Bayesian Analysis of Posterior Predictions from GP model

If I remove shape how would I get the posterior distributions for all 13 parameters? Should I just declare them all separately?