[julia-users] Re: Setting RNG seeds for parallel computing

[Gray Calhoun]

On Tuesday, July 15, 2014 5:25:11 PM UTC-5, Gray Calhoun wrote:
> Hi everyone, I'm trying to start using Julia for some Monte
> Carlo simulations (not MCMC) which I'd like to parallelize. I
> haven't found any documentation for setting the RNG's seed for
> parallelization. The naive approach gives different results
> than non-parallel execution (which is not surprising).
[rest of message cut]

Thanks to everyone who replied to my original email.  Just to follow
up, it turns out that `pmap` gives a simple solution, since it manages
the parallel tasks from the main process. The following code gives the
general idea:

```
addprocs(1)
srand(1)
function rvproducer()
    for i=1:4
        produce(rand())
    end
end

parallelrvs = pmap(Task(rvproducer)) do x
    println(x) ## verify that the work is done on different processes
    return x
end

srand(1)
sort(parallelrvs) == sort(rand(4)) ## returns true
```

--Gray