object, the returned quantity is the square root of the estimated dispersion parameter.

Value

A nonnegative number

Author

Sanford Weisberg, sandy@umn.edu

Examples

m1 <- lm(prestige ~ income + education, data=Duncan)
sigmaHat(m1)
#> [1] 13.36903