Faster method of creating list of looped row vectors?

1 vue (au cours des 30 derniers jours)
mfas
mfas le 6 Juil 2015
Commenté : mfas le 6 Juil 2015
At the moment I have the following code:
kres=1;
kz=0;
K=[];
for kx=-3:kres:3
for ky=-3:kres:3
k=[kx,ky,kz];
K=[K;k];
end
end
to achieve a list of row vectors in the form:
[-3,-3,0;-3,-2,0;-3,-1,0;-3,0,0;....;3,2,0;3,3,0]
However when kres is reduced to < 0.01 this loop takes a long time to compute. Is there a faster way to achieve the same result without having to have a loop within a loop?

Réponse acceptée

Bjorn Gustavsson
Bjorn Gustavsson le 6 Juil 2015
kres=1; kz=0; K=[]; This is the way I'd go about it:
x = -3:kres:3;
y = -3:kres:3;
z = 0;
[x,y,z] = meshgrid(x,y,z);
K = [x(:),y(:),z(:)];
HTH

Plus de réponses (2)

Keith Hooks
Keith Hooks le 6 Juil 2015
You'll see quite a bit of improvement if you pre-allocate K. I understand the coding is not as clean, but the speed improvement is close to 5X for the 0.1 resolution.
Original: Elapsed time is 0.027472 seconds .
With pre-allocation: Elapsed time is 0.005479 seconds .
kres=.1;
kz=0;
kx = -3:kres:3;
ky = -3:kres:3;
n = length(kx);
m = length(ky);
K=zeros(n*m,3);
for i=1:n
for j=1:m
k=[kx(i),ky(j),kz];
K((i-1)*n + j,:) = k;
end
end

Thorsten
Thorsten le 6 Juil 2015
Modifié(e) : Thorsten le 6 Juil 2015
kres=0.01;
kx = -3:kres:3;
N = numel(kx);
k1 = repmat(kx, [N 1]);
K2 = [k1(:) repmat(kx', [N 1]) repmat(0, [N^2 1])];

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