The best explanation I found on this topic. No fancy equations, or formulas; a very intuitive approach to the plot. This video deserves much more views Keep up the good work!
This is a great explanation of how kde bivariate plots works. Intuitively, you can guess what the figure tells you, but this video tells why so nicely!
Is it possible to get the information on the contour area ? What I mean is after computing the probability density function of a 2D gaussian distribution, the hill will be in the third dimension, Can I get the values of the projection of this (till 3 sigma values of the variables) ?
In most instances, you need not implement the underlying code behind the jointplot function which plots a contour plot. But, in case you are curious about how it's actually implemented, here is one of the simple ways to do it. Take various 2D points(with x and y coordinates) and compute the z (the density) at each of these 2D points.. Now join the 2D points which have the same value of z into a line and color it with darkness proportional to the value of density (z).
