fill values based on value for matched IDs

March 23, 2022

I am looking for a data.table solution to a simply problem. I have data like so

library(data.table)
data <- data.table(
  id = seq(1:20),
  match_id = c( rep(1,5), rep(2,5), rep(3,5), rep(4,5)  ),
  exp = c( 1, rep(0,4), 1 , rep(0,4), 1 , rep(0,4), 1 , rep(0,4) )
)
data[, var := ifelse(exp==1, runif(1,min=0,max=100), 0), by = id]

I want, for each unique match_id where exp==0 to fill var based on the value of var for exp==1. Data to end up with:

data_want <- data.table(
  id = seq(1:20),
  match_id = c( rep(1,5), rep(2,5), rep(3,5), rep(4,5)  ),
  exp = c( 1, rep(0,4), 1 , rep(0,4), 1 , rep(0,4), 1 , rep(0,4) ),
  var = c(rep(data[1]$var,5), rep(data[6]$var,5), rep(data[11]$var,5), rep(data[16]$var,5))
)

>Solution :

data[, var:=var[which(exp==1)], by=match_id]

       id match_id   exp       var
    <int>    <num> <num>     <num>
 1:     1        1     1 75.342705
 2:     2        1     0 75.342705
 3:     3        1     0 75.342705
 4:     4        1     0 75.342705
 5:     5        1     0 75.342705
 6:     6        2     1 81.822968
 7:     7        2     0 81.822968
 8:     8        2     0 81.822968
 9:     9        2     0 81.822968
10:    10        2     0 81.822968
11:    11        3     1  3.309884
12:    12        3     0  3.309884
13:    13        3     0  3.309884
14:    14        3     0  3.309884
15:    15        3     0  3.309884
16:    16        4     1  2.047301
17:    17        4     0  2.047301
18:    18        4     0  2.047301
19:    19        4     0  2.047301
20:    20        4     0  2.047301
       id match_id   exp       var