apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-04-28 06:16:24 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

2D plots fixed

Browse source
This commit is contained in:
Ricardo 2013-11-07 17:35:41 +00:00
parent d2db4c6688
commit ae6648e0cf
1 changed files with 7 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

17
GPy/core/gp_base.py
View file

@ -37,7 +37,7 @@ class GPBase(Model):
# the end # the end
def posterior_samples_f(self,X,size=10,which_parts='all',full_cov=True): def posterior_samples_f(self,X,size=10,which_parts='all'):
""" """
Samples the posterior GP at the points X. Samples the posterior GP at the points X.
@ -51,16 +51,13 @@ class GPBase(Model):
:type full_cov: bool. :type full_cov: bool.
:returns: Ysim: set of simulations, a Numpy array (N x samples). :returns: Ysim: set of simulations, a Numpy array (N x samples).
""" """
m, v = self._raw_predict(X, which_parts=which_parts, full_cov=full_cov) m, v = self._raw_predict(X, which_parts=which_parts, full_cov=True)
v = v.reshape(m.size,-1) if len(v.shape)==3 else v v = v.reshape(m.size,-1) if len(v.shape)==3 else v
if not full_cov: Ysim = np.random.multivariate_normal(m.flatten(), v, size).T
Ysim = np.random.multivariate_normal(m.flatten(), np.diag(v.flatten()), size).T
else:
Ysim = np.random.multivariate_normal(m.flatten(), v, size).T
return Ysim return Ysim
def posterior_samples(self,X,size=10,which_parts='all',full_cov=True,noise_model=None): def posterior_samples(self,X,size=10,which_parts='all',noise_model=None):
""" """
Samples the posterior GP at the points X. Samples the posterior GP at the points X.
@ -76,7 +73,7 @@ class GPBase(Model):
:type noise_model: integer. :type noise_model: integer.
:returns: Ysim: set of simulations, a Numpy array (N x samples). :returns: Ysim: set of simulations, a Numpy array (N x samples).
""" """
Ysim = self.posterior_samples_f(X, size, which_parts=which_parts, full_cov=full_cov) Ysim = self.posterior_samples_f(X, size, which_parts=which_parts, full_cov=True)
if isinstance(self.likelihood,Gaussian): if isinstance(self.likelihood,Gaussian):
noise_std = np.sqrt(self.likelihood._get_params()) noise_std = np.sqrt(self.likelihood._get_params())
Ysim += np.random.normal(0,noise_std,Ysim.shape) Ysim += np.random.normal(0,noise_std,Ysim.shape)
@ -209,11 +206,11 @@ class GPBase(Model):
x, y = np.linspace(xmin[0], xmax[0], resolution), np.linspace(xmin[1], xmax[1], resolution) x, y = np.linspace(xmin[0], xmax[0], resolution), np.linspace(xmin[1], xmax[1], resolution)
#predict on the frame and plot #predict on the frame and plot
if use_raw_predict: if plot_raw:
m, _ = self._raw_predict(Xgrid, which_parts=which_parts) m, _ = self._raw_predict(Xgrid, which_parts=which_parts)
Y = self.likelihood.Y Y = self.likelihood.Y
else: else:
m, _, _, _ = self.predict(Xgrid, which_parts=which_parts) m, _, _, _ = self.predict(Xgrid, which_parts=which_parts,num_samples=100) #FIXME we need a balance between accuracy and speed to define num_samples
Y = self.likelihood.data Y = self.likelihood.data
for d in which_data_ycols: for d in which_data_ycols:
m_d = m[:,d].reshape(resolution, resolution).T m_d = m[:,d].reshape(resolution, resolution).T