assembling a correlation matrix

1 次查看（过去 30 天）
Bush Merc
Bush Merc 2013-10-19
评论： Azzi Abdelmalek 2013-10-19
Hi,
I am creating a correlation matrix (H) using the following routine. However for large values of M,N it takes a long time to create H. I was wondering if there other ways of creating H more efficiently.
X=15;
%
N=10;M=20;
K=randi(X,M,N);
L=randi(X,M,N);
H=zeros(X,X);
for k=1:X
for l=1:X
H(k,l)=hsum(k,l,K,L,M,N);
end
end
%where hsum is the following function
function ret=hsum(x,y,C,G,M,N)
ret=0;
for i=1:M
for j=1:N
if(C(i,j)==x && G(i,j)==y)
ret= ret+1;
end
end
end
Thanks Mark

请先登录，再回答此问题。

回答（1 个）

Azzi Abdelmalek
Azzi Abdelmalek 2013-10-19
You can change your function
function ret=hsum(x,y,C,G,M,N)
a=C(1:M,1:N);
b=G(1:M,1:N);
ret=sum(find(C==x and G==y));
  2 个评论
Azzi Abdelmalek
Azzi Abdelmalek 2013-10-19
编辑：Azzi Abdelmalek 2013-10-19
Bush Merc commented
Thanks for the suggestion. Unfortunately it is slower than the "for" loop.
function ret=hsum(x,y,C,G,M,N)
ret=sum(logical(find(C==x & G==y)));
Azzi Abdelmalek
Azzi Abdelmalek 2013-10-19
编辑：Azzi Abdelmalek 2013-10-19
try
unction ret=hsum(x,y,C,G,M,N)
ret=sum(C==x & G==y);
%If you want to add a comment, click on comment on this answer

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Creating and Concatenating Matrices 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by