Merge pull request #800 from sighellan/devel

Minor fix to multioutput regression example, to clarify code + typo.
This commit is contained in:
Zhenwen Dai 2020-03-09 12:15:41 +00:00 committed by GitHub
commit 4ac66e2336
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23

View file

@ -620,8 +620,10 @@ def multioutput_gp_with_derivative_observations():
# Then create the model, we give everything in lists, the order of the inputs indicates the order of the outputs
# Now we have the regular observations first and derivative observations second, meaning that the kernels and
# the likelihoods must follow the same order. Crosscovariances are automatically taken car of
m = GPy.models.MultioutputGP(X_list=[x, xd], Y_list=[y, yd], kernel_list=[se, se_der], likelihood_list = [gauss, gauss])
# the likelihoods must follow the same order. Crosscovariances are automatically taken care of
m = GPy.models.MultioutputGP(X_list=[x, xd], Y_list=[y, yd],
kernel_list=[se, se_der],
likelihood_list=[gauss, gauss_der])
# Optimize the model
m.optimize(messages=0, ipython_notebook=False)