def dpop_petcu():
comm = infrastructure.communication.InProcessCommunicationLayer()
# Variables definition:
x0 = Variable('x0', ['a', 'b', 'c'])
x1 = Variable('x1', ['a', 'b', 'c'])
x2 = Variable('x2', ['a', 'b', 'c'])
x3 = Variable('x3', ['a', 'b', 'c'])
# relation between variables
r1_0 = relations.NAryMatrixRelation([x1, x0],
[[2, 2, 3], [5, 3, 7], [6, 3, 1]])
r2_1 = relations.NAryMatrixRelation([x2, x1],
[[0, 2, 1], [3, 4, 6], [5, 2, 5]])
r3_1 = relations.NAryMatrixRelation([x3, x1],
[[6, 2, 3], [3, 3, 2], [4, 4, 1]])
al0 = DpopAlgo(x0)
al1 = DpopAlgo(x1)
al2 = DpopAlgo(x2)
al3 = DpopAlgo(x3)
al0.add_child(x1)
al1.add_child(x2)
al1.add_child(x3)
al2.set_parent(x1)
al2.add_relation(r2_1)
al3.set_parent(x1)
al3.add_relation(r3_1)
al1.set_parent(x0)
al1.add_relation(r1_0)
a0 = Agent('a0', comm)
a0.add_computation(al0)
a1 = Agent('a1', comm)
a1.add_computation(al1)
a2 = Agent('a2', comm)
a2.add_computation(al2)
a3 = Agent('a3', comm)
a3.add_computation(al3)
results, _, _ = infrastructure.synchronous_single_run([a0, a1, a2, a3])
if results == {'x0': 'a', 'x1': 'c', 'x2': 'b', 'x3': 'a'}:
logging.info('SUCCESS !! ')
return 0
else:
logging.info('invalid result found, needs some debugging ...' +
str(results))
return 1
def graph_coloring_no_prefs():
# Extremely simple graph coloring problem
# Three variables, 2 constraints, no cycle
# modelled as 3 variables and two factors, each variable and each factor
# has its own agent.
# This problem has a strong symmetry and several optimal solutions:
# applying maxsum does not allow to make a choice and each variable keeps
# its full domain available, which means that for any value di from
# domain Di of variable xi, there exists an optimal assignment in which
# xi take di.
#
# To select a solution, an extra step would be needed : select a value
# for one variable and walk the tree to get the value for other variables
# induced by this first choice.
d = ['R', 'G']
v1 = Variable('v1', d)
v2 = Variable('v2', d)
v3 = Variable('v3', d)
# cost table for the factor, simply reflects the fact that neighbor
# variables shoud not have the same value
BIN_DIFF_TABLE_2 = [[1, 0], [0, 1]]
# Factor between v1 and v2
f12 = relations.NAryMatrixRelation([v1, v2],
name='f12',
matrix=BIN_DIFF_TABLE_2)
# Factor between v2 and v3
f23 = relations.NAryMatrixRelation([v2, v3],
name='f23',
matrix=BIN_DIFF_TABLE_2)
# Map the factor graph to agents and solve it
node_agents = distribue_agent_for_all([v1, v2, v3], [f12, f23])
# this sample does not work with our standard run method :
# no factor is a leaf, which means that no start message is sent !
# We need to initiate messages manually
for n in node_agents:
for a in n.computations:
if hasattr(a, '_init_msg'):
a._init_msg()
res, _, _ = infrastructure.synchronous_single_run(node_agents)
print("FINISHED ! " + str(res))
def graph_coloring_with_prefs():
# In this setup, we introduce preferences for each variable
# as we are minimizing, we express preferences as cost (with lower cost
# for preferred value)
# v1 prefers 'R' p(v1) = [ -0.1, +0.1]
# v2 and v3 prefer 'G' p(v2) = p(v3) = [ +0.1, -0.1]
# With This preferences, MaxSum now correctly selects the optimal assignment
# v1 = R, v2 = G , v3 = R
d = ['R', 'G']
v1 = Variable('v1', d)
v2 = Variable('v2', d)
v3 = Variable('v3', d)
# Factor (cost) between v1 and v2
# f12 = (x1 == x2) + p(x1) + p(x2)
# where (x1 == x2) is 1 if x1 equals x2 and 0 otherwise
BIN_DIFF_TABLE_PREFS_12 = [[1, 0.2], [-0.1, 1]]
f12 = relations.NAryMatrixRelation([v1, v2],
name='f12',
matrix=BIN_DIFF_TABLE_PREFS_12)
# f12 = maxsum.BinaryValueTable('f12', v1, v2, BIN_DIFF_TABLE_PREFS_12)
# Factor between v2 and v3
BIN_DIFF_TABLE_PREFS_23 = [[1.2, 0], [0, 0.8]]
f23 = relations.NAryMatrixRelation([v2, v3],
name='f23',
matrix=BIN_DIFF_TABLE_PREFS_23)
# f23 = maxsum.BinaryValueTable('f23', v2, v3, BIN_DIFF_TABLE_PREFS_23)
# Map the factor graph to agents and solve it
node_agents = distribue_agent_for_all([v1, v2, v3], [f12, f23])
# this sample does not work with our standard run method :
# no factor is a leaf, which means that no start message is sent !
# We need to initiate messages manually
for n in node_agents:
for a in n.computations:
if hasattr(a, '_init_msg'):
a._init_msg()
res, _, _ = infrastructure.synchronous_single_run(node_agents)
print("FINISHED ! " + str(res))
def maxsum_smartlights_multiplecomputationagent_costvariable():
"""
This sample use variable with integrated cost.
* 3 lights l1, l2 & l3
each light can have a luminosity level in the 0-9 range
The energy cost is a linear function of the luminosity level, with l1
more efficient than l2 and l3
* one scene action y1, the room luminosity level
y1 = (l1 + l2 + l3 )/3
y1 domain is also between 0-9
* one rule :
l3 must be off AND y1 Must be 5
"""
# building the Factor Graph
# Lights :
l1 = VariableWithCostFunc('l1', list(range(10)), lambda x: x * 0.5)
l2 = VariableWithCostFunc('l2', list(range(10)), lambda x: x)
l3 = VariableWithCostFunc('l3', list(range(10)), lambda x: x)
# Scene action
y1 = Variable('y1', list(range(10)))
@relations.AsNAryFunctionRelation(l1, l2, l3, y1)
def scene_rel(l1, l2, l3, y1):
if y1 == round(l1 / 3.0 + l2 / 3.0 + l3 / 3.0):
return 0
return INFNT
# Rule
@relations.AsNAryFunctionRelation(l3, y1)
def rule_rel(l3, y1):
"""
This rule means : target luminosity if 5, and x3 is off.
:param x3:
:param y1:
:return:
"""
return 10 * (abs(y1 - 5) + l3)
# Create computation for factors and variables
# Light 1
algo_l1 = amaxsum.VariableAlgo(l1, [scene_rel.name])
# Light 2
algo_l2 = amaxsum.VariableAlgo(l2, [scene_rel.name])
# Light 3
algo_l3 = amaxsum.VariableAlgo(l3, [scene_rel.name, rule_rel.name])
# Scene
algo_y1 = amaxsum.VariableAlgo(y1, [rule_rel.name, scene_rel.name])
algo_scene = amaxsum.FactorAlgo(scene_rel)
# Rule
algo_rule = amaxsum.FactorAlgo(rule_rel)
# Distribution of the computation on the three physical light-bulb nodes.
# We have 9 computations to distribute on 3 agents, mapping the 3 light
# bulbs.
comm = infrastructure.communication.InProcessCommunicationLayer()
a1 = infrastructure.Agent('Bulb1', comm)
a1.add_computation(algo_l1)
a1.add_computation(algo_scene)
a1.add_computation(algo_y1)
a2 = infrastructure.Agent('Bulb2', comm)
a2.add_computation(algo_l2)
a3 = infrastructure.Agent('Bulb3', comm)
a3.add_computation(algo_l3)
a3.add_computation(algo_rule)
dcop_agents = [a1, a2, a3]
results, _, _ = infrastructure.synchronous_single_run(dcop_agents, 5)
print(results)
if results == {'l1': 9, 'y1': 5, 'l3': 0, 'l2': 5}:
logging.info('SUCCESS !! ')
return 0
else:
logging.info('invalid result found, needs some debugging ...' +
str(results))
return 1
type: intention
function: 1 if v1 == v2 else 0
diff_2_3:
type: intention
function: 1 if v3 == v2 else 0
agents:
a1:
capacity: 100
a2:
capacity: 100
a3:
capacity: 100
a4:
capacity: 100
a5:
capacity: 100
"""
dcop = load_dcop(dcop_yaml)
cg = build_computation_graph(dcop)
mapping = distribute(cg, dcop.agents)
agents = deploy_on_local_agents(cg, mapping)
results, _, _ = synchronous_single_run(agents)
print("FINISHED ! " + str(results))