The sizes would not work out if you did that. You are passing in 81 x 51 matrices X and Y, which become known by the names x and y in the function. Inside the function, you want to replace x with a vector of two elements. Then you have
z=3.*sin(0.5+0.25.*x.*y).*cos(x)
with x now replaced by a vector of two elements, x.*y would be either 1 x 2 or 2 x 1 (depending on row or column vector) times an 81 x 51 matrix, and that cannot work.
But perhaps you did not want to replace the first x in that expression, only the second. In that case you would have 3.*sin(0.5+0.25.*x.*y) be an 81 x 51 result, and you would be trying to multiply it by cos() of either a 1 x 2 or 2 x 1, and that cannot work for size reasons either.
I am not at all clear as to why you think you need to construct that vector? If you just leave that line out then you get a reasonable result.
