Normal Q-Q Plot

Plots empirical quantiles of the studentized residuals from a linear model against the theoretical quantiles of a the normal distribution.

qq_plot(x, reps = 100, conf = 0.95, n.labels = 3, alpha = 0.4)

Arguments

x

an object of class "lm".

reps

integer; number of bootstrap replications for the confidence envelope (default=100)

conf

numeric; size of confidence interval (default=0.95).

n.labels

integer; the number of largest residuals to label ( default = 3).

alpha

numeric; transparency for plotted points (0 to 1, default=0.4).

Value

a ggplot2 graph

Details

The function is a modification of the qqPlot function in the car package, using ggplot2 rather than base graphics. A robust linear regression line is provided by the rlm function in the MASS package.

See also

qqPlot, qqnorm, rlm

Examples

mtcars$am <- factor(mtcars$am)
fit <- lm(mpg ~ wt + am + disp + hp, mtcars)
qq_plot(fit)