Reputation: 161
I have created a list which itself consists of several lists. I am trying to determine an efficient way to get the "key" from the "value". I.e. if I specify ("cat" or "dog"), ("fish" or "chicken"), ("horse" or "donkey") how can I return "pet", "food", and "work" respectively. I have tried to create a method with a for loop since I am not sure how else to approach with indexing by name.
pet <- c("cat", "dog")
food <- c("fish", "chicken")
work <- c("horse", "donkey")
types <- c("pet", "food", "work")
animal.list <- vector(mode = "list", length = length(types))
names(animal.list) <- types
for (i in types)
{
animal.list[[i]] <- vector(mode = "list", length = length(c("a", "b")))
names(animal.list[[i]]) <- c("a", "b")
animal.list[[i]][["a"]] <- eval(parse(text = i))[[1]]
animal.list[[i]][["b"]] <- eval(parse(text = i))[[2]]
}
My attempt would look something like this, but hopefully I could use some kind of which( %in%) statement to do it more efficiently/compactly.
f <- function(x)
{
ret <- NULL
for (i in animals)
{
if(x == animal.list[[i]][["a"]] | x == animal.list[[i]][["b"]])
{
ret <- i
}
}
}
Upvotes: 4
Views: 2932
Reputation: 52687
You can create a lookup table with stack
, and then use match
to find the values:
animals <- stack(list(pet=pet, food=food, work=work))
f <- function(x) as.character(animals[match(x, animals[[1]]), 2])
Then:
f("cat")
# [1] "pet"
f("horse")
# [1] "work"
Note %in%
is just a variation on match
.
You can also use R's built in character lookup:
animal.vec <- as.character(animals[[2]])
names(animal.vec) <- animals[[1]]
animal.vec[c("cat", "horse")]
# cat horse
# "pet" "work
Upvotes: 7
Reputation: 72769
You can select multiple elements of a list with a single bracket rather than a double. Combining that with any
and %in%
:
# ("cat" or "dog")
idx <- sapply( animal.list, function(x) any( x %in% c("cat", "dog")) )
names( animal.list )[ idx ]
Upvotes: 1