for looop convert to arrayfun

3 vues (au cours des 30 derniers jours)
Question Mr
Question Mr le 29 Avr 2021
Modifié(e) : Matt J le 30 Avr 2021
I want to convert these for loops into an arrayfun. Always give me an error
"All of the input arguments must be of the same size and shape.
Previous inputs had size 31 in dimension 1. Input #5 has size 1"
[X,Y,Z] = sph2cart(ThetaRad,PhiRad,1.0);
nElec = length(M.lab);
EF(nElec,nElec) = 0;
for i = 1:nElec;
for j = 1:nElec;
EF(i,j) = 1 - ( ( (X(i) - X(j))^2 + ...
(Y(i) - Y(j))^2 + (Z(i) - Z(j))^2 ) / 2 );
end;
end;
I tried to create a new funciton:
function EF = gpuefg(X,Y,Z, maxiter)
EF(maxiter,maxiter) = 0;
for i = 1:maxiter;
for j = 1:maxiter;
EF(i,j) = 1 - ( ( (X(i) - X(j))^2 + ...
(Y(i) - Y(j))^2 + (Z(i) - Z(j))^2 ) / 2 );
end;
end;
end
And call it
EFG = gpuArray(EF);
.
.
.
EFG = arrayfun(@gpuefg,X,Y,Z, nElec, 'UniformOutput', false );
But not working, how should I do this?
  2 commentaires
David Hill
David Hill le 29 Avr 2021
Please explain what you are trying to do. What are your inputs and expected outputs?
Question Mr
Question Mr le 29 Avr 2021
Here is my original code: (not the whole is just the point)
[X,Y,Z] = sph2cart(ThetaRad,PhiRad,1.0);
nElec = length(M.lab);
EF(nElec,nElec) = 0;
for i = 1:nElec;
for j = 1:nElec;
EF(i,j) = 1 - ( ( (X(i) - X(j))^2 + ...
(Y(i) - Y(j))^2 + (Z(i) - Z(j))^2 ) / 2 );
end;
end;
output is a matrix
And I tried to create my own function for change two "for" loop
function EF = gpuefg(X,Y,Z, maxiter)
EF(maxiter,maxiter) = 0;
for i = 1:maxiter;
for j = 1:maxiter;
EF(i,j) = 1 - ( ( (X(i) - X(j))^2 + ...
(Y(i) - Y(j))^2 + (Z(i) - Z(j))^2 ) / 2 );
end;
end;
end
Output is matrix like in the original "EF(i,j)" (well that is the plan)
When I called:
EFG = gpuArray(EF);
...
EFG = arrayfun(@gpuefg,X,Y,Z, nElec, 'UniformOutput', false );
It's crashed but I dont know why,
EFG is a gpuArray,
X,Y,Z my function's inputs
nElec is my function's maxiter
So the point is, I want to swap these two "for loops" for one "arryfun"

Connectez-vous pour commenter.

Réponse acceptée

Matt J
Matt J le 29 Avr 2021
Modifié(e) : Matt J le 29 Avr 2021
The way to do this with arrayfun, (which I don't think will be optimal here) would be,
[I,J]=ndgrid(gpuArray(1:nElec));
fun=@(i,j) 1 - ( ( (X(i) - X(j))^2 + (Y(i) - Y(j))^2 + (Z(i) - Z(j))^2 ) / 2 );
EF= arrayfun(fun,I,J);
  4 commentaires
Question Mr
Question Mr le 30 Avr 2021
I have another question. I tried you solution but I got this error:
"Use of functional workspace is not supported.
For more information see Tips."
What should I do now? I've wroten the code into the .m file
Matt J
Matt J le 30 Avr 2021
Modifié(e) : Matt J le 30 Avr 2021
Why did you Accept the answer if it didn't work? Why not just go with the other answer, which I told you should be better than arrayfun?

Connectez-vous pour commenter.

Plus de réponses (1)

Matt J
Matt J le 29 Avr 2021
Modifié(e) : Matt J le 29 Avr 2021
The more efficient way to do what you are attempting is with the 2 lines of code,
X=X(:); Y=Y(:); Z=Z(:);
EF = 1 - ( (X-X.').^2 + (Y-Y.').^2 + (Z-Z.').^2)/2 ;

Catégories

En savoir plus sur Loops and Conditional Statements dans Help Center et File Exchange

Community Treasure Hunt

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

Start Hunting!

Translated by