inst/exercise/course/LongitudinalNormalExerciseCode.R

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

fit.data<- read.csv(file='./LongitudinalNormalExerciseData.csv',
                    header=TRUE, sep=',')
library(reshape2)

#creating longform dataset
data1<- melt(fit.data[,c('id','gender','age','oxygen1','oxygen2','oxygen3')],
             id.vars=c('id','gender','age'),variable.name='oxygenlevel',value.name='oxygen')
data2<- melt(fit.data[,c('runtime1','runtime2','runtime3')],variable.name='runtimelevel',value.name='runtime')
data3<- melt(fit.data[,c('pulse1','pulse2','pulse3')],variable.name='pulselevel',value.name='pulse')
data4<- cbind(data1,data2)
longform.data<- cbind(data4,data3)

condition<- ifelse(longform.data$oxygenlevel=='oxygen1',1,ifelse(longform.data$oxygenlevel=='oxygen2',2,3))

#plotting histogram with normal density curve
library(rcompanion)

plotNormalHistogram(longform.data$pulse)

#fitting random slope and intercept model
library(nlme)

summary(fitted.model<- lme(pulse~gender+age+oxygen+runtime+condition, random=~1+condition|id,
control=lmeControl(opt='optim'), data=longform.data))

#prediction
print(predict(fitted.model, data.frame(id=21, gender='F', age=36, condition=1, oxygen=40.2, 
runtime=10.3),level=0))