How to plot contour for gaussian copula (etc.)

From matlab help center, I can generate Gaussian copula random variables and plot it like this:

n = 500;

rng('default') % for reproducibility
U = copularnd('Gaussian',[1 .8; .8 1],n);
plot(U(:,1),U(:,2),'.')
title('{\it\rho} = 0.8')
xlabel('U1')
ylabel('U2')

The question is, how could i plot the "contour plots" of bivariate copula models like this (the right pic):

It may probabilily be marginally normalized, but i don't know how to do it by function "contour".

Thanks for your helping!!

[Mann Baidi]

Hi,

Have you tried 'copulapdf' function? If not, you can refer to the following link:

https://www.mathworks.com/help/stats/copulapdf.html

[Mathieu NOE]

you can make a contour plot from scattered data using this Fex submission :

Contour Plot for Scattered Data - File Exchange - MATLAB Central (mathworks.com)

Answer 1

[Gautam]

You can use the “copulapdf” function to plot a bivariate gaussian copula pdf and then use the “contour” on these pdf values function for the contour plot.

The code below generates the pdf

u = linspace(0,1,25); 
[u1,u2] = meshgrid(u,u); 

y = copulapdf('Gaussian',[u1(:),u2(:)],[1 .8; .8 1]); 
surf(u1,u2,reshape(y,25,25)); 

Using the “contour” command on this pdf will give you the contour plot.

contour(reshape(y, 25,25), 15);