How do I select all unique combinations of two columns in an R data frame?

Question

I have a correlation matrix that I put in a dataframe like so:

row | var1 | var2 | cor
1   | A    | B    | 0.6
2   | B    | A    | 0.6
3   | A    | C    | 0         


        

Answer 1

Here is yet another tidyverse partial solution, as I have dropped the row column:

library(tidyverse)

data.cor <-
  read.table(
    h = T,
    sep = "|",
    stringsAsFactors = F,
    text = "row | var1 | var2 | cor
            1   | A    | B    | 0.6
            2   | B    | A    | 0.6
            3   | A    | C    | 0.4
            4   | C    | A    | 0.4"
  ) %>%
  mutate_if(is.character, ~ trimws(.))

data.cor
#>   row var1 var2 cor
#> 1   1    A    B 0.6
#> 2   2    B    A 0.6
#> 3   3    A    C 0.4
#> 4   4    C    A 0.4

df <- data.cor %>%
  gather(var, val, var1:var2) %>%
  distinct(cor, val) %>%
  group_by(cor) %>%
  mutate(x = paste("var", 1:n(), sep = "")) %>%
  spread(x, val) %>% 
  ungroup()

df
#> # A tibble: 2 x 3
#>     cor var1  var2 
#>     
#> 1   0.4 A     C    
#> 2   0.6 A     B

^{Created on 2019-04-18 by the reprex package (v0.2.1)}