pbc3 <- read.csv2("http://staff.pubhealth.ku.dk/~linearpredictors/datafiles/Pbc3.csv", sep = ";", dec = ".", header=TRUE, colClasses = c("factor","factor","factor","factor","numeric", "factor","factor","numeric","numeric", "numeric","numeric","numeric","numeric","numeric","factor"), na.strings="." ) ## followup in years pbc3$followup <- pbc3$days/365.25 ## quintiles of bilirubin ## would we like to set the 1st quintile to 0? quintiles <- quantile(pbc3$bili,seq(0,1,0.2)) ## quintile groups of bilirubin pbc3$biligroup <- cut(pbc3$bili, quintiles, include.lowest=TRUE, # include also lowest observed value labels = c(1:5) # default labelling uses intervals )