Hi, just to revisit something you said a few weeks ago about this line of code
y_ = gp.marginal_likelihood(“y”,X=x,y=y, noise=s+er)
Could you expand a bit more on why this is technically incorrect. I do agree and it really doesn’t make much difference to my model, but I need to justify not including it to my supervisor. For context the error I am including is standard deviation from the measured stress values and was found in the experimental/modelling stages and provided to me in a dataset