apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-06-02 14:45:15 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

renamed opimizer methods to unobscure gradients and objective

Browse source
This commit is contained in:
Max Zwiessele 2015-01-12 11:35:57 +00:00
parent ec42011617
commit dff577d22b
1 changed files with 8 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

14
GPy/core/model.py
View file

@ -151,7 +151,7 @@ class Model(Parameterized):
"""
return -(self._log_likelihood_gradients() + self._log_prior_gradients())
def _grads(self, x):
def _objective_grads(self, x):
"""
Gets the gradients from the likelihood and the priors.
@ -197,7 +197,7 @@ class Model(Parameterized):
return np.inf
return obj
def _objective_grads(self, x):
def _objective_and_grads(self, x):
try:
self.optimizer_array = x
obj_f, obj_grads = self.objective_function(), self._transform_gradients(self.objective_function_gradients())
@ -233,9 +233,7 @@ class Model(Parameterized):
"""
if self.is_fixed:
print 'nothing to optimize'
if self.size == 0:
if self.is_fixed or self.size == 0:
print 'nothing to optimize'
if not self.update_model():
@ -255,7 +253,7 @@ class Model(Parameterized):
optimizer = optimization.get_optimizer(optimizer)
opt = optimizer(start, model=self, **kwargs)
opt.run(f_fp=self._objective_grads, f=self._objective, fp=self._grads)
opt.run(f_fp=self._objective_and_grads, f=self._objective, fp=self._objective_grads)
self.optimization_runs.append(opt)
@ -312,7 +310,7 @@ class Model(Parameterized):
# evaulate around the point x
f1 = self._objective(x + dx)
f2 = self._objective(x - dx)
gradient = self._grads(x)
gradient = self._objective_grads(x)
dx = dx[transformed_index]
gradient = gradient[transformed_index]
@ -358,7 +356,7 @@ class Model(Parameterized):
print "No free parameters to check"
return
gradient = self._grads(x).copy()
gradient = self._objective_grads(x).copy()
np.where(gradient == 0, 1e-312, gradient)
ret = True
for nind, xind in itertools.izip(param_index, transformed_index):