apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-11 21:12:38 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

adjusted parameters to report their changes

Browse source
This commit is contained in:
Max Zwiessele 2013-11-03 13:58:15 +00:00
parent 067206e83e
commit 6feb5dd2f1
10 changed files with 84 additions and 57 deletions
Download patch file Download diff file Expand all files Collapse all files

8
GPy/kern/kern.py
View file

@ -402,7 +402,7 @@ class kern(Parameterized):
"""Compute the gradient of the diagonal of the covariance function with respect to the parameters."""
assert X.shape[1] == self.input_dim
assert dL_dKdiag.size == X.shape[0]
target = np.zeros(self.num_params)
target = np.zeros(self.size)
[p.dKdiag_dtheta(dL_dKdiag, X[:, i_s], target[ps]) for p, i_s, ps in zip(self._parameters_, self.input_slices, self._param_slices_)]
return self._transform_gradients(target)
@ -418,7 +418,7 @@ class kern(Parameterized):
return target
def dpsi0_dtheta(self, dL_dpsi0, Z, mu, S):
target = np.zeros(self.num_params)
target = np.zeros(self.size)
[p.dpsi0_dtheta(dL_dpsi0, Z[:, i_s], mu[:, i_s], S[:, i_s], target[ps]) for p, ps, i_s in zip(self._parameters_, self._param_slices_, self.input_slices)]
return self._transform_gradients(target)
@ -433,7 +433,7 @@ class kern(Parameterized):
return target
def dpsi1_dtheta(self, dL_dpsi1, Z, mu, S):
target = np.zeros((self.num_params))
target = np.zeros((self.size))
[p.dpsi1_dtheta(dL_dpsi1, Z[:, i_s], mu[:, i_s], S[:, i_s], target[ps]) for p, ps, i_s in zip(self._parameters_, self._param_slices_, self.input_slices)]
return self._transform_gradients(target)
@ -480,7 +480,7 @@ class kern(Parameterized):
def dpsi2_dtheta(self, dL_dpsi2, Z, mu, S):
"""Gradient of the psi2 statistics with respect to the parameters."""
target = np.zeros(self.num_params)
target = np.zeros(self.size)
[p.dpsi2_dtheta(dL_dpsi2, Z[:, i_s], mu[:, i_s], S[:, i_s], target[ps]) for p, i_s, ps in zip(self._parameters_, self.input_slices, self._param_slices_)]
# compute the "cross" terms

22
GPy/kern/parts/rbf.py
View file

@ -55,6 +55,8 @@ class RBF(Kernpart):
self.lengthscale.add_observer(self, self.update_lengthscale)
self.add_parameters(self.variance, self.lengthscale)
self.update_lengthscale(self.lengthscale)
self.parameters_changed()
# initialize cache
#self._Z, self._mu, self._S = np.empty(shape=(3, 1))
#self._X, self._X2, self._params_save = np.empty(shape=(3, 1))
@ -65,7 +67,8 @@ class RBF(Kernpart):
'extra_link_args' : ['-lgomp']}
def on_input_change(self, X):
self._K_computations(X, None)
#self._K_computations(X, None)
pass
def update_lengthscale(self, l):
self.lengthscale2 = np.square(self.lengthscale)
@ -74,8 +77,8 @@ class RBF(Kernpart):
# reset cached results
#self._X, self._X2, self._params_save = np.empty(shape=(3, 1))
#self._Z, self._mu, self._S = np.empty(shape=(3, 1)) # cached versions of Z,mu,S
#self._X, self._X2 = np.empty(shape=(2, 1))
#self._Z, self._mu, self._S = np.empty(shape=(3, 1)) # cached versions of Z,mu,S
self._X, self._X2 = np.empty(shape=(2, 1))
self._Z, self._mu, self._S = np.empty(shape=(3, 1)) # cached versions of Z,mu,S
pass
# def _get_params(self):
# return np.hstack((self.variance, self.lengthscale))
@ -98,17 +101,16 @@ class RBF(Kernpart):
# return ['variance'] + ['lengthscale_%i' % i for i in range(self.lengthscale.size)]
def K(self, X, X2, target):
if self._X is None or X.base is not self._X.base or X2 is not None:
import pdb;pdb.set_trace()
self._K_computations(X, X2)
#if self._X is None or X.base is not self._X.base or X2 is not None:
self._K_computations(X, X2)
target += self.variance * self._K_dvar
def Kdiag(self, X, target):
np.add(target, self.variance, target)
def dK_dtheta(self, dL_dK, X, X2, target):
if self._X is None or X.base is not self._X.base or X2 is not None:
self._K_computations(X, X2)
#if self._X is None or X.base is not self._X.base or X2 is not None:
self._K_computations(X, X2)
target[0] += np.sum(self._K_dvar * dL_dK)
if self.ARD:
dvardLdK = self._K_dvar * dL_dK
@ -156,8 +158,8 @@ class RBF(Kernpart):
target[0] += np.sum(dL_dKdiag)
def dK_dX(self, dL_dK, X, X2, target):
if self._X is None or X.base is not self._X.base or X2 is not None:
self._K_computations(X, X2)
#if self._X is None or X.base is not self._X.base or X2 is not None:
self._K_computations(X, X2)
if X2 is None:
_K_dist = 2*(X[:, None, :] - X[None, :, :])
else: