Creating a new column with paste; constrict output so the smallest number comes first?

[jhodzic]

Hi all,

I have the following head of data

  OID IN_FID NEAR_FID NEAR_DIST
1   1      1       49  63.10777
2   2      2       63 154.24751
3   3      3        4 383.29095
4   4      4        3 383.29095
5   5      5        4 444.65797
6   6      6      171 365.11103

I would like to combine the IN_FID with NEAR_FID, and I did so by doing

paste(df$IN_FID, df$NEAR_FID, sep ="")

but, does anyone know if there is a way to arrange it such that the smaller of the two numbers (IN_FID vs NEAR_FID, always comes first?

Thanks!

[mattwarkentin]

This should do the trick. Customize as needed.

library(tidyverse)
library(glue)

dd <-
  tribble(
    ~IN_FID, ~NEAR_FID,
    1, 49,
    2, 63,
    3, 4,
    4, 3,
    5, 4,
    6, 17
  )

map2_chr(dd$IN_FID, dd$NEAR_FID, 
     ~glue_collapse(sort(c(.x, .y)), sep = " vs. "))
#> [1] "1 vs. 49" "2 vs. 63" "3 vs. 4"  "3 vs. 4"  "4 vs. 5"  "6 vs. 17"

[arnabp]

Another way of achieving the same thing ... Using the mapply function to find min and max and then combining as desired. Code snippet below might help

df1 <- data.frame(in_fid = c(1, 2, 3, 4, 5, 6), near_fid = c(49, 63, 4, 3, 4, 171))

paste(mapply(min, df1$in_fid, df1$near_fid),
mapply(max, df1$in_fid, df1$near_fid), sep = " ")

Thanks
Arnab P

[jhodzic]

Oh this is great! Thank you! Is there any way to transmute() this, i.e. to take out the IN_FID and NEAR_FID columns after I make the new column?

Thanks