Model structure validation

[ToennisStef]

Is there was already some implementation of the model structure validation as it is indicated in the Bayesian Optimization Book from Garnett? Is there already some function or workflow to compute the model marginal posterior or the model structure posterior?

Model structural posterior:

Model marginal posterior:

Or is there a function or workflow implemented like in the book. Working with the simple example of the book:

Where the M5 stands for the Matern kernel and Lin for the Linear kernel. + indicates the summation of the covariance functions and x indicates the multiplication of the covariance functions.

Is there already some implementation of a function or a workflow that returns the probability of a model given the data compared to other models?