You could try the table functions from
https://github.com/rened/FunctionalData.jl
julia> Pkg.update(); Pkg.add("FunctionalData")
julia> ptable((x,y)->x+y, 1:2, 1:3)
2x3 Array{Int64,2}:
2 3 4
3 4 5
table does not parallelize, ptable uses all workers, and ltable uses only the
local workers.
In case your results cannot be stored in a dense numerical array, use ptableany
instead.
I just added this - if it does not work for your case please file an issue.
Am 19.02.2015 um 01:40 schrieb DumpsterDoofus <[email protected]>:
> I sometimes generate arrays using array comprehension syntax like this:
>
> array = [func(a,k) for a in 0.0:0.01:1.0, k in 0.0:0.01:1.0];
>
> or alternately using `broadcast` like this:
>
> array = broadcast(func, linspace(0,1,101)', linspace(0,1,101))
>
> This is the Julia equivalent to the following Mathematica code:
>
> array = Outer[func, Range[0,1,0.01], Range[0,1,0.01]];
>
> To parallelize in Mathematica, you typically wrap the expression in
> `Parallelize`:
>
> array = Parallelize[Outer[func, Range[0,1,0.01], Range[0,1,0.01]]];
>
> Is there a simple way to "parallel broadcast" or "parallel array
> comprehension" in Julia yet? I know I could manually chop the workload into
> chunks, and then feed them to available worker processes using `@spawnat`,
> and then fuse the results together into one large array, but I was wondering
> if there was a non-kludgy way to automatically distribute array comprehension
> syntax.
