r - foreach unable to find object within function - r

Question

I have a function defined below named algor, which is translated from MATLAB to R. In order to make the function faster, I am using the foreach construct for the first time. I have the complete function code below:
library("ramify")
library("foreach")
algor <- function (vc) {
# initialize A, ybar, and Ia
A <- 0
ybar <- 0
Ia <- 0
# x is the first column of vc
x <- vc[, 1, drop = FALSE]
# y is the second column of vc
y <- vc[, 2, drop = FALSE]
# n is the length of x
n <- length(x)
foreach(i = 1:(n-1), .combine = 'c', .export = c("A", "ybar", "Ia", "x", "y")) %do% {
A <- A + 0.5 * (x[i] - x[i+1]) * (y[i] + y[i+1])
ybar <- ybar + (1 / 6) * (x[i] - x[i+1]) * (y[i] ^ 2 + y[i] * y[i+1] + y[i+1] ^ 2)
Ia <- Ia + (1 / 12) * (x[i] - x[i+1]) * (y[i] ^ 3 + y[i] ^ 2 * y[i+1] + y[i] * y[i+1] ^ 2 + y[i+1] ^ 3)
}
props <- mat("A, Ia, ybar", eval = TRUE)
return(props)
}
inner <- mat("0, 300; 300, 300; 300, 695; 0, 695; 0, 300")
algor(inner)
Although I have exported A, ybar, Ia, x, and y I am getting an error that the object A is not found, which is below:
Error in eval(parse(text = paste0("c(", paste0(char_vals, collapse = ","), :
object 'A' not found
Called from: eval(parse(text = paste0("c(", paste0(char_vals, collapse = ","),
")")))
How do I get foreach to recognize the defined objects: A, ybar, Ia, x, and y?
Thank you.

Answer 1

Try defining the .GlobalEnv variables within the foreach loop in every call.
library("ramify")
library("foreach")
algor <- function (vc) {
# initialize A, ybar, and Ia
A <- 0
ybar <- 0
Ia <- 0
# x is the first column of vc
x <- vc[, 1, drop = FALSE]
# y is the second column of vc
y <- vc[, 2, drop = FALSE]
# n is the length of x
n <- length(x)
foreach(i = 1:(n-1), .combine = 'c', .export = c("A", "ybar", "Ia", "x", "y")) %do% {
.GlobalEnv$A <- A
.GlobalEnv$ybar <- ybar
.GlobalEnv$Ia <- Ia
A <- A + 0.5 * (x[i] - x[i+1]) * (y[i] + y[i+1])
ybar <- ybar + (1 / 6) * (x[i] - x[i+1]) * (y[i] ^ 2 + y[i] * y[i+1] + y[i+1] ^ 2)
Ia <- Ia + (1 / 12) * (x[i] - x[i+1]) * (y[i] ^ 3 + y[i] ^ 2 * y[i+1] + y[i] * y[i+1] ^ 2 + y[i+1] ^ 3)
}
props <- mat("A, Ia, ybar", eval = TRUE)
return(props)
}
inner <- mat("0, 300; 300, 300; 300, 695; 0, 695; 0, 300")
algor(inner)
This returns:
[,1] [,2] [,3]
[1,] 118500 30870237500 58953750