def p_object_given_feature(self,feature, priors):
probability_object_is_feature_given_object = self.table_to_function()
try:
object_given_features=dist.bayesEvidence(priors, probability_object_is_feature_given_object, feature)
except:
print "feature: ",feature, " not found, will return prior"
object_given_features = priors
return object_given_features
def getNextValues(self, state, inp):
"""
@param state: Distribution over states of the subject machine,
represented as a C{dist.Dist} object
@param inp: A pair C{(o, i)} of the observation (output) and input
of the subject machine on this time step.
"""
(o, i) = inp
if self.model.sensorDisplayFun:
self.model.sensorDisplayFun(o)
sGo = dist.bayesEvidence(state, self.model.observationDistribution, o)
if self.verbose:
print "after obs", o, sGo
dSPrime = dist.totalProbability(sGo, self.model.transitionDistribution(i))
if self.verbose:
print "after trans", i, dSPrime
if self.model.beliefDisplayFun:
self.model.beliefDisplayFun(dSPrime)
return (dSPrime, dSPrime)
def getNextValues(self, state, inp):
"""
@param state: Distribution over states of the subject machine,
represented as a C{dist.Dist} object
@param inp: A pair C{(o, i)} of the observation (output) and input
of the subject machine on this time step.
"""
(o, i) = inp
if self.model.sensorDisplayFun:
self.model.sensorDisplayFun(o)
sGo = dist.bayesEvidence(state, self.model.observationDistribution, o)
if self.verbose:
print 'after obs', o, sGo
dSPrime = dist.totalProbability(sGo,
self.model.transitionDistribution(i))
if self.verbose:
print 'after trans', i, dSPrime
if self.model.beliefDisplayFun:
self.model.beliefDisplayFun(dSPrime)
return (dSPrime, dSPrime)
