Remove rows if the swap also exist in the data frame in R

Question

I am trying to remove those rows if the swap also exists in the data frame.

For example, if I have a data frame:

1 2
1 3
1 4
2 4
4 2
2 1

Then the row (1,2), (2,4) will be removed because (2,1) and (4,2) are also in the df. Is there any fast and neat way to do it? Thank you!

Solution

You can row-wise sort the columns and then select only the unique ones :

library(dplyr)

df %>%
 mutate(col1 = pmin(V1, V2), 
        col2 = pmax(V1, V2)) %>%
 distinct(col1, col2)

#  col1 col2
#1    1    2
#2    1    3
#3    1    4
#4    2    4

---

Using base R :

df1 <- transform(df, col1 = pmin(V1, V2), col2 = pmax(V1, V2))
df[!duplicated(df1[3:4]), ]

data

df <- structure(list(V1 = c(1L, 1L, 1L, 2L, 4L, 2L), V2 = c(2L, 3L, 
4L, 4L, 2L, 1L)), class = "data.frame", row.names = c(NA, -6L))

Comments

[Ronak Shah]

Can same row be repeated twice? For eg - (1, 2) and (1, 2) ?

[DigiPath]

no, if (1, 2) is in the list then (2,1) can not be in the list

[DigiPath]

I like the dplyr solution, it is pretty fast! Thank you!

[Ronak Shah]

No, wait. Do you want to remove both the original and the swap? meaning (1, 2) and (2, 1) ? My answer only removes (2, 1).

[Ronak Shah]

In case if you want to remove both, you can use df[!(duplicated(df1[3:4]) | duplicated(df1[3:4], fromLast = TRUE)), ]

[DigiPath]

No I mean remain either (1,2) or (2,1) so I think your answer meets my need. Thank you!