Faster way of applying `all` function?

Question

Suppose i have a logical array :

dimensions = c(10,100,200,300)
x = runif(prod(dimensions))>0.3
dim(x) = dimensions

Is there a faster way to get the result :

y = colMeans(apply(x,2:4,all))

?

Apply is quite slow, and this use case is quite straightforward, hence i wander...

Answer 1

Use colSums:

system.time(
  y <- colMeans(apply(x,2:4,all))
  )
# user      system     elapsed 
#6.916       0.058       6.978 

system.time(
z <- colMeans(colSums(x, dims = 1) == dim(x)[1])
)
# user      system     elapsed 
#0.117       0.000       0.117

identical(y, z)
#[1] TRUE