Using multiple inputs with nftool

My task is to approximate an eggholder function ( https://www.sfu.ca/~ssurjano/egg.html ).

I am currently trying to figure out how to choose both of these 2 variables (x and y) as inputs in nftool.

If it's easier to do with another command, I could use it instead of nftool.

x = linspace(-512, 512, 1025);
y = linspace(-512, 512, 1025)'; 
for i=1:length(x)
    for j =1:length(y)
        z(i,j) = -(y(j,1)+47) .* sin(sqrt(abs(y(j,1)+x(1,i)/2+47))) - x(1,i) .* sin(sqrt(abs(x(1,i)-(y(j,1)+47))));
    end
end

Answer 1

[Sarthak]

Hi Zdenek,

As per my understanding you can reshape your inputs to a single input matrix.

inputs = [x(:), y(:)]; 

This inputs can then be used for any suitable function.

[Zdenek Vrana]

Thank you, it worked. However, I have another issue.

After training the network, every new input (eg. one pair of X and Y values) gives me 1025 predicted values. I think it's because the inputs (for traning) are only 2 rows, but the function (the result of those 2 inputs) is a 1025x1025 matrix.

How do I solve this?