Say your mean is m, and your covariance is S.
m = [1 2];
S = [2 1; 1 10];
The easy way: If you have the statistics toolbox, there is a function called "gmdistribution"
G = gmdistribution(m,S)
F = @(x,y) pdf(G,[x y])
ezcontour(F)
The hard way: Find the equation for a multivariate normal distribution on Wikipedia:
It's at the top right hand side, and type that long expression into MATLAB:
F = @(x,y) (1/2*pi) * (1/sqrt(det(S))) * exp(-0.5* ...
ezcontour(F)
