I have a dataset like the below:

df <- 
  read.table(textConnection("CustomerID   Code1 Code2   
A 1 1 
A 0 1 
A 1 1 
B 1 0 
B 0 1 
C 1 1 
C 1 1 
D 1 0                         
D 0 1 
D 1 1 "), header=TRUE)

What I am looking to do is loop through each CustomerID and, if the first row of the group has Code1 = 1 and Code2 = 1, change the Code2 value to 0 for that row. So in the above case, below would be the correct final dataset - the change only applies to the very first row in each customer group, if it has a 1 for both Code1 and Code2.

Any help very much appreciated, thank you.

CodePudding user response：

Here's a solution using dplyr::case_when(). Group by CustomerID, create an index within each group using row_number(), and apply the case_when logic to the first row.

library(dplyr)
#> Warning: package 'dplyr' was built under R version 4.2.1
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

df <-
  read.table(textConnection("CustomerID   Code1 Code2
A 1 1
A 0 1
A 1 1
B 1 0
B 0 1
C 1 1
C 1 1
D 1 0
D 0 1
D 1 1 "),
header = TRUE)

df |>
  group_by(CustomerID) |>
  mutate(
    row = row_number(),
    Code2 = case_when(row == 1 &
                        Code1 == 1 & Code2 == 1 ~ as.integer(0),
                      TRUE ~ Code2)
  ) |>
  ungroup() |>
  select(-row)
#> # A tibble: 10 × 3
#>    CustomerID Code1 Code2
#>    <chr>      <int> <int>
#>  1 A              1     0
#>  2 A              0     1
#>  3 A              1     1
#>  4 B              1     0
#>  5 B              0     1
#>  6 C              1     0
#>  7 C              1     1
#>  8 D              1     0
#>  9 D              0     1
#> 10 D              1     1

^{Created on 2022-10-26 with reprex v2.0.2}