# Compute the partial derivative numerically

76 views (last 30 days)
F.O on 21 Dec 2017
Edited: F.O on 21 Dec 2017
Hi, I want to compute the first and second partial derivative with respect to x, y for this function
x0=0
y0=0
x=[-1:0.1:1];
y=[-2:0.1:2];
v=x+exp(-((x-x0).^2+(y-y0).^2))

Star Strider on 21 Dec 2017
Example
x0=0;
y0=0;
x=[-1:0.1:1];
y=[-2:0.1:2];
[X,Y] = meshgrid(x, y);
v = @(x,y) x+exp(-((x-x0).^2+(y-y0).^2));
figure(1)
surf(X, Y, v(X,Y), 'FaceAlpha',0.5, 'EdgeColor',[0.3 0.3 0.7])
hold on
surf(X, Y, dX, 'EdgeColor','g')
surf(X, Y, dY, 'EdgeColor','r')
hold off
grid on
legend('v(x,y)', 'dX', 'dY')
xlabel('\bfX')
ylabel('\bfY')
view(40, 20)
figure(2)
contour(X, Y, v(X,Y))
hold on
quiver(X, Y, dX, dY)
hold off

Show 1 older comment
Star Strider on 21 Dec 2017
The easiest way is to use the related del2 (link) function. Remember to multiply it by 4 to get the second derivative:
d2Z = 4*del2(v(X,Y));
referencing my earlier code for the function call. See the del2 documentation for details, since it works a bit differently than gradient.
F.O on 21 Dec 2017