CODEWITHSUNDEEP
rfunctionfor-loop
rnorouzian
rnorouzian

Reputation: 7517

Making a function loopable in R

I'm trying to loop over f in my R code below. But f itself is a function (i.e., is.function(f) gives TRUE).

I'm wondering given that f is a function, is it possible to loop over it?

Note: f comes from f <- approxfun(...) and needs to remain as a function so I can use integrate(f, ...).

Here is my R code:

den = list()
x = rep(0, 2)
d = list(c(rbeta(1e4, 5, 5), rbeta(1e4, 2, 5)))
f = as.function()       ## What should this be? ##

for(i in 1:2){
den[[i]] <- density(d[[i]])
f[i] <- approxfun(den[[i]]$x, den[[i]]$y, yleft = 0, yright = 0) # how can it become loopable?
x[i] <- integrate(f[i], 0, .1)[[1]]
}

Upvotes: 2

Views: 86

Answers (2)

Rui Barradas
Rui Barradas

Reputation: 76663

Though there already is an accepted answer, I will post mine, since I believe that the main error in your code comes from trying too hard, you are complicating too much.

First of all, you don't need a list of functions f, you only need one function to use with integrate, like you say in the question.

Second, the way you created d uses an unnecessary call to c. This call will make of its arguments one vector, not two vectors of random beta numbers, which is what you want.

So the full code is just a simplification of yours.

set.seed(4149)    # makes the results reproducible

den = list()
x = rep(0, 2)
d = list(rbeta(1e4, 5, 5), rbeta(1e4, 2, 5))

for(i in 1:2){
    den[[i]] <- density(d[[i]])
    f <- approxfun(den[[i]]$x, den[[i]]$y, yleft = 0, yright = 0) # how can it become loopable?
    x[i] <- integrate(f, 0, .1)[[1]]
}

x
#[1] 0.001244482 0.110921241

Note the call to set.seed, when using RNG's it is a good idea to set the seed, in order to have reproducible data and results.

Upvotes: 1

kangaroo_cliff
kangaroo_cliff

Reputation: 6222

In the following, class(f[1]) is a list while class(f[[1]]) is a function. So, you simply need to have f[[i]] in the integrate function.

den = list()
x = rep(0, 2)

d <- list()
d[[1]] <- rbeta(1e4, 5, 5) 
d[[2]] <- rbeta(1e4, 2, 5)

f <- list()     

for(i in 1:2){

  den[[i]] <- density(d[[i]])
  f[[i]] <- approxfun(den[[i]]$x, den[[i]]$y, yleft = 0, yright = 0) 
  x[i] <- integrate(f[[i]], 0, .1)[[1]]

}

Upvotes: 1

Related Questions