inst/example/course/LongitudinalLogisticExampleCode.R

In ocrug/AdvancedRegression: Advaced Regression Models With SAS And R

dosages.data<- read.csv(file="./LongitudinalLogisticExampleData.csv", 
                            header=TRUE, sep=",")

#creating longform dataset and time variable
library(reshape2)
longform<- melt(dosages.data, id.vars=c('patid', 'dosage','gender'), variable.name='weekn', value.name='effects')
week<- ifelse(longform$weekn=='week1',1,ifelse(longform$weekn=='week3',3,ifelse(longform$weekn=='week7',7,16)))

#fitting random slope and intercept logistic model
library(lme4)
summary(fitted.model<- glmer(effects~dosage+gender+week+(1+(week|patid)), data=longform,family=binomial(link='logit')))

#using the fitted model for prediction
print(predict(fitted.model, data.frame(patid=29, dosage='A', gender='F',week=7),re.form=NA, type='response'))