"""Probability models (Chapter 13-15)"""
from collections import defaultdict
from functools import reduce
from agents import Agent
from utils import *
[docs]
def DTAgentProgram(belief_state):
"""
[Figure 13.1]
A decision-theoretic agent.
"""
def program(percept):
belief_state.observe(program.action, percept)
program.action = max(belief_state.actions(), key=belief_state.expected_outcome_utility)
return program.action
program.action = None
return program
# ______________________________________________________________________________
[docs]
class ProbDist:
"""A discrete probability distribution. You name the random variable
in the constructor, then assign and query probability of values.
>>> P = ProbDist('Flip'); P['H'], P['T'] = 0.25, 0.75; P['H']
0.25
>>> P = ProbDist('X', {'lo': 125, 'med': 375, 'hi': 500})
>>> P['lo'], P['med'], P['hi']
(0.125, 0.375, 0.5)
"""
def __init__(self, var_name='?', freq=None):
"""If freq is given, it is a dictionary of values - frequency pairs,
then ProbDist is normalized."""
self.prob = {}
self.var_name = var_name
self.values = []
if freq:
for (v, p) in freq.items():
self[v] = p
self.normalize()
def __getitem__(self, val):
"""Given a value, return P(value)."""
try:
return self.prob[val]
except KeyError:
return 0
def __setitem__(self, val, p):
"""Set P(val) = p."""
if val not in self.values:
self.values.append(val)
self.prob[val] = p
[docs]
def normalize(self):
"""Make sure the probabilities of all values sum to 1.
Returns the normalized distribution.
Raises a ZeroDivisionError if the sum of the values is 0."""
total = sum(self.prob.values())
if not np.isclose(total, 1.0):
for val in self.prob:
self.prob[val] /= total
return self
[docs]
def show_approx(self, numfmt='{:.3g}'):
"""Show the probabilities rounded and sorted by key, for the
sake of portable doctests."""
return ', '.join([('{}: ' + numfmt).format(v, p) for (v, p) in sorted(self.prob.items())])
def __repr__(self):
return "P({})".format(self.var_name)
[docs]
class JointProbDist(ProbDist):
"""A discrete probability distribute over a set of variables.
>>> P = JointProbDist(['X', 'Y']); P[1, 1] = 0.25
>>> P[1, 1]
0.25
>>> P[dict(X=0, Y=1)] = 0.5
>>> P[dict(X=0, Y=1)]
0.5"""
def __init__(self, variables):
self.prob = {}
self.variables = variables
self.vals = defaultdict(list)
def __getitem__(self, values):
"""Given a tuple or dict of values, return P(values)."""
values = event_values(values, self.variables)
return ProbDist.__getitem__(self, values)
def __setitem__(self, values, p):
"""Set P(values) = p. Values can be a tuple or a dict; it must
have a value for each of the variables in the joint. Also keep track
of the values we have seen so far for each variable."""
values = event_values(values, self.variables)
self.prob[values] = p
for var, val in zip(self.variables, values):
if val not in self.vals[var]:
self.vals[var].append(val)
[docs]
def values(self, var):
"""Return the set of possible values for a variable."""
return self.vals[var]
def __repr__(self):
return "P({})".format(self.variables)
[docs]
def event_values(event, variables):
"""Return a tuple of the values of variables in event.
>>> event_values ({'A': 10, 'B': 9, 'C': 8}, ['C', 'A'])
(8, 10)
>>> event_values ((1, 2), ['C', 'A'])
(1, 2)
"""
if isinstance(event, tuple) and len(event) == len(variables):
return event
else:
return tuple([event[var] for var in variables])
# ______________________________________________________________________________
[docs]
def enumerate_joint_ask(X, e, P):
"""
[Section 13.3]
Return a probability distribution over the values of the variable X,
given the {var:val} observations e, in the JointProbDist P.
>>> P = JointProbDist(['X', 'Y'])
>>> P[0,0] = 0.25; P[0,1] = 0.5; P[1,1] = P[2,1] = 0.125
>>> enumerate_joint_ask('X', dict(Y=1), P).show_approx()
'0: 0.667, 1: 0.167, 2: 0.167'
"""
assert X not in e, "Query variable must be distinct from evidence"
Q = ProbDist(X) # probability distribution for X, initially empty
Y = [v for v in P.variables if v != X and v not in e] # hidden variables.
for xi in P.values(X):
Q[xi] = enumerate_joint(Y, extend(e, X, xi), P)
return Q.normalize()
[docs]
def enumerate_joint(variables, e, P):
"""Return the sum of those entries in P consistent with e,
provided variables is P's remaining variables (the ones not in e)."""
if not variables:
return P[e]
Y, rest = variables[0], variables[1:]
return sum([enumerate_joint(rest, extend(e, Y, y), P) for y in P.values(Y)])
# ______________________________________________________________________________
[docs]
class BayesNet:
"""Bayesian network containing only boolean-variable nodes."""
def __init__(self, node_specs=None):
"""Nodes must be ordered with parents before children."""
self.nodes = []
self.variables = []
node_specs = node_specs or []
for node_spec in node_specs:
self.add(node_spec)
[docs]
def add(self, node_spec):
"""Add a node to the net. Its parents must already be in the
net, and its variable must not."""
node = BayesNode(*node_spec)
assert node.variable not in self.variables
assert all((parent in self.variables) for parent in node.parents)
self.nodes.append(node)
self.variables.append(node.variable)
for parent in node.parents:
self.variable_node(parent).children.append(node)
[docs]
def variable_node(self, var):
"""Return the node for the variable named var.
>>> burglary.variable_node('Burglary').variable
'Burglary'"""
for n in self.nodes:
if n.variable == var:
return n
raise Exception("No such variable: {}".format(var))
[docs]
def variable_values(self, var):
"""Return the domain of var."""
return [True, False]
def __repr__(self):
return 'BayesNet({0!r})'.format(self.nodes)
[docs]
class DecisionNetwork(BayesNet):
"""An abstract class for a decision network as a wrapper for a BayesNet.
Represents an agent's current state, its possible actions, reachable states
and utilities of those states."""
def __init__(self, action, infer):
"""action: a single action node
infer: the preferred method to carry out inference on the given BayesNet"""
super(DecisionNetwork, self).__init__()
self.action = action
self.infer = infer
[docs]
def best_action(self):
"""Return the best action in the network"""
return self.action
[docs]
def get_utility(self, action, state):
"""Return the utility for a particular action and state in the network"""
raise NotImplementedError
[docs]
def get_expected_utility(self, action, evidence):
"""Compute the expected utility given an action and evidence"""
u = 0.0
prob_dist = self.infer(action, evidence, self).prob
for item, _ in prob_dist.items():
u += prob_dist[item] * self.get_utility(action, item)
return u
[docs]
class BayesNode:
"""A conditional probability distribution for a boolean variable,
P(X | parents). Part of a BayesNet."""
def __init__(self, X, parents, cpt):
"""X is a variable name, and parents a sequence of variable
names or a space-separated string. cpt, the conditional
probability table, takes one of these forms:
* A number, the unconditional probability P(X=true). You can
use this form when there are no parents.
* A dict {v: p, ...}, the conditional probability distribution
P(X=true | parent=v) = p. When there's just one parent.
* A dict {(v1, v2, ...): p, ...}, the distribution P(X=true |
parent1=v1, parent2=v2, ...) = p. Each key must have as many
values as there are parents. You can use this form always;
the first two are just conveniences.
In all cases the probability of X being false is left implicit,
since it follows from P(X=true).
>>> X = BayesNode('X', '', 0.2)
>>> Y = BayesNode('Y', 'P', {T: 0.2, F: 0.7})
>>> Z = BayesNode('Z', 'P Q',
... {(T, T): 0.2, (T, F): 0.3, (F, T): 0.5, (F, F): 0.7})
"""
if isinstance(parents, str):
parents = parents.split()
# We store the table always in the third form above.
if isinstance(cpt, (float, int)): # no parents, 0-tuple
cpt = {(): cpt}
elif isinstance(cpt, dict):
# one parent, 1-tuple
if cpt and isinstance(list(cpt.keys())[0], bool):
cpt = {(v,): p for v, p in cpt.items()}
assert isinstance(cpt, dict)
for vs, p in cpt.items():
assert isinstance(vs, tuple) and len(vs) == len(parents)
assert all(isinstance(v, bool) for v in vs)
assert 0 <= p <= 1
self.variable = X
self.parents = parents
self.cpt = cpt
self.children = []
[docs]
def p(self, value, event):
"""Return the conditional probability
P(X=value | parents=parent_values), where parent_values
are the values of parents in event. (event must assign each
parent a value.)
>>> bn = BayesNode('X', 'Burglary', {T: 0.2, F: 0.625})
>>> bn.p(False, {'Burglary': False, 'Earthquake': True})
0.375"""
assert isinstance(value, bool)
ptrue = self.cpt[event_values(event, self.parents)]
return ptrue if value else 1 - ptrue
[docs]
def sample(self, event):
"""Sample from the distribution for this variable conditioned
on event's values for parent_variables. That is, return True/False
at random according with the conditional probability given the
parents."""
return probability(self.p(True, event))
def __repr__(self):
return repr((self.variable, ' '.join(self.parents)))
# Burglary example [Figure 14.2]
T, F = True, False
burglary = BayesNet([('Burglary', '', 0.001),
('Earthquake', '', 0.002),
('Alarm', 'Burglary Earthquake',
{(T, T): 0.95, (T, F): 0.94, (F, T): 0.29, (F, F): 0.001}),
('JohnCalls', 'Alarm', {T: 0.90, F: 0.05}),
('MaryCalls', 'Alarm', {T: 0.70, F: 0.01})])
# ______________________________________________________________________________
[docs]
def enumeration_ask(X, e, bn):
"""
[Figure 14.9]
Return the conditional probability distribution of variable X
given evidence e, from BayesNet bn.
>>> enumeration_ask('Burglary', dict(JohnCalls=T, MaryCalls=T), burglary
... ).show_approx()
'False: 0.716, True: 0.284'"""
assert X not in e, "Query variable must be distinct from evidence"
Q = ProbDist(X)
for xi in bn.variable_values(X):
Q[xi] = enumerate_all(bn.variables, extend(e, X, xi), bn)
return Q.normalize()
[docs]
def enumerate_all(variables, e, bn):
"""Return the sum of those entries in P(variables | e{others})
consistent with e, where P is the joint distribution represented
by bn, and e{others} means e restricted to bn's other variables
(the ones other than variables). Parents must precede children in variables."""
if not variables:
return 1.0
Y, rest = variables[0], variables[1:]
Ynode = bn.variable_node(Y)
if Y in e:
return Ynode.p(e[Y], e) * enumerate_all(rest, e, bn)
else:
return sum(Ynode.p(y, e) * enumerate_all(rest, extend(e, Y, y), bn)
for y in bn.variable_values(Y))
# ______________________________________________________________________________
[docs]
def elimination_ask(X, e, bn):
"""
[Figure 14.11]
Compute bn's P(X|e) by variable elimination.
>>> elimination_ask('Burglary', dict(JohnCalls=T, MaryCalls=T), burglary
... ).show_approx()
'False: 0.716, True: 0.284'"""
assert X not in e, "Query variable must be distinct from evidence"
factors = []
for var in reversed(bn.variables):
factors.append(make_factor(var, e, bn))
if is_hidden(var, X, e):
factors = sum_out(var, factors, bn)
return pointwise_product(factors, bn).normalize()
[docs]
def is_hidden(var, X, e):
"""Is var a hidden variable when querying P(X|e)?"""
return var != X and var not in e
[docs]
def make_factor(var, e, bn):
"""Return the factor for var in bn's joint distribution given e.
That is, bn's full joint distribution, projected to accord with e,
is the pointwise product of these factors for bn's variables."""
node = bn.variable_node(var)
variables = [X for X in [var] + node.parents if X not in e]
cpt = {event_values(e1, variables): node.p(e1[var], e1)
for e1 in all_events(variables, bn, e)}
return Factor(variables, cpt)
[docs]
def pointwise_product(factors, bn):
"""Multiply a sequence of factors together into a single factor over the union of their
variables, using the Bayes net ``bn`` to enumerate variable values."""
return reduce(lambda f, g: f.pointwise_product(g, bn), factors)
[docs]
def sum_out(var, factors, bn):
"""Eliminate var from all factors by summing over its values."""
result, var_factors = [], []
for f in factors:
(var_factors if var in f.variables else result).append(f)
result.append(pointwise_product(var_factors, bn).sum_out(var, bn))
return result
[docs]
class Factor:
"""A factor in a joint distribution."""
def __init__(self, variables, cpt):
self.variables = variables
self.cpt = cpt
[docs]
def pointwise_product(self, other, bn):
"""Multiply two factors, combining their variables."""
variables = list(set(self.variables) | set(other.variables))
cpt = {event_values(e, variables): self.p(e) * other.p(e) for e in all_events(variables, bn, {})}
return Factor(variables, cpt)
[docs]
def sum_out(self, var, bn):
"""Make a factor eliminating var by summing over its values."""
variables = [X for X in self.variables if X != var]
cpt = {event_values(e, variables): sum(self.p(extend(e, var, val)) for val in bn.variable_values(var))
for e in all_events(variables, bn, {})}
return Factor(variables, cpt)
[docs]
def normalize(self):
"""Return my probabilities; must be down to one variable."""
assert len(self.variables) == 1
return ProbDist(self.variables[0], {k: v for ((k,), v) in self.cpt.items()})
[docs]
def p(self, e):
"""Look up my value tabulated for e."""
return self.cpt[event_values(e, self.variables)]
[docs]
def all_events(variables, bn, e):
"""Yield every way of extending e with values for all variables."""
if not variables:
yield e
else:
X, rest = variables[0], variables[1:]
for e1 in all_events(rest, bn, e):
for x in bn.variable_values(X):
yield extend(e1, X, x)
# ______________________________________________________________________________
# [Figure 14.12a]: sprinkler network
sprinkler = BayesNet([('Cloudy', '', 0.5),
('Sprinkler', 'Cloudy', {T: 0.10, F: 0.50}),
('Rain', 'Cloudy', {T: 0.80, F: 0.20}),
('WetGrass', 'Sprinkler Rain',
{(T, T): 0.99, (T, F): 0.90, (F, T): 0.90, (F, F): 0.00})])
# ______________________________________________________________________________
[docs]
def prior_sample(bn):
"""
[Figure 14.13]
Randomly sample from bn's full joint distribution.
The result is a {variable: value} dict.
"""
event = {}
for node in bn.nodes:
event[node.variable] = node.sample(event)
return event
# _________________________________________________________________________
[docs]
def rejection_sampling(X, e, bn, N=10000):
"""
[Figure 14.14]
Estimate the probability distribution of variable X given
evidence e in BayesNet bn, using N samples.
Raises a ZeroDivisionError if all the N samples are rejected,
i.e., inconsistent with e.
>>> random.seed(47)
>>> rejection_sampling('Burglary', dict(JohnCalls=T, MaryCalls=T),
... burglary, 10000).show_approx()
'False: 0.7, True: 0.3'
"""
counts = {x: 0 for x in bn.variable_values(X)} # bold N in [Figure 14.14]
for j in range(N):
sample = prior_sample(bn) # boldface x in [Figure 14.14]
if consistent_with(sample, e):
counts[sample[X]] += 1
return ProbDist(X, counts)
[docs]
def consistent_with(event, evidence):
"""Is event consistent with the given evidence?"""
return all(evidence.get(k, v) == v for k, v in event.items())
# _________________________________________________________________________
[docs]
def likelihood_weighting(X, e, bn, N=10000):
"""
[Figure 14.15]
Estimate the probability distribution of variable X given
evidence e in BayesNet bn.
>>> random.seed(1017)
>>> likelihood_weighting('Burglary', dict(JohnCalls=T, MaryCalls=T),
... burglary, 10000).show_approx()
'False: 0.702, True: 0.298'
"""
W = {x: 0 for x in bn.variable_values(X)}
for j in range(N):
sample, weight = weighted_sample(bn, e) # boldface x, w in [Figure 14.15]
W[sample[X]] += weight
return ProbDist(X, W)
[docs]
def weighted_sample(bn, e):
"""
Sample an event from bn that's consistent with the evidence e;
return the event and its weight, the likelihood that the event
accords to the evidence.
"""
w = 1
event = dict(e) # boldface x in [Figure 14.15]
for node in bn.nodes:
Xi = node.variable
if Xi in e:
w *= node.p(e[Xi], event)
else:
event[Xi] = node.sample(event)
return event, w
# _________________________________________________________________________
[docs]
def gibbs_ask(X, e, bn, N=1000):
"""[Figure 14.16]"""
assert X not in e, "Query variable must be distinct from evidence"
counts = {x: 0 for x in bn.variable_values(X)} # bold N in [Figure 14.16]
Z = [var for var in bn.variables if var not in e]
state = dict(e) # boldface x in [Figure 14.16]
for Zi in Z:
state[Zi] = random.choice(bn.variable_values(Zi))
for j in range(N):
for Zi in Z:
state[Zi] = markov_blanket_sample(Zi, state, bn)
counts[state[X]] += 1
return ProbDist(X, counts)
[docs]
def markov_blanket_sample(X, e, bn):
"""Return a sample from P(X | mb) where mb denotes that the
variables in the Markov blanket of X take their values from event
e (which must assign a value to each). The Markov blanket of X is
X's parents, children, and children's parents."""
Xnode = bn.variable_node(X)
Q = ProbDist(X)
for xi in bn.variable_values(X):
ei = extend(e, X, xi)
# [Equation 14.12]
Q[xi] = Xnode.p(xi, e) * product(Yj.p(ei[Yj.variable], ei) for Yj in Xnode.children)
# (assuming a Boolean variable here)
return probability(Q.normalize()[True])
# _________________________________________________________________________
[docs]
class HiddenMarkovModel:
"""A Hidden markov model which takes Transition model and Sensor model as inputs"""
def __init__(self, transition_model, sensor_model, prior=None):
self.transition_model = transition_model
self.sensor_model = sensor_model
self.prior = prior or [0.5, 0.5]
[docs]
def sensor_dist(self, ev):
"""Return the sensor (observation) distribution corresponding to the evidence ``ev``:
the first row of the sensor model when ``ev`` is True, otherwise the second."""
if ev is True:
return self.sensor_model[0]
else:
return self.sensor_model[1]
[docs]
def forward(HMM, fv, ev):
"""Perform one forward (filtering) step of an HMM: project the forward message ``fv``
through the transition model, weight it by the sensor distribution for evidence ``ev``,
and return the normalized next forward message. [Figure 15.4]"""
prediction = vector_add(scalar_vector_product(fv[0], HMM.transition_model[0]),
scalar_vector_product(fv[1], HMM.transition_model[1]))
sensor_dist = HMM.sensor_dist(ev)
return normalize(element_wise_product(sensor_dist, prediction))
[docs]
def backward(HMM, b, ev):
"""Perform one backward step of an HMM: weight the backward message ``b`` by the sensor
distribution for evidence ``ev`` and propagate it through the transition model, returning
the normalized previous backward message. [Figure 15.4]"""
sensor_dist = HMM.sensor_dist(ev)
prediction = element_wise_product(sensor_dist, b)
return normalize(vector_add(scalar_vector_product(prediction[0], HMM.transition_model[0]),
scalar_vector_product(prediction[1], HMM.transition_model[1])))
[docs]
def forward_backward(HMM, ev):
"""
[Figure 15.4]
Forward-Backward algorithm for smoothing. Computes posterior probabilities
of a sequence of states given a sequence of observations.
"""
t = len(ev)
ev.insert(0, None) # to make the code look similar to pseudo code
fv = [[0.0, 0.0] for _ in range(len(ev))]
b = [1.0, 1.0]
sv = [[0, 0] for _ in range(len(ev))]
fv[0] = HMM.prior
for i in range(1, t + 1):
fv[i] = forward(HMM, fv[i - 1], ev[i])
for i in range(t, -1, -1):
sv[i - 1] = normalize(element_wise_product(fv[i], b))
b = backward(HMM, b, ev[i])
sv = sv[::-1]
return sv
[docs]
def viterbi(HMM, ev):
"""
[Equation 15.11]
Viterbi algorithm to find the most likely sequence. Computes the best path and the
corresponding probabilities, given an HMM model and a sequence of observations.
"""
t = len(ev)
ev = ev.copy()
ev.insert(0, None)
m = [[0.0, 0.0] for _ in range(len(ev) - 1)]
# the recursion is initialized with m1 = forward(P(X0), e1)
m[0] = forward(HMM, HMM.prior, ev[1])
# keep track of maximizing predecessors
backtracking_graph = []
for i in range(1, t):
m[i] = element_wise_product(HMM.sensor_dist(ev[i + 1]),
[max(element_wise_product(HMM.transition_model[0], m[i - 1])),
max(element_wise_product(HMM.transition_model[1], m[i - 1]))])
backtracking_graph.append([np.argmax(element_wise_product(HMM.transition_model[0], m[i - 1])),
np.argmax(element_wise_product(HMM.transition_model[1], m[i - 1]))])
# computed probabilities
ml_probabilities = [0.0] * (len(ev) - 1)
# most likely sequence
ml_path = [True] * (len(ev) - 1)
# the construction of the most likely sequence starts in the final state with the largest probability, and
# runs backwards; the algorithm needs to store for each xt its predecessor xt-1 maximizing its probability
i_max = np.argmax(m[-1])
for i in range(t - 1, -1, -1):
ml_probabilities[i] = m[i][i_max]
ml_path[i] = True if i_max == 0 else False
if i > 0:
i_max = backtracking_graph[i - 1][i_max]
return ml_path, ml_probabilities
[docs]
def baum_welch(HMM, observations, iterations=100):
"""
[Section 20.3]
Baum-Welch algorithm: the instance of EM that learns the parameters of a
Hidden Markov Model (transition model, sensor model and prior) from a single
sequence of boolean 'observations', starting from the initial guess in 'HMM'.
Each iteration runs a (scaled) forward-backward pass to compute the smoothed
state marginals gamma_t(i) = P(X_t=i | e_1:T) and transition marginals
xi_t(i,j) = P(X_t=i, X_t+1=j | e_1:T) (E-step), then re-estimates every
parameter as the corresponding normalized expected count (M-step)::
prior_i = gamma_0(i)
A_ij = sum_t xi_t(i, j) / sum_t gamma_t(i)
sensor_oi = sum_{t: e_t = o} gamma_t(i) / sum_t gamma_t(i)
Returns a new HiddenMarkovModel with the learned parameters.
"""
A = np.array(HMM.transition_model, dtype=float)
prior = np.array(HMM.prior, dtype=float)
# sensor[0] = P(e=True | state), sensor[1] = P(e=False | state)
sensor = np.array(HMM.sensor_model, dtype=float)
obs = list(observations)
n, t_max = len(prior), len(obs)
for _ in range(iterations):
# emission vectors b_t(i) = P(e_t | X_t = i), recomputed from current sensor
B = np.array([sensor[0] if e else sensor[1] for e in obs])
# E-step: scaled forward (alpha) and backward (beta) messages
alpha, c = np.zeros((t_max, n)), np.zeros(t_max)
alpha[0] = prior * B[0]
c[0] = alpha[0].sum()
alpha[0] /= c[0]
for t in range(1, t_max):
alpha[t] = B[t] * (alpha[t - 1] @ A)
c[t] = alpha[t].sum()
alpha[t] /= c[t]
beta = np.zeros((t_max, n))
beta[-1] = 1
for t in range(t_max - 2, -1, -1):
beta[t] = (A @ (B[t + 1] * beta[t + 1])) / c[t + 1]
# smoothed state and transition marginals (normalized, so the per-step
# scaling factors cancel out)
gamma = alpha * beta
gamma /= gamma.sum(axis=1, keepdims=True)
xi = np.zeros((t_max - 1, n, n))
for t in range(t_max - 1):
xi[t] = alpha[t][:, None] * A * B[t + 1] * beta[t + 1]
xi[t] /= xi[t].sum()
# M-step: re-estimate every parameter from the expected counts
prior = gamma[0]
A = xi.sum(axis=0) / gamma[:-1].sum(axis=0)[:, None]
mask = np.array(obs, dtype=bool)
p_true = gamma[mask].sum(axis=0) / gamma.sum(axis=0)
sensor = np.array([p_true, 1 - p_true])
return HiddenMarkovModel(A.tolist(), sensor.tolist(), prior.tolist())
# _________________________________________________________________________
[docs]
def fixed_lag_smoothing(e_t, HMM, d, ev, t):
"""
[Figure 15.6]
Smoothing algorithm with a fixed time lag of 'd' steps.
Computes the smoothed estimate P(X_{t-d} | e_{1:t}) for the slice that lies
'd' steps in the past, given the evidence sequence ev = [e_1, ..., e_t].
Returns None when there is not yet enough evidence (t <= d).
"""
if t <= d:
return None
T_model = np.array(HMM.transition_model)
# forward message advanced over e_1 .. e_{t-d}
f = HMM.prior
for i in range(t - d):
f = forward(HMM, f, ev[i])
# backward transformation accumulated over the lag window e_{t-d+1} .. e_t
B = np.eye(len(f))
for i in range(t - d, t):
O_i = np.diag(HMM.sensor_dist(ev[i]))
B = B @ T_model @ O_i
return normalize((np.array(f) * (B @ np.ones(len(f)))).tolist())
# _________________________________________________________________________
[docs]
def particle_filtering(e, N, HMM):
"""Particle filtering considering two states variables."""
dist = [0.5, 0.5]
# Weight Initialization
w = [0 for _ in range(N)]
# STEP 1
# Propagate one step using transition model given prior state
dist = vector_add(scalar_vector_product(dist[0], HMM.transition_model[0]),
scalar_vector_product(dist[1], HMM.transition_model[1]))
# Assign state according to probability
s = ['A' if probability(dist[0]) else 'B' for _ in range(N)]
w_tot = 0
# Calculate importance weight given evidence e
for i in range(N):
if s[i] == 'A':
# P(U|A)*P(A)
w_i = HMM.sensor_dist(e)[0] * dist[0]
if s[i] == 'B':
# P(U|B)*P(B)
w_i = HMM.sensor_dist(e)[1] * dist[1]
w[i] = w_i
w_tot += w_i
# Normalize all the weights
for i in range(N):
w[i] = w[i] / w_tot
# Limit weights to 4 digits
for i in range(N):
w[i] = float("{0:.4f}".format(w[i]))
# STEP 2
s = weighted_sample_with_replacement(N, s, w)
return s
# _________________________________________________________________________
[docs]
class KalmanFilter:
"""
[Section 15.4]
Kalman filter for a linear-Gaussian dynamical system. The hidden state
evolves and is observed according to the linear-Gaussian model
x_{t+1} = F x_t + noise, noise ~ N(0, Sigma_x) (transition model)
z_t = H x_t + noise, noise ~ N(0, Sigma_z) (sensor model)
where F is the transition matrix, H the sensor matrix, Sigma_x the
transition (process) noise covariance and Sigma_z the sensor (measurement)
noise covariance. Because the family of Gaussians is closed under the
Bayesian filtering update, the forward message stays Gaussian and is fully
described by a mean vector and a covariance matrix at every step.
"""
def __init__(self, transition_model, sensor_model, transition_noise, sensor_noise):
self.F = np.atleast_2d(transition_model) # transition matrix
self.H = np.atleast_2d(sensor_model) # sensor matrix
self.Sigma_x = np.atleast_2d(transition_noise) # transition noise covariance
self.Sigma_z = np.atleast_2d(sensor_noise) # sensor noise covariance
[docs]
def predict(self, mean, cov):
"""Time update: project the Gaussian estimate one step forward through F."""
mean = self.F @ mean
cov = self.F @ cov @ self.F.T + self.Sigma_x
return mean, cov
[docs]
def update(self, mean, cov, z):
"""Measurement update: condition the predicted Gaussian on observation z."""
# Kalman gain [Equation 15.21]
K = cov @ self.H.T @ np.linalg.inv(self.H @ cov @ self.H.T + self.Sigma_z)
mean = mean + K @ (np.atleast_1d(z) - self.H @ mean)
cov = (np.eye(cov.shape[0]) - K @ self.H) @ cov
return mean, cov
[docs]
def filter(self, mean, cov, z):
"""One predict-then-update cycle for a single new observation z."""
mean, cov = self.predict(mean, cov)
return self.update(mean, cov, z)
[docs]
def kalman_filter(KF, mean0, cov0, observations):
"""
[Section 15.4]
Run the Kalman filter 'KF' over a sequence of 'observations', starting from
the Gaussian prior N(mean0, cov0). Returns, for each time step, the filtered
Gaussian estimate as a (mean, covariance) pair.
"""
mean, cov = np.atleast_1d(mean0).astype(float), np.atleast_2d(cov0).astype(float)
estimates = []
for z in observations:
mean, cov = KF.filter(mean, cov, z)
estimates.append((mean, cov))
return estimates
# _________________________________________________________________________
[docs]
class DynamicBayesNet:
"""
[Section 15.5]
A dynamic Bayesian network for a stationary first-order Markov process. It is
specified by a prior network over the state variables at slice 0 and a single
transition + sensor network describing, for one time step, the distribution of
each state variable (given the previous slice) and of each evidence variable
(given the current slice). The DBN can be 'unrolled' into an ordinary BayesNet
spanning any number of slices and then queried with the exact inference
algorithms; in particular filtering is the query for the last state variable
given the whole evidence sequence.
Each spec is a (variable, parents, cpt) triple as for a BayesNode. In a
transition spec, a parent named '<var>_prev' refers to state variable <var> at
the previous slice; every other parent refers to the current slice.
"""
def __init__(self, prior, transition, sensors):
self.prior = prior
self.transition = transition
self.sensors = sensors
self.state_variables = [spec[0] for spec in prior]
self.evidence_variables = [spec[0] for spec in sensors]
@staticmethod
def _rename(parents, t, t_prev):
"""Map the parent names of a slice template to concrete unrolled names."""
if isinstance(parents, str):
parents = parents.split()
return [f'{p[:-len("_prev")]}_{t_prev}' if p.endswith('_prev') else f'{p}_{t}' for p in parents]
[docs]
def unroll(self, steps):
"""Unroll the DBN into a BayesNet over slices 0..steps (evidence at 1..steps)."""
specs = [(f'{var}_0', self._rename(parents, 0, 0), cpt) for var, parents, cpt in self.prior]
for t in range(1, steps + 1):
for var, parents, cpt in self.transition + self.sensors:
specs.append((f'{var}_{t}', self._rename(parents, t, t - 1), cpt))
return BayesNet(specs)
[docs]
def filter(self, evidence, query, infer=elimination_ask):
"""
Filtering: the posterior over 'query' at the last slice given the whole
observation sequence. 'evidence' is a list of dicts, one per time step
t = 1, 2, ..., each mapping evidence variables to their observed values.
"""
steps = len(evidence)
net = self.unroll(steps)
e = {f'{var}_{t}': val for t, obs in enumerate(evidence, 1) for var, val in obs.items()}
return infer(f'{query}_{steps}', e, net)
# _________________________________________________________________________
# TODO: Implement continuous map for MonteCarlo similar to Fig25.10 from the book
[docs]
class MCLmap:
"""Map which provides probability distributions and sensor readings.
Consists of discrete cells which are either an obstacle or empty"""
def __init__(self, m):
self.m = m
self.nrows = len(m)
self.ncols = len(m[0])
# list of empty spaces in the map
self.empty = [(i, j) for i in range(self.nrows) for j in range(self.ncols) if not m[i][j]]
[docs]
def sample(self):
"""Returns a random kinematic state possible in the map"""
pos = random.choice(self.empty)
# 0N 1E 2S 3W
orient = random.choice(range(4))
kin_state = pos + (orient,)
return kin_state
[docs]
def ray_cast(self, sensor_num, kin_state):
"""Returns distance to nearest obstacle or map boundary in the direction of sensor"""
pos = kin_state[:2]
orient = kin_state[2]
# sensor layout when orientation is 0 (towards North)
# 0
# 3R1
# 2
delta = ((sensor_num % 2 == 0) * (sensor_num - 1), (sensor_num % 2 == 1) * (2 - sensor_num))
# sensor direction changes based on orientation
for _ in range(orient):
delta = (delta[1], -delta[0])
range_count = 0
while 0 <= pos[0] < self.nrows and 0 <= pos[1] < self.nrows and not self.m[pos[0]][pos[1]]:
pos = vector_add(pos, delta)
range_count += 1
return range_count
[docs]
def monte_carlo_localization(a, z, N, P_motion_sample, P_sensor, m, S=None):
"""
[Figure 25.9]
Monte Carlo localization algorithm
"""
def ray_cast(sensor_num, kin_state, m):
return m.ray_cast(sensor_num, kin_state)
M = len(z)
S_ = [0] * N
W_ = [0] * N
v = a['v']
w = a['w']
if S is None:
S = [m.sample() for _ in range(N)]
for i in range(N):
S_[i] = P_motion_sample(S[i], v, w)
W_[i] = 1
for j in range(M):
z_ = ray_cast(j, S_[i], m)
W_[i] = W_[i] * P_sensor(z[j], z_)
S = weighted_sample_with_replacement(N, S_, W_)
return S