unexpected symbol in "tt" function for categorical variable

Asked Mar 09 '21 at 19:14

Active Mar 09 '21 at 19:24

Viewed 146 times

I have a dataset with time, event and 4 predictor variables of which "Q.V4" is a categorical variable which needs to be transformed using tt function. This variable has 4 levels - Q1, Q2, Q3, Q4

F1 <- data.frame(
  time=c(102,48,84,72,66,90), 
  event=c(0,1,0,0,0,0), 
  v1=c(72,76,74,76,72,61),
  v2=c(1,1,1,1,1,1), v3=c(1,1,2,2,1,2), 
  Q.V4=c("Q2","Q4","Q3","Q1","Q2","Q1"))

coxph(Surv(time,event)~v1 + v2 + v3 + tt(Q.V4), data = F1, 
 tt = function(Q.V4,time,...){matrix <- model.matrix(~Q.V4)[,-1] matrix*time)})

I am getting the following error message

unexpected symbol in "function(Q.V4,time,...){ matrix <- model.matrix(~Q.V4)[,-1]"

I would really appreciate if you can please show me where I am making the mistake and help me with the right code at your earliest. Thanks in advance.

edited Mar 09 '21 at 19:24

asked Mar 09 '21 at 19:14

Biostats

1

you're missing something between `[,-1]` and `matrix*time`. Looks like a linebreak? – SmokeyShakers Mar 09 '21 at 19:19
Thanks for pointing out. I removed the gap. It is the error message that I am getting. I don't know why...please help me...thanks – Biostats Mar 09 '21 at 19:24
If you want two statements on the same line, you need a ";" in between them. Try: `coxph(Surv(time,event)~v1 + v2 + v3 + tt(Q.V4), data = F1, tt = function(Q.V4,time,...){matrix <- model.matrix(~Q.V4)[,-1];matrix*time}) ` – MrFlick Mar 09 '21 at 19:24
Hi @MrFlick, your method is not working....Please help – Biostats Mar 09 '21 at 19:35
MrFlick's suggestion works for me. Notice he also removed the `)` before the `}`, which shouldn't be there – Hobo Mar 09 '21 at 19:42
1

Oh yes! Now I got it...really appreciate your help @MrFlick.... – Biostats Mar 09 '21 at 19:53
Thanks to you as well @Hobo – Biostats Mar 09 '21 at 19:53

unexpected symbol in "tt" function for categorical variable

0 Answers0