What is the best way to combine two or more DelayedArray to a high dimension DelayedArray with low cost of memory?

for example:

a <- array(runif(1500000), dim=c(10000, 30, 5))
A <- DelayedArray(a)
b <- array(runif(1500000), dim=c(10000, 30, 5))
B <- DelayedArray(b)
DelayedArray(arrayc(A, B), dim=c(10000, 30, 5, 2))

Thank you!

There's no general abind() method for DelayedArray at this time. However, for your particular example, you can achieve it with some trickery using aperm() with arbind():

suppressPackageStartupMessages(library(DelayedArray))

# NOTE: I add a 4th dimension to the array
a <- array(runif(1500000), dim=c(10000, 30, 5, 1))
A <- DelayedArray(a)
b <- array(runif(1500000), dim=c(10000, 30, 5, 1))
B <- DelayedArray(b)
C <- aperm(arbind(aperm(A, 4:1), aperm(B, 4:1)), 4:1)
dim(C)
#> [1] 10000    30     5     2

Created on 2018-03-16 by the reprex package (v0.2.0).

This is memory-efficient as aperm() only permutes some index variables and not the data itself.