Apparently theano.shared does not work with pm.gp.Latent.prior for testval either. I tried to do the “using hold-out testing values” approach as suggested in the FAQ. I’d like to do something like J_init = theano.shared(np.zeros(n_coords)) gp = pm.gp.Latent(cov_func=cov) J = gp.prior('J', X, testv…

Testval for gp.Latent cannot be theano.shared?

bwengals November 11, 2017, 12:27am 5

Right, I see what you mean now. You certainly could if that works for you. I think using a numpy array is the most well worn usage path here though.

Topic		Replies	Views
Test_value shape errors with theano.shared Questions	13	2586	March 27, 2018
Latent GP prior on X=theano.shared node with changing shape Questions theano	2	636	November 17, 2017
Inserting pymc symbollic variables into theano shared array Questions	2	423	December 27, 2018
Problem with shared variable for means Questions theano	2	1131	June 14, 2019
Shapes of shared variables Questions theano	3	987	June 29, 2018