substract two strings in dplyr row wise for R dataframe - r

Question

Have two columns and need a third substracting the two using dplyr.
Very simple example for the sake of clarity. Split/separate approach not valid in my case.
x <- c("FRANCE","GERMANY","RUSSIA")
y <- c("Paris FRANCE", "Berlin GERMANY", "Moscow RUSSIA")
cities <- data.frame(x,y)
cities
x y
1 FRANCE Paris FRANCE
2 GERMANY Berlin GERMANY
3 RUSSIA Moscow RUSSIA
Expected results:
x y new
1 FRANCE Paris FRANCE Paris
2 GERMANY Berlin GERMANY Berlin
3 RUSSIA Moscow RUSSIA Moscow
What I've tried so far (to no avail):
this gets the very same df but removing the city (contrary as desired)
cities %>% mutate(new = setdiff(x,y))
x y new
1 FRANCE Paris FRANCE FRANCE
2 GERMANY Berlin GERMANY GERMANY
3 RUSSIA Moscow RUSSIA RUSSIA
On the contrary, setdiff in reverse order gets same initial data
cities %>% mutate(new = setdiff(y,x))
x y new
1 FRANCE Paris FRANCE Paris FRANCE
2 GERMANY Berlin GERMANY Berlin GERMANY
3 RUSSIA Moscow RUSSIA Moscow RUSSIA
Using gsub to remove worked just for first row issuing a warning
cities %>% mutate(new = gsub(x,"",y))
Warning message:
In gsub(x, "", y) :
argument 'pattern' has length > 1 and only the first element will be used
x y new
1 FRANCE Paris FRANCE Paris
2 GERMANY Berlin GERMANY Berlin GERMANY
3 RUSSIA Moscow RUSSIA Moscow RUSSIA

Answer 1

We can use stringr::str_replace:
library(tidyverse)
cities %>%
mutate_if(is.factor, as.character) %>%
mutate(new = trimws(str_replace(y, x, "")))
# x y new
#1 FRANCE Paris FRANCE Paris
#2 GERMANY Berlin GERMANY Berlin
#3 RUSSIA Moscow RUSSIA Moscow

Answer 2

Here is a solution with base R:
x <- c("FRANCE","GERMANY","RUSSIA")
y <- c("Paris FRANCE", "Berlin GERMANY", "Moscow RUSSIA")
cities <- data.frame(x,y,stringsAsFactors = F)
cities$new = mapply(function(a,b)
{setdiff(strsplit(a,' ')[[1]],strsplit(b,' ')[[1]])}, cities$y, cities$x)
Output:
x y new
1 FRANCE Paris FRANCE Paris
2 GERMANY Berlin GERMANY Berlin
3 RUSSIA Moscow RUSSIA Moscow
Hope this helps!