def propose(self):
# Propose new values using normal distribution
if self.proposal_distribution == "Normal":
# New normal deviate, centred on current value
new_val = rnormal(self.stochastic.value, self.adaptive_scale_factor * self.proposal_sd)
# Round before setting proposed value
self.stochastic.value = round_array(new_val)
elif self.proposal_distribution == "Poisson":
k = shape(self.stochastic.value)
# Add or subtract (equal probability) Poisson sample
new_val = self.stochastic.value + rpoisson(self.adaptive_scale_factor * self.proposal_sd) * (-ones(k))**(random(k)>0.5)
if self._positive:
# Enforce positive values
self.stochastic.value = abs(new_val)
else:
self.stochastic.value = new_val
elif self.proposal_distribution == "Prior":
self.stochastic.random()
def propose(self):
# Propose new values using normal distribution
if self.proposal_distribution == "Normal":
# New normal deviate, centred on current value
new_val = rnormal(self.stochastic.value, self.adaptive_scale_factor * self.proposal_sd)
# Round before setting proposed value
self.stochastic.value = round_array(new_val)
elif self.proposal_distribution == "Poisson":
k = shape(self.stochastic.value)
# Add or subtract (equal probability) Poisson sample
new_val = self.stochastic.value + rpoisson(self.adaptive_scale_factor * self.proposal_sd) * (-ones(k))**(random(k)>0.5)
if self._positive:
# Enforce positive values
self.stochastic.value = abs(new_val)
else:
self.stochastic.value = new_val
elif self.proposal_distribution == "Prior":
self.stochastic.random()
def propose(self):
"""
This method proposes values for stochastics based on the empirical
covariance of the values sampled so far.
The proposal jumps are drawn from a multivariate normal distribution.
"""
arrayjump = np.dot(self.proposal_sd, np.random.normal(size=self.proposal_sd.shape[0]))
if self.verbose > 2:
print 'Jump :', arrayjump
# Update each stochastic individually.
for stochastic in self.stochastics:
jump = arrayjump[self._slices[stochastic]]
if np.iterable(stochastic.value):
jump = np.reshape(arrayjump[self._slices[stochastic]],np.shape(stochastic.value))
if self.isdiscrete[stochastic]:
jump = round_array(jump)
stochastic.value = stochastic.value + jump
def propose(self):
"""
This method proposes values for stochastics based on the empirical
covariance of the values sampled so far.
The proposal jumps are drawn from a multivariate normal distribution.
"""
arrayjump = np.dot(self.proposal_sd, np.random.normal(size=self.proposal_sd.shape[0]))
if self.verbose > 2:
print 'Jump :', arrayjump
# Update each stochastic individually.
for stochastic in self.stochastics:
jump = arrayjump[self._slices[stochastic]]
if np.iterable(stochastic.value):
jump = np.reshape(arrayjump[self._slices[stochastic]],np.shape(stochastic.value))
if self.isdiscrete[stochastic]:
jump = round_array(jump)
stochastic.value = stochastic.value + jump
def _set_stochastics(self, p):
for stochastic in self.stochastics:
if self.stochastic_type_dict[stochastic] is int:
stochastic.value = round_array(reshape(ravel(p)[self._slices[stochastic]],shape(stochastic.value)))
else:
stochastic.value = reshape(ravel(p)[self._slices[stochastic]],shape(stochastic.value))