how to run two matlab scripts in parallel?

210 views (last 30 days)
I need to run two scripts in parallel and simultaneously, run the first script and the second script at the same time.

Accepted Answer

KSSV
KSSV on 2 Jan 2017
You can use parfor to run two functions simultaneously. Eg:
funcs1 = {@fun1, @fun2} ; % let fun1, fun2 be two functions
arguments = {inp1,inp2 ;inp1,inp2} ; % write the inputs of each function
solutions = cell(1,2); % initialize the solution
% use of parfor
parfor ii = 1:2
solutions1{ii}=funcs1{ii}(arguments1{ii,:});
end
M = solutions1{1} ; N = solutions1{2} ; % assign the results

More Answers (4)

José-Luis
José-Luis on 2 Jan 2017
Edited: José-Luis on 2 Jan 2017
You could use parpool(). More flexible than parfor() but you'd need the parallel computing toolbox.
Another alternative is to start two sessions of Matlab and run the different scripts there.
  4 Comments
Walter Roberson
Walter Roberson on 25 Sep 2020
SPMD with labSend() and labReceive() is natural for that kind of work.
Hamza Ashraf
Hamza Ashraf on 26 Sep 2020
Can you explain it with code i dont know how to use spmd with labsend and lab receive

Sign in to comment.


Austin Bond
Austin Bond on 17 Apr 2018
should the 'arguments' and 'solutions' variables really be 'arguments1' and 'solutions1'?

Sean de Wolski
Sean de Wolski on 17 Apr 2018

Just call batch directly or right-click on the script in the current folder browser and select "Run as batch job"


MathWorks Support Team
MathWorks Support Team on 2 Sep 2022
You can use Parallel Computing Toolbox to run two MATLAB functions simultaneously. (If you have scripts, then first wrap them up as functions as described here in the doc Create Functions in Files). The parfeval function allows you to run two functions simultaneously on a parallel pool. Each call to parfeval causes the specified function to be invoked on a worker with the provided inputs. You also need to specify up front how many outputs you want. Your code might end up like this:
numOut = 1; % Both functions here have only a single output
f1 = parfeval(@myFirstFunction, numOut, in1, in2); % request myFirstFunction(in1,in2)
f2 = parfeval(@mySecondFunction, numOut, in3); % request mySecondFunction(in3)
out1 = fetchOutputs(f1); % this waits for myFirstFunction to complete and then gets the result
out2 = fetchOutputs(f2);
If you need to process the results as soon as either function completes, you could use fetchNext instead of fetchOutputs.
  1 Comment
Walter Roberson
Walter Roberson on 2 Sep 2022
These days, you can also use parfeval() and background threads without needing the Parallel Computing Toolbox; see backgroundPool

Sign in to comment.

Categories

Find more on Parallel Computing Fundamentals in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!