This is a dumb example, but I would like to learn a bit more about its syntax. I was expecting that adding @threads right in front of “for” would work out to make an array.
[f(x) @threads for x in 1:max_X]
[@threads f(x) for x in 1:max_X]
However, none of them works.
As far as I know, you can’t use Threads.@threads with a comprehension.
You can though use a threaded map. For example,
using Distributed
pmap(f, x)
and it’ll apply f to each element of x in a parallel way.
Alternatively, you can use the map from the package ThreadsX.
Note that while parallel this is not multithreaded. Instead it uses multiple worker processes (which btw need to be started first).
fetch.([Threads.@spawn f(i) for i in x]) usually does the trick for me. I don’t know if it will be the most performant but it’s good enough if i quickly want to parallelize something
There is also ThreadsX.map
Thank you very much, everyone.
I’ve decided to stick with multithreading due to the cluster system.
ThreadsX seems very impressive!