A basic approach is
x <- rchisq(1000, 4)
qqplot(qchisq(ppoints(x), 4), x)
abline(0,1)
Substitute whatever distribution and parameters that might apply. Notice that
if you don't have a location-scale family of distributions, you need to compare
to the identity line, not just look for linearity.
Is there any R package that can plot Q-Q plot for loglogistic,lognormal and
2 parameter exponential distributions or is there a way to check the model
fit for these distributions in R ?
[[alternative HTML version deleted]]
__
R-help@r-project.or
2 matches
Mail list logo
