varCheck {glmxdiag} | R Documentation |
According to GLM theory, the true variance function can be estimated as (y-mu)^2/phi. Comparing this quantity with the variance function assumed by the model suggests if the chosen one is appropriate. If the model is correct, points should be arranged near the bisector; otherwise this plot suggests the direction of the error made by chosing the variance function.
varCheck(model, xlab, ylab, pch, col, lcol, ...)
model |
a model supported by |
xlab |
title for the x axis. |
ylab |
title for the y axis. |
pch |
type of points. |
col |
color of the points. |
lcol |
color of the bisector. |
... |
further arguments passed to |
Doesn't return a value, called for side effects.
Giuseppe Reale
data(stopping, package = 'glmxdiag')
mod <- glm(Distance ~ Speed, family = Gamma(link = 'sqrt'), data = stopping)
varCheck(mod)