0

I know strictly programming question are out of skope on this platform, but I was wondering if any of you might have an answer for what I want to do. I am completly new in R, or coding.

This is an example of my data:

 X1  year individual partner 
  <dbl> <dbl> <chr> <chr>      
1     1  2015  A     B             
2     2  2015  A     B             
3     3  2015  B     A             
4     4  2015  C     A              
5     5  2015  C     D             
6     6  2015  D     C

I want to create a new column which will give me the id of the dyad. A dyad will be a unique combination of individual and partner. I can merge the columns together, but in that case A_B and B_A will be two different dyads however these two are actually the same dyad.

An example of what I want to achieve:

  X1  year individual partner dyad 
  <dbl> <dbl> <chr> <chr> <chr>
1     1  2015  A     B     A_B
2     2  2015  A     B     A_B
3     3  2015  B     A     A_B
4     4  2015  C     A     C_A
5     5  2015  C     D     C_D
6     6  2015  D     C     C_D

I there a way to merge two columns together but make a rule for how these are combined in the new column which would be appropriate in my case?

Thank you for your help.

valo
  • 25
  • 5
  • 1
    In 4th row shouldn't be `A_C`? – Duck Jul 12 '20 at 15:23
  • @Duck - It can be joined A_C or C_A. However, I want it to be the same whether A is the partner and C the individual, or C is the partner and A is the individual. Sorry, don't know if that is clear? – valo Jul 12 '20 at 15:31

1 Answers1

0

You can try this:

#Data
df <- structure(list(X1 = 1:6, year = c(2015L, 2015L, 2015L, 2015L, 
2015L, 2015L), individual = c("A", "A", "B", "C", "C", "D"), 
    partner = c("B", "B", "A", "A", "D", "C")), row.names = c("1", 
"2", "3", "4", "5", "6"), class = "data.frame")

#Create index
df$id <- apply(df[,c(3:4)],1,function(x)paste(sort(x),collapse = "_"))

  X1 year individual partner  id
1  1 2015          A       B A_B
2  2 2015          A       B A_B
3  3 2015          B       A A_B
4  4 2015          C       A A_C
5  5 2015          C       D C_D
6  6 2015          D       C C_D
Duck
  • 39,058
  • 13
  • 42
  • 84
  • Thank you so much that worked perfectly! Could you please explain to me how does the function work so I can understand what its doing? If you have the time! – valo Jul 12 '20 at 15:36
  • @ValsRo Great! For sure, apply with `margin=1` compute some function at row levels. You have to select the columns where you want to apply it `df[,c(3:4)]` and define the function. In your case `function(x)paste(sort(x),collapse = "_")` takes the elements in columns 3,4 and for each row it sorts `sort()` and then combine using `paste()`. The argument `collapse="_"` is used in order to add a sep :) – Duck Jul 12 '20 at 15:42
  • That's great! Thanks again for both the help and the explanation :) – valo Jul 12 '20 at 15:48