Assign discrete values to meshgrid elements

I have a system of differential equations that takes in a pair of scalars (say C and D) and outputs a scalar E after evaluating the system numerically. I'm able to generate multiple E-values for an array of C- and D-values by using nested for loops. How can I plot the resulting values as a surface against a grid of the independent variables, rather than as a 3D scatter plot?

E is essentially a function of C and D, but cannot be expressed in closed form and must be evaluated numerically.

for i=1:n
    C=0.1*i;
    for j=1:n
        D=0.01*j;
        while ti<to
            [t, y]=ode23(@myFunction, [ti to], [0; 0; 0]);
            E(i,j)=y(end, 2);
        end
    end
end

% The following code will work, but the axes won't be oriented correctly
[C,D]=meshgrid(0:0.1:0.1*n,0:0.01:0.01*n);
surf(C,D,E);

Answer 1

[Walter Roberson]

Switch to surf(C, D, E.') or else change meshgrid to ndgrid()

[Ritwik Rakshit]

Transposing worked perfectly. Is there any documentation about the interaction of meshgrid and surf, or why transposition is necessary?

[Walter Roberson]

The syntax is surf(X, Y, Z) , and all three matrices must be the same size. But in MATLAB X is the rows and Y is the columns .

 [C, D] = meshgrid( 1:M, 1:N)

gives C as being N x M but

 for i : M
     for j = 1 : N
         E(i, j) = something
     end
 end

gives E as M x N.