linear regression - Trying to get confidence/prediction intervals with `predict.lm` in R, but I keep getting an error regarding my dichotomous variable -
i have dataset looks this: time size type 1 22 151 0 2 31 92 0 3 26 175 0 4 35 31 0 5 27 104 0 6 5 277 0 7 17 210 0 8 24 120 0 9 9 290 0 10 21 238 0 11 33 164 1 12 20 272 1 13 16 295 1 14 43 68 1 15 36 85 1 16 26 224 1 17 25 166 1 18 18 305 1 19 35 124 1 20 19 246 1 my aim simple: run regression , confidence/prediction intervals. i run regression so: fit.lm1<-lm(time~size+type,data=project3) i want 95% confidence intervals , 95% prediction intervals mean time when size equal 200. want cis/pis type = 0 , type = 1. code is: new_val <- data.frame(size= c(200,200),type=c(1,0)) ci<-predict(fit.lm1,newdata=new_val,interval="confidence") pi<-predict(fit.lm1,newdata=new_val,interval="prediction") i following errors: error: variable 'type' fitted type "factor" type "numeric"