Quantcast
Channel: First steps - JuliaLang
Viewing all articles
Browse latest Browse all 2795

Trivial question about avoiding concurrency in a shared matrix between processes

$
0
0

Good evening everyone.

As the title suggests, I have a very trivial question to ask (I’m a beginner with julia).

In my code, I would like to have several (parallel) processes do calculations (quite long and complex) and each produce a matrix.

Then I would like all these processes to add each its own matrix into a single matrix. In practice, I would like to do a parallel reduction with an entire matrix instead of a single variable.

I had thought of using SharedArrays, that is, to declare a single shared matrix external to the parallel cycle distributed among the processes and then have each process add its matrix elements to this shared matrix, but what I have read is not reliable, since a concurrency would be created.

I hope it is clear what I intend to do, but if not I will try to explain myself better.

Thank you for your help!

1 post - 1 participant

Read full topic


Viewing all articles
Browse latest Browse all 2795

Latest Images

Trending Articles



Latest Images