CODEWITHSUNDEEP
rfunction
Daniel
Daniel

Reputation: 33

R: Defining functions within a loop

Defining multiple functions in a loop:

par <- 1:2    #parameters for functions
qF <- list()  #list I will write the functions into

for(i in 1:2){
  qF[[i]] <- function(p){qnorm(p, mean = par[i])}
}

My result:

>qF
[[1]]
function (p) 
{
qnorm(p, mean = par[i])
}

[[2]]
function (p) 
{
qnorm(p, mean = par[i])
}

The functions are the same! What I WANT my result to be:

>qF
[[1]]
function (p) 
{
 qnorm(p, mean = par[1])
}

[[2]]
 function (p) 
{
 qnorm(p, mean = par[2])
}

Is there any way to do this?

Upvotes: 3

Views: 1904

Answers (2)

Rorschach
Rorschach

Reputation: 32426

If you want to create the functions dynamically you need for the par[i] to evaluate each time, otherwise all the par[i] will be evaluated when the functions are called, and i will be the value at the end of the loop.

for(i in 1:2){
    qF[[i]] <- local({
        mu <- par[i]
        function(...) { qnorm(..., mean = mu) }   
    })
}

You could also substitute the variable into the function body

for(i in 1:2){
    qF[[i]] <- eval(substitute(
        function(...) qnorm(..., mean = mu)), list(mu=par[i]))
}

And you can see what mu is in each function's environment

sapply(qF, function(f) mget("mu", environment(f)))
# $mu
# [1] 1
# 
# $mu
# [1] 2

Upvotes: 3

Colonel Beauvel
Colonel Beauvel

Reputation: 31171

You can do:

library(functional)

funcs = lapply(1:2, function(u) Curry(qnorm, mean=u)) 

funcs[[1]](0.77)
#[1] 1.738847

funcs[[2]](0.77)
#[1] 2.738847

Or if you do not appreciate spicy meals:

funcs = lapply(1:2, function(u) function(...) qnorm(..., mean=u))

Upvotes: 2

Related Questions