Context: I am trying to model Probabilistic power flow in Gaussian Process framework using gpml toolbox. Question: The problem with the command [ymu1, ys1,f,s] = gp(hyp, @infGaussLik, meanfunc, covfunc, likfunc, X, v,x_test); is that I'm getting the same variances for every prediction of bus voltage. I', certain that the model should predict generation bus voltage with zero mean. What is the right way to get the variances of each prediction in gpml?
Thank You!