apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-06-05 14:55:15 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

adding limited support for svg to have differnet number of latent functions to columns of Y

Browse source
This commit is contained in:
James Hensman 2015-04-29 21:40:51 +01:00
parent 49701b33c0
commit 3e4f272808
1 changed files with 7 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

10
GPy/core/svgp.py
View file

@ -9,7 +9,7 @@ from ..inference.latent_function_inference import SVGP as svgp_inf
class SVGP(SparseGP): class SVGP(SparseGP):
def __init__(self, X, Y, Z, kernel, likelihood, mean_function=None, name='SVGP', Y_metadata=None, batchsize=None): def __init__(self, X, Y, Z, kernel, likelihood, mean_function=None, name='SVGP', Y_metadata=None, batchsize=None, num_latent_functions=None):
""" """
Stochastic Variational GP. Stochastic Variational GP.
@ -41,8 +41,12 @@ class SVGP(SparseGP):
SparseGP.__init__(self, X_batch, Y_batch, Z, kernel, likelihood, mean_function=mean_function, inference_method=inf_method, SparseGP.__init__(self, X_batch, Y_batch, Z, kernel, likelihood, mean_function=mean_function, inference_method=inf_method,
name=name, Y_metadata=Y_metadata, normalizer=False) name=name, Y_metadata=Y_metadata, normalizer=False)
self.m = Param('q_u_mean', np.zeros((self.num_inducing, Y.shape[1]))) #assume the number of latent functions is one per col of Y unless specified
chol = choleskies.triang_to_flat(np.tile(np.eye(self.num_inducing)[:,:,None], (1,1,Y.shape[1]))) if num_latent_functions is None:
num_latent_functions = Y.shape[1]
self.m = Param('q_u_mean', np.zeros((self.num_inducing, num_latent_functions)))
chol = choleskies.triang_to_flat(np.tile(np.eye(self.num_inducing)[:,:,None], (1,1,num_latent_functions)))
self.chol = Param('q_u_chol', chol) self.chol = Param('q_u_chol', chol)
self.link_parameter(self.chol) self.link_parameter(self.chol)
self.link_parameter(self.m) self.link_parameter(self.m)