Returns a leverage measure for each sample point in the data.
Usage
leverage(object, ...)
## Default S3 method:
leverage(object, ...)
## S3 method for class 'lm'
leverage(object, ...)
Arguments
object
a fitted model object.
...
additional arguments are passed to the dispatched method.
Details
For least-squares linear models the leverages are the diagonal elements of the hat matrix. This function is generic. The default method assigns a leverage of p/n to every sample point.
Value
a numeric vector containing the leverages.
Examples
## Example from lm:
ctl <- c(4.17,5.58,5.18,6.11,4.50,4.61,5.17,4.53,5.33,5.14)
trt <- c(4.81,4.17,4.41,3.59,5.87,3.83,6.03,4.89,4.32,4.69)
group <- gl(2,10,20, labels=c("Ctl","Trt"))
weight <- c(ctl, trt)
lm.D9 <- lm(weight ~ group)
leverage(lm.D9)