How to reset a loop?

Question

Hi everyone I have this dataset

library(dplyr)
library(tidyr)
input<- frame_data(
  ~member_id,   ~fill_date   ,  ~drug, ~days_supply, 
  "603",       "02/17/2005",    "a",    30,         
  "603",       "06/13/2005",    "a",    30,       
  "603",       "08/11/2005",    "a",    30,           
  "603",       "06/12/2006",    "b",    15,
  "603",       "05/09/2006",    "b",    30
  )

I am trying to create a variable called "time" which indicates the number of time the variable "drug" appears. So the output should look like this

output<- frame_data(
  ~member_id,   ~fill_date   ,  ~drug, ~days_supply, ~time,
  "603",       "02/17/2005",    "a",    30,           1,
  "603",       "06/13/2005",    "a",    30,           2,
  "603",       "08/11/2005",    "a",    30,           3,
  "603",       "06/12/2006",    "b",    15,           1,
  "603",       "05/09/2006",    "b",    30            2
  )

in other terms I'am looking for a sort of loop that can reset every time the "drug" variable changes I've tried this code

time<-1
i<-2
j<-1
while (i <=nrow(input)){
  if (input[i,3,drop=]==input[i-1,3,drop=]){
    j<-i
    time<-c(time,j)
  }else{
    j<-1
    time<-c(time,j)
  }
  i<-i+1
}

but of course it dosen't work since the i could not be reset because it indicates the row index while verifying the condition in the same times.

Thank you for your help

Answer 1

You're using dplyr, so use group_by, not a loop.

input %>% group_by(drug) %>% mutate(time = 1:n())

Probably you want to add member_id to the group by as well, but since you don't mention I don't include. If so, just group_by(drug, member_id) instead of group_by(drug).