Fill a nxm matrix with vectorised expression

Hi all,

here my doubt: i have a matrix M with dimensions nxm. The matrix is initially initialised with zeros, and i fill each M(i,j) calling another function:

M = zeros(n,m)
for i = 1:n
    for j = 1:m
        M(i,j) = another_function(several_inputs);
    end
end

is it possible to fill the matrix M without the for loops, but instead using a vectorised expression?

Thanks

[Konstantin]

What is the condition for selecting the function? Is the function different for each cell of the matrix, or is it possible, that the same function will be used for several cells?

[KSSV]

It depends on what your another_function is.

[Andrea Agosti]

Hi, thanks for your aswer. The inputs of the another_function are different for each i,j.

Thanks

[Konstantin]

sometimes it works to vectorize the function itself

matrix = another_function(matrix)

Answer 1

[Matt J]

No, there is no way to vectorize an element-wise function generically, without considering the specifics of the function. You can avoid for-loop syntax by using arrayfun, but this simply hides the loops from view, it doesn't replace them with something more efficient.

[Andrea Agosti]

thanks!