Displaying 1 result from an estimated 1 matches for "decupl".
Did you mean:
dedupl
2006 Feb 27
1
Different deviance residuals in a (similar?!?) glm example
...19, 338,
246, 205, 148, 112, 75, 52, 32, 23, 15, 7, 6, 3) # just
as example
deaths.sim <- rpois(length(age), exposures*hazard.sim) # simulating
deaths from a poisson distribution (Brillinger, 1986)
my.offset <- log(exposures) # offset for the poisson regression
# new dataset: decupleing the sample size
deaths.sim1 <- deaths.sim * 10
exposures1 <- exposures * 10
my.offset1 <- log(exposures1)
# fitting the first dataset
fit <- glm(formula = deaths.sim ~ age + offset(my.offset),
family = poisson(link = "log"))
res <- residuals(fit, type=&quo...