apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-24 14:15:14 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

[GPU] gradient check ready

Browse source
This commit is contained in:
Zhenwen Dai 2014-04-02 11:48:27 +01:00
parent b90a867232
commit 24cc9c1bc3
2 changed files with 39 additions and 44 deletions
Download patch file Download diff file Expand all files Collapse all files

2
GPy/kern/_src/psi_comp/ssrbf_psi_gpucomp.py
View file

@ -234,7 +234,7 @@ try:
double psi2exp1_c = psi2exp1[IDX_NMMQ(n,m1,m2,q)]; double psi2exp1_c = psi2exp1[IDX_NMMQ(n,m1,m2,q)];
double psi2exp2_c = psi2exp2[IDX_MMQ(m1,m2,q)]; double psi2exp2_c = psi2exp2[IDX_MMQ(m1,m2,q)];
double dZ = Z1_c - Z2_c; double dZ = Z2_c - Z1_c;
double muZ = mu[IDX_NQ(n,q)] - (Z1_c+Z2_c)/2.0; double muZ = mu[IDX_NQ(n,q)] - (Z1_c+Z2_c)/2.0;
double Z2 = Z1_c*Z1_c+Z2_c*Z2_c; double Z2 = Z1_c*Z1_c+Z2_c*Z2_c;
double denom = 2.0*S_c/l_c+1.0; double denom = 2.0*S_c/l_c+1.0;

81
GPy/kern/_src/rbf.py
View file

@ -138,24 +138,21 @@ class RBF(Stationary):
def gradients_Z_expectations(self, dL_dpsi1, dL_dpsi2, Z, variational_posterior): def gradients_Z_expectations(self, dL_dpsi1, dL_dpsi2, Z, variational_posterior):
# Spike-and-Slab GPLVM # Spike-and-Slab GPLVM
if isinstance(variational_posterior, variational.SpikeAndSlabPosterior): if isinstance(variational_posterior, variational.SpikeAndSlabPosterior):
dL_dpsi1_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi1)) if self.useGPU:
dL_dpsi2_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi2)) dL_dpsi1_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi1))
gZ = self.psicomp.gradients_Z_expectations(dL_dpsi1_gpu, dL_dpsi2_gpu, self.variance, self.lengthscale, Z, variational_posterior) dL_dpsi2_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi2))
return self.psicomp.gradients_Z_expectations(dL_dpsi1_gpu, dL_dpsi2_gpu, self.variance, self.lengthscale, Z, variational_posterior)
else:
_, _, _, _, _, _dpsi1_dZ, _ = ssrbf_psi_comp._psi1computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob)
_, _, _, _, _, _dpsi1_dZ, _ = ssrbf_psi_comp._psi1computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob) _, _, _, _, _, _dpsi2_dZ, _ = ssrbf_psi_comp._psi2computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob)
_, _, _, _, _, _dpsi2_dZ, _ = ssrbf_psi_comp._psi2computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob)
#psi1
#psi1 grad = (dL_dpsi1[:, :, None] * _dpsi1_dZ).sum(axis=0)
grad = (dL_dpsi1[:, :, None] * _dpsi1_dZ).sum(axis=0)
#psi2
#psi2 grad += (dL_dpsi2[:, :, :, None] * _dpsi2_dZ).sum(axis=0).sum(axis=1)
grad += (dL_dpsi2[:, :, :, None] * _dpsi2_dZ).sum(axis=0).sum(axis=1)
return grad
print np.abs(gZ - grad).max()
return grad
elif isinstance(variational_posterior, variational.NormalPosterior): elif isinstance(variational_posterior, variational.NormalPosterior):
l2 = self.lengthscale **2 l2 = self.lengthscale **2
@ -179,32 +176,30 @@ class RBF(Stationary):
def gradients_qX_expectations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, Z, variational_posterior): def gradients_qX_expectations(self, dL_dpsi0, dL_dpsi1, dL_dpsi2, Z, variational_posterior):
# Spike-and-Slab GPLVM # Spike-and-Slab GPLVM
if isinstance(variational_posterior, variational.SpikeAndSlabPosterior): if isinstance(variational_posterior, variational.SpikeAndSlabPosterior):
dL_dpsi1_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi1)) if self.useGPU:
dL_dpsi2_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi2)) dL_dpsi1_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi1))
gmu,gS,gg = self.psicomp.gradients_qX_expectations(dL_dpsi1_gpu, dL_dpsi2_gpu, self.variance, self.lengthscale, Z, variational_posterior) dL_dpsi2_gpu = gpuarray.to_gpu(np.asfortranarray(dL_dpsi2))
return self.psicomp.gradients_qX_expectations(dL_dpsi1_gpu, dL_dpsi2_gpu, self.variance, self.lengthscale, Z, variational_posterior)
else:
ndata = variational_posterior.mean.shape[0] ndata = variational_posterior.mean.shape[0]
_, _, _dpsi1_dgamma, _dpsi1_dmu, _dpsi1_dS, _, _ = ssrbf_psi_comp._psi1computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob) _, _, _dpsi1_dgamma, _dpsi1_dmu, _dpsi1_dS, _, _ = ssrbf_psi_comp._psi1computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob)
_, _, _dpsi2_dgamma, _dpsi2_dmu, _dpsi2_dS, _, _ = ssrbf_psi_comp._psi2computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob) _, _, _dpsi2_dgamma, _dpsi2_dmu, _dpsi2_dS, _, _ = ssrbf_psi_comp._psi2computations(self.variance, self.lengthscale, Z, variational_posterior.mean, variational_posterior.variance, variational_posterior.binary_prob)
#psi1 #psi1
grad_mu = (dL_dpsi1[:, :, None] * _dpsi1_dmu).sum(axis=1) grad_mu = (dL_dpsi1[:, :, None] * _dpsi1_dmu).sum(axis=1)
grad_S = (dL_dpsi1[:, :, None] * _dpsi1_dS).sum(axis=1) grad_S = (dL_dpsi1[:, :, None] * _dpsi1_dS).sum(axis=1)
grad_gamma = (dL_dpsi1[:,:,None] * _dpsi1_dgamma).sum(axis=1) grad_gamma = (dL_dpsi1[:,:,None] * _dpsi1_dgamma).sum(axis=1)
#psi2 #psi2
grad_mu += (dL_dpsi2[:, :, :, None] * _dpsi2_dmu).reshape(ndata,-1,self.input_dim).sum(axis=1) grad_mu += (dL_dpsi2[:, :, :, None] * _dpsi2_dmu).reshape(ndata,-1,self.input_dim).sum(axis=1)
grad_S += (dL_dpsi2[:, :, :, None] * _dpsi2_dS).reshape(ndata,-1,self.input_dim).sum(axis=1) grad_S += (dL_dpsi2[:, :, :, None] * _dpsi2_dS).reshape(ndata,-1,self.input_dim).sum(axis=1)
grad_gamma += (dL_dpsi2[:,:,:, None] * _dpsi2_dgamma).reshape(ndata,-1,self.input_dim).sum(axis=1) grad_gamma += (dL_dpsi2[:,:,:, None] * _dpsi2_dgamma).reshape(ndata,-1,self.input_dim).sum(axis=1)
if self.group_spike_prob:
grad_gamma[:] = grad_gamma.mean(axis=0)
print np.abs(gmu-grad_mu).max(),np.abs(gS-grad_S).max(),np.abs(gg-grad_gamma).max() if self.group_spike_prob:
grad_gamma[:] = grad_gamma.mean(axis=0)
return grad_mu, grad_S, grad_gamma
return grad_mu, grad_S, grad_gamma
elif isinstance(variational_posterior, variational.NormalPosterior): elif isinstance(variational_posterior, variational.NormalPosterior):