How to simplify a code to remove a for loop and make it faster?

Hello,

I'm still learning vectorization and haven't figured out how to simplify this code:

B=zeros(1,a^2);

for j=1:a^2;
    B(:)=B(:)+C(:,j)*D(j);
end;

in which B is a vector of 1 row by a^2 columns, C is an a x a matrix, and D is a vector of 1 row by a^2 columns

So basically what is happening is I take all the columns of C (one by one) and multiply them by D and then add them to B but this is repeated column by column on C.

I appreciate if anyone can show me how to do this.

Thank you very much.

Answer 1

[Iain]

One of these ought to do it:

 B = C' * D;

or

 B = C' * D';

You might need to transpose B to get what you want....

Answer 2

[Azzi Abdelmalek]

Edit

a=4
C=rand(a^2);
D=rand(1,a^2);
B=sum(bsxfun(@times, C,D),2)

[carlos Uribe]

sorry my bad...C is an a^2 x a^2 array

[Azzi Abdelmalek]

Look at edited answer

Answer 3

[Andrei Bobrov]

   B = D*C';