glmxdiag (version 1.0.0)

leverage: Visualize Leverages

Description

Graphical visualization of leverages for each observation within input model. Corresponds to the main diagonal of the hat matrix, i.e H[i,i].

Usage

leverage(object, label.id, n.label.id, xlab, ylab, pos,
         hline, lcol, lwd, lty, ...)

Arguments

object

object of class 'influence', see Details.

label.id

labels of observations, should be a vector with n elements.

n.label.id

number of observations with highest leverage to label in the plot.

xlab

title for the x axis.

ylab

title for the y axis.

pos

position of observations labels. Values 1, 2, 3, 4 respectively indicate below, left, above, right.

hline

numeric, where to position the horizontal line; useful to individuate points that exceed a specific treshold. Defaults to 2*p/n.

lcol

color of line.

lwd

width of line.

lty

type of line.

...

further arguments passed to plot

Value

Doesn't return a value, called for side effects.

Details

This function only works with objects given as output from influenceDiag.

A rule of thumbs says that an observations has high leverage if H[i,i] > 2*p/n where p is the number of coefficients and n the sample size.

Examples

# NOT RUN {
data("mtcars")
mod <- glm(mpg ~ cyl + hp + carb, family = Gamma, data = mtcars)
inf <- influenceDiag(mod)
leverage(inf, label.id = rownames(mtcars))
# }