apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-09 12:02:38 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix add kernel and VarDTC_minibatch speed tuning

Browse source
This commit is contained in:
Zhenwen Dai 2014-07-01 18:01:33 +01:00
parent 129985998c
commit 216de32c0c
3 changed files with 18 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

12
GPy/inference/latent_function_inference/var_dtc_parallel.py
View file

@ -94,8 +94,12 @@ class VarDTC_minibatch(LatentFunctionInference):
for n_start in xrange(0,num_data,self.batchsize): for n_start in xrange(0,num_data,self.batchsize):
n_end = min(self.batchsize+n_start, num_data) n_end = min(self.batchsize+n_start, num_data)
Y_slice = Y[n_start:n_end] if (n_end-n_start)==num_data:
X_slice = X[n_start:n_end] Y_slice = Y
X_slice = X
else:
Y_slice = Y[n_start:n_end]
X_slice = X[n_start:n_end]
if het_noise: if het_noise:
b = beta[n_start] b = beta[n_start]
@ -347,7 +351,9 @@ def update_gradients(model, mpi_comm=None):
while not isEnd: while not isEnd:
isEnd, n_range, grad_dict = model.inference_method.inference_minibatch(model.kern, X, model.Z, model.likelihood, Y) isEnd, n_range, grad_dict = model.inference_method.inference_minibatch(model.kern, X, model.Z, model.likelihood, Y)
if isinstance(model.X, VariationalPosterior): if isinstance(model.X, VariationalPosterior):
if mpi_comm ==None: if (n_range[1]-n_range[0])==X.shape[0]:
X_slice = X
elif mpi_comm ==None:
X_slice = model.X[n_range[0]:n_range[1]] X_slice = model.X[n_range[0]:n_range[1]]
else: else:
X_slice = model.X[model.N_range[0]+n_range[0]:model.N_range[0]+n_range[1]] X_slice = model.X[model.N_range[0]+n_range[0]:model.N_range[0]+n_range[1]]

7
GPy/kern/_src/add.py
View file

@ -63,13 +63,16 @@ class Add(CombinationKernel):
target = np.zeros(X.shape) target = np.zeros(X.shape)
[target.__iadd__(p.gradients_X_diag(dL_dKdiag, X)) for p in self.parts] [target.__iadd__(p.gradients_X_diag(dL_dKdiag, X)) for p in self.parts]
return target return target
@Cache_this(limit=2, force_kwargs=['which_parts'])
def psi0(self, Z, variational_posterior): def psi0(self, Z, variational_posterior):
return reduce(np.add, (p.psi0(Z, variational_posterior) for p in self.parts)) return reduce(np.add, (p.psi0(Z, variational_posterior) for p in self.parts))
@Cache_this(limit=2, force_kwargs=['which_parts'])
def psi1(self, Z, variational_posterior): def psi1(self, Z, variational_posterior):
return reduce(np.add, (p.psi1(Z, variational_posterior) for p in self.parts)) return reduce(np.add, (p.psi1(Z, variational_posterior) for p in self.parts))
@Cache_this(limit=2, force_kwargs=['which_parts'])
def psi2(self, Z, variational_posterior): def psi2(self, Z, variational_posterior):
psi2 = reduce(np.add, (p.psi2(Z, variational_posterior) for p in self.parts)) psi2 = reduce(np.add, (p.psi2(Z, variational_posterior) for p in self.parts))
#return psi2 #return psi2

8
GPy/kern/_src/psi_comp/__init__.py
View file

@ -10,7 +10,7 @@ import sslinear_psi_comp
class PSICOMP_RBF(Pickleable): class PSICOMP_RBF(Pickleable):
@Cache_this(limit=1, ignore_args=(0,)) @Cache_this(limit=2, ignore_args=(0,))
def psicomputations(self, variance, lengthscale, Z, variational_posterior): def psicomputations(self, variance, lengthscale, Z, variational_posterior):
if isinstance(variational_posterior, variational.NormalPosterior): if isinstance(variational_posterior, variational.NormalPosterior):
return rbf_psi_comp.psicomputations(variance, lengthscale, Z, variational_posterior) return rbf_psi_comp.psicomputations(variance, lengthscale, Z, variational_posterior)
@ -19,7 +19,7 @@ class PSICOMP_RBF(Pickleable):
else: else:
raise ValueError, "unknown distriubtion received for psi-statistics" raise ValueError, "unknown distriubtion received for psi-statistics"
@Cache_this(limit=1, ignore_args=(0,1,2,3)) @Cache_this(limit=2, ignore_args=(0,1,2,3))
def psiDerivativecomputations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, lengthscale, Z, variational_posterior): def psiDerivativecomputations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, lengthscale, Z, variational_posterior):
if isinstance(variational_posterior, variational.NormalPosterior): if isinstance(variational_posterior, variational.NormalPosterior):
return rbf_psi_comp.psiDerivativecomputations(dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, lengthscale, Z, variational_posterior) return rbf_psi_comp.psiDerivativecomputations(dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, lengthscale, Z, variational_posterior)
@ -30,7 +30,7 @@ class PSICOMP_RBF(Pickleable):
class PSICOMP_Linear(Pickleable): class PSICOMP_Linear(Pickleable):
@Cache_this(limit=1, ignore_args=(0,)) @Cache_this(limit=2, ignore_args=(0,))
def psicomputations(self, variance, Z, variational_posterior): def psicomputations(self, variance, Z, variational_posterior):
if isinstance(variational_posterior, variational.NormalPosterior): if isinstance(variational_posterior, variational.NormalPosterior):
raise NotImplementedError raise NotImplementedError
@ -39,7 +39,7 @@ class PSICOMP_Linear(Pickleable):
else: else:
raise ValueError, "unknown distriubtion received for psi-statistics" raise ValueError, "unknown distriubtion received for psi-statistics"
@Cache_this(limit=1, ignore_args=(0,1,2,3)) @Cache_this(limit=2, ignore_args=(0,1,2,3))
def psiDerivativecomputations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, Z, variational_posterior): def psiDerivativecomputations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, variance, Z, variational_posterior):
if isinstance(variational_posterior, variational.NormalPosterior): if isinstance(variational_posterior, variational.NormalPosterior):
raise NotImplementedError raise NotImplementedError