Thanks very much for the detailed example. This particular description is very helpful >There is an exponential falloff between these groups of some similarity

Thanks very much for the detailed example. This particular description is very helpful >There is an exponential falloff between these groups of some similarity measure, but its continuous not discrete. That's a lot of food for thoughts for me to digest, however I still struggle to imagine the more refined structure that GPs can capture via the form of covariance function. More reading I guess. Thanks again :-D