Replace duplicate elements in dplyr pipe

Question

I want to replace duplicated elements within a group

df <- data.frame(A=c("a", "a", "a", "b", "b", "c"), group = c(1, 1, 2, 2, 2, 3))

I want to keep the first element of the group, while replacing anything else with NA. Something like:

df <- df %>% 
     group_by(group) %>%
     mutate(B = first(A))

Which doesn't produce what I want. What I want instead is B <- c(a, NA, a, NA, NA, c)

Answer 1

Use replace with duplicated:

df %>% group_by(group) %>% mutate(B = replace(A, duplicated(A), NA))

# A tibble: 6 x 2
# Groups:   group [3]
#       A group
#  <fctr> <dbl>
#1      a     1
#2     NA     1
#3      a     2
#4      b     2
#5     NA     2
#6      c     3

Or if keep only the first element:

df %>% 
    group_by(group) %>% 
    mutate(B = ifelse(row_number() == 1, as.character(A), NA))

# A tibble: 6 x 2
# Groups:   group [3]
#      A group
#  <chr> <dbl>
#1     a     1
#2  <NA>     1
#3     a     2
#4  <NA>     2
#5  <NA>     2
#6     c     3

OR use replace:

df %>% 
    group_by(group) %>% 
    mutate(B = replace(A, row_number() > 1, NA))

# A tibble: 6 x 2
# Groups:   group [3]
#       A group
#  <fctr> <dbl>
#1      a     1
#2     NA     1
#3      a     2
#4     NA     2
#5     NA     2
#6      c     3

Answer 2

In data.table you could do:

library(data.table)
setDT(df)[, B := c(A[1], rep(NA, .N - 1)), by = group]

Or same logic in dplyr:

library(dplyr)
df %>% group_by(group) %>% mutate(B = c(as.character(A[1]), rep(NA, n() - 1)))

# A tibble: 6 x 3
# Groups:   group [3]
#       A group     B
#  <fctr> <dbl> <chr>
#1      a     1     a
#2      a     1  <NA>
#3      a     2     a
#4      b     2  <NA>
#5      b     2  <NA>
#6      c     3     c