CODEWITHSUNDEEP
rdplyrduplicates
Oluf Sändrostrom
Oluf Sändrostrom

Reputation: 39

How to deduplicate values within a column without group_by in R?

I have one column (a) with only two types of values: NULL and TRUE. I want to create 'a_deduplicated' column out of 'a', which is taking the first TRUE value and make the rest NULL. I do not have any other columns to group by it.

Could you please help me to write a code in R to get the 'a_deduplicated' column?

a a_deduplicated
NA NA
NA NA
NA NA
TRUE TRUE
TRUE NA
NA NA
TRUE TRUE
TRUE NA
TRUE NA
NA NA
NA NA
TRUE TRUE
TRUE NA
TRUE NA
TRUE NA

df to run in R:

df <- structure(list(a = c(NA, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, TRUE), 
                      a_deduplicated = c(NA, NA, NA, TRUE, NA, NA, TRUE, NA, NA, NA, NA, TRUE, NA, NA, NA)), class = c("tbl_df","tbl", "data.frame"), 
 row.names = c(NA, -15L))

Upvotes: 0

Views: 85

Answers (4)

Yuriy Saraykin
Yuriy Saraykin

Reputation: 8880

df <- structure(list(a = c(NA, NA, NA, TRUE, TRUE, NA, TRUE, TRUE, TRUE, NA, NA, TRUE, TRUE, TRUE, TRUE), 
                      a_deduplicated = c(NA, NA, NA, TRUE, NA, NA, TRUE, NA, NA, NA, NA, TRUE, NA, NA, NA)), class = c("tbl_df","tbl", "data.frame"), 
 row.names = c(NA, -15L))

library(tidyverse)
df %>% 
  group_by(grp = data.table::rleid(a)) %>% 
  mutate(res = if_else(row_number() == 1, a, NA), .keep = c("unused")) %>% 
  select(-grp) %>% 
  ungroup()

#> # A tibble: 15 x 3
#>      grp a_deduplicated res  
#>    <int> <lgl>          <lgl>
#>  1     1 NA             NA   
#>  2     1 NA             NA   
#>  3     1 NA             NA   
#>  4     2 TRUE           TRUE 
#>  5     2 NA             NA   
#>  6     3 NA             NA   
#>  7     4 TRUE           TRUE 
#>  8     4 NA             NA   
#>  9     4 NA             NA   
#> 10     5 NA             NA   
#> 11     5 NA             NA   
#> 12     6 TRUE           TRUE 
#> 13     6 NA             NA   
#> 14     6 NA             NA   
#> 15     6 NA             NA

library(data.table)
setDT(df)[, res := ifelse(seq(.N) == 1, a, NA), by = rleid(a)][]
#>        a a_deduplicated  res
#>  1:   NA             NA   NA
#>  2:   NA             NA   NA
#>  3:   NA             NA   NA
#>  4: TRUE           TRUE TRUE
#>  5: TRUE             NA   NA
#>  6:   NA             NA   NA
#>  7: TRUE           TRUE TRUE
#>  8: TRUE             NA   NA
#>  9: TRUE             NA   NA
#> 10:   NA             NA   NA
#> 11:   NA             NA   NA
#> 12: TRUE           TRUE TRUE
#> 13: TRUE             NA   NA
#> 14: TRUE             NA   NA
#> 15: TRUE             NA   NA

Created on 2022-08-15 by the reprex package (v2.0.1)

Upvotes: 1

maike
maike

Reputation: 485

Similar to Mael's answer, but using the lag function

require(tidyverse)

df <- df %>%  
  mutate(
    new = ifelse(a & is.na(lag(a)), TRUE, NA)
  )

df
#> # A tibble: 15 × 3
#>    a     a_deduplicated new  
#>    <lgl> <lgl>          <lgl>
#>  1 NA    NA             NA   
#>  2 NA    NA             NA   
#>  3 NA    NA             NA   
#>  4 TRUE  TRUE           TRUE 
#>  5 TRUE  NA             NA   
#>  6 NA    NA             NA   
#>  7 TRUE  TRUE           TRUE 
#>  8 TRUE  NA             NA   
#>  9 TRUE  NA             NA   
#> 10 NA    NA             NA   
#> 11 NA    NA             NA   
#> 12 TRUE  TRUE           TRUE 
#> 13 TRUE  NA             NA   
#> 14 TRUE  NA             NA   
#> 15 TRUE  NA             NA

require(waldo)
#> Lade nötiges Paket: waldo
waldo::compare(df$a_deduplicated, df$new)
#> ✔ No differences

Upvotes: 3

Ma&#235;l
Ma&#235;l

Reputation: 52329

You can do:

transform(df, a_new = ifelse(a & is.na(diff(c(0, a))), T, NA))

      a a_deduplicated a_new
1    NA             NA    NA
2    NA             NA    NA
3    NA             NA    NA
4  TRUE           TRUE  TRUE
5  TRUE             NA    NA
6    NA             NA    NA
7  TRUE           TRUE  TRUE
8  TRUE             NA    NA
9  TRUE             NA    NA
10   NA             NA    NA
11   NA             NA    NA
12 TRUE           TRUE  TRUE
13 TRUE             NA    NA
14 TRUE             NA    NA
15 TRUE             NA    NA

Upvotes: 2

akshaymoorthy
akshaymoorthy

Reputation: 346

Solution using data.table:

setDT(df)
df[, flag:=shift(a,1,type="lag")]
df[, dedup:=ifelse(is.na(flag) & a == TRUE,a,NA)]
df


       a a_deduplicated flag dedup
 1:   NA             NA   NA    NA
 2:   NA             NA   NA    NA
 3:   NA             NA   NA    NA
 4: TRUE           TRUE   NA  TRUE
 5: TRUE             NA TRUE    NA
 6:   NA             NA TRUE    NA
 7: TRUE           TRUE   NA  TRUE
 8: TRUE             NA TRUE    NA
 9: TRUE             NA TRUE    NA
10:   NA             NA TRUE    NA
11:   NA             NA   NA    NA
12: TRUE           TRUE   NA  TRUE
13: TRUE             NA TRUE    NA
14: TRUE             NA TRUE    NA
15: TRUE             NA TRUE    NA

Upvotes: 1

Related Questions