CODEWITHSUNDEEP
r
mahmoud fathy
mahmoud fathy

Reputation: 391

R integrate : evaluation of function gave a result of wrong type

I am having this annoying error again in R that I thought I understood well and I could avoid it for enough time. But now it's starting to haunt me again.

inner.f.y <- function(y)
{
    cat("length(y)   ", length(y), "\n")

    t<-2*y*exp((exp(-1*1i)-1)*y)

    cat("length(t)   ", length(t), "\n")
    t
}
integrate(inner.f.y, lower = 0.01, upper = 8)

This is the output on the console

> integrate(inner.f.y, lower = 0.01, upper = 8)
length(y)    21 
length(t)    21 
Error in integrate(inner.f.y, lower = 0.01, upper = 8) : 
  evaluation of function gave a result of wrong type

So what can it be this issue ?

Upvotes: 1

Views: 1753

Answers (1)

scoa
scoa

Reputation: 19857

See ?integrate (my emphasis):

f : an R function taking a numeric first argument and returning a numeric vector of the same length.

Your function inner.f.y does not return a numeric vector, but a complex one:

a <- inner.f.y(1)

is.numeric(a)
[1] FALSE

is.complex(a)
[1] TRUE

As @Ben Bolker suggests, you could use elliptic::myintegrate instead:

elliptic::myintegrate(inner.f.y, lower = 0.01, upper = 8)
length(y)    21 
length(t)    21 
length(y)    21 
length(t)    21 
[1] -1.150277-1.380324i

Upvotes: 3

Related Questions