How to apply bsxfun for floating point matrices

I have 2 floating point matrices.

How can i multiply corresponding entries in the matrices using bsxfun ?

Does bsxfun only work with numeric matrices ?

[Stephen23]

According to the documentation floating point arrays are a numeric data type . And bsxfun is certainly able to use floating point arrays.

Can you please post a comment with the complete code and error message.

If the two matrices are the same size (or one is a scalar), then you do not need to use bsxfun, you can simply use normal multiplication :

A.*B

Answer 1

[Matt J]

bsxfun works with both numeric and logical matrices. However, if the matrices are the same size, you should just do C=A.*B. If they are not the same size, then the syntax is

    C=bsxfun(@times,A,B);

but you need to be sure the dimensions that disagree are singletons in one of the matrices.

[Sreejith]

I used bsxfun to multiply 2 floating point matrices.

When i ran the program, it showed an error message

"error using bsxfun : operands must be numeric arrays"

How can i multiply corresponding entries in 2 floating point matrices ?

Is there any way ?

[Matt J]

You need to show us how you generated the input matrices, or attach them here in a .mat file. If your inputs really are floating point matrices, then MATLAB should see them as numeric.

[Sreejith]

my matrices are:

U 2 x 1250 matrix and R 1250 x 1250 matrix

both stores floating point values

i wrote the following code to store sum of product of their corresponding entries and when i ran the program the error was shown

j = 1;

while j < 1251

    s = sum(bsxfun(@times,u(c,:)',r(:,j)));
    minlist.data(j,1) = s;
    j = j + 1;
end

[Matt J]

You should attach a .mat file containing u and r so that we can check.

However, the line

     s = sum(bsxfun(@times,u(c,:)',r(:,j)));

is really the same as

      s=u(c,:)*r(:,j);

as people in other threads have told you.