CODEWITHSUNDEEP
r
Dileep Kumar Patchigolla
Dileep Kumar Patchigolla

Reputation: 997

How can we replace elements in a vector in R?

Lets say, I have a vector

animal <- c('cat','snake','cat','pigeon','snake')

And a dataframe called

map <- data.frame(find=c('cat','snake','pigeon'),replace=c('mammal','reptile','bird')

Now I want animal to be modified using the map by matching each element of animal to the replace column of map. So I am expecting:

animal <- c('mammal','reptile','mammal','bird','reptile')

How can I do this without using loops over each element of my first vector?

Upvotes: 1

Views: 6351

Answers (3)

Marco
Marco

Reputation: 9624

You can treat animal as a factor and rename its levels. For example, using the plyr package:

library(plyr)
animal <- c('cat','snake','cat','pigeon','snake')
animal2 <- revalue(animal, c("cat" = "mammal", 
                             "snake" = "reptile", 
                             "pigeon" = "bird"))

> animal
[1] "cat"    "snake"  "cat"    "pigeon" "snake" 
> animal2
[1] "mammal"  "reptile" "mammal"  "bird"    "reptile"

To make it automatic as required in the comment below

repl <- as.character(map$replace)
names(repl) <- map$find
animal2 <- revalue(animal, repl)

> animal
[1] "cat"    "snake"  "cat"    "pigeon" "snake" 
> animal2
[1] "mammal"  "reptile" "mammal"  "bird"    "reptile"

Upvotes: 2

mark999
mark999

Reputation: 185

You can use the match function:

> animal <- as.character(map[match(animal, map$find), "replace"])
> animal
[1] "mammal"  "reptile" "mammal"  "bird"    "reptile"

Upvotes: 3

Tim
Tim

Reputation: 7474

For this purpose I use simple recode function. As input you need vector to be changed denoted as x, vector of values to be changes from and a vector of replacement values to (so from[1] is recoded to to[1]), you can specify other value and all the values that are not in from will be recoded to other. You can find the function pasted below.

recode <- function(x, from, to, other) {
  stopifnot(length(from) == length(to))

  new_x <- x
  k <- length(from)
  for (i in 1:k) new_x[x == from[i]] <- to[i]

  if (!missing(other) && length(other) > 1) {
    new_x[!(x %in% from)] <- other[1]
    warning("'other' has length > 1 and only the first element will be used")
  }

  new_x
}

using your own example the usage would be

recode(animal, map$find, map$replace)

Upvotes: 1

Related Questions