fabric {BayesDA} | R Documentation |
Numbers of faults found in each of 32 rolls of fabric produced in a particular factory. Also given is the length of the roll.
data(fabric)
A data frame with 32 observations on the following 2 variables.
length
length of roll
faults
number of faults in roll
The book uses this for exercise 5. page 441
data(fabric) str(fabric) names(fabric) # Identity link: with(fabric, plot(faults ~ length)) # log link: with(fabric, plot(faults ~ length, log="y")) # Fitting poisson regression models: mod1 <- glm(faults ~ length-1, data=fabric, family=poisson) OK <- require(MCMCpack) if(OK) mod2 <- MCMCpoisson(faults ~ length-1, data=fabric, b0=0, B0=0.0001) summary(mod1) confint(mod1) if(OK) summary(mod2) # The exercise is to investigate overdispersion ...