proc_genmod: Emulator of SAS proc_genmod output

In collinn/sassyR: A sassy way to use R

Description

This function emulates some of the output returned by proc_genmod in SAS. It is simply a wrapper for a handful of diagnostics, organized in a slightly convenient way. This function does not fit any actual models, and thus must be passed a fitted object of class 'glm'

Usage

proc_genmod(fit, scale = "pearson")

Arguments

fit	A fitting 'glm' or 'glm.nb' object
scale	Matching SAS argument, uses either pearson or deviance for estimation of dispersion parameter, takes either "pearson" or "deviance"

Examples

library(MASS)

## Examples match course notes (4.5-4.7)
data(ACHD)

fit1 <- glm(Mort ~ Age_Gp, family = poisson, offset = Log_Pop, data = ACHD)
fit2 <- glm.nb(Mort ~ Age_Gp + offset(Log_Pop), data = ACHD)
fit3 <- glm(cbind(Mort, Pop-Mort) ~ Age_Gp,
            family = binomial(link = "log"), data = ACHD)
fit4 <- glm(cbind(Mort, Pop-Mort) ~ Age_Gp,
            family = binomial(link = "logit"), data = ACHD)
fit5 <- glm(Log_Mort ~ Age_Gp, family = gaussian,
            offset = Log_Pop, data = ACHD)

proc_genmod(fit1)
proc_genmod(fit2)
proc_genmod(fit3)
proc_genmod(fit4)
proc_genmod(fit5)

collinn/sassyR documentation built on May 1, 2020, 1:59 a.m.