Optimize code without for loop

My code uses 6 nested for loops, it is basically used to build 3D matrix with data in specific order from 2D matrix with data in random order.

These lines work as expected but take significant times to execute.

Is there any alternative approach of programming which lead to avoid all these for loops and therefore speed up the execution ?

for z=1:1001
    p=0;
    for j=1:length(DOFNameRef)
        for n=1:length(DOFDir)
            for i=1:length(DOFNamePoint)
                for l=1:length(DOFDir)
                    p = p+1;
                    for k=1:Mlig*Mcol
                       c(k) = contains(Resp(k),DOFNamePoint(i)) && contains(Resp(k),DOFDir(l)) && contains(Ref(k), DOFNameRef(j)) && contains(Ref(k),DOFDir(n));                                 
                    end
                    [NumLigne, NumCol] = NLignNCol(p,Mlig, Mcol);
                    M(NumLigne,NumCol,z) = Data(z, c); 
                end
            end
        end
    end
end

Answer 1

[AKennedy]

You can consider:

• Vectorization: Try to replace loops with vectorized operations. Eg., you can use logical indexing instead of the contains function inside the innermost loop.
• Preallocation: Preallocate all matrices and arrays before the loops. This avoids dynamically resizing arrays, which is computationally expensive.
• Parallel Computing: If you have the Parallel Computing Toolbox, you can use parfor to parallelize the outer loops.
• Logical Indexing: Instead of using nested loops to find elements that meet certain conditions, use logical indexing.

[Marc Servagent]

Thanks,

the vectorization helped to sigificantly improve the situation.