def test_compare_agents() :
environment = agents.VacuumeEnvironment
agentList = [agents.TableDrivenVacuumAgent,agents.RandomVacuumAgent, agents.ReflexVacuumAgent]
result = agents.compare_agents(environment, agentList,[(1,0),(2,0),(0,0)])
for i in range(0,len(result)):
print("{}'s avg performance: {}".format(result[i][0].__name__,result[i][1]))
#runs a single vacuum environment with a specific agent and step count
class vacuumSim():
def __init__(self):
self.vacuumEnvironment = agents.VacuumeEnvironment()
def startSim(self):
self.vacuumEnvironment.add_thing(agents.TraceAgent(agents.ReflexVacuumAgent()))
#self.vacuumEnvironment.add_thing(agents.TraceAgent(agents.TableDrivenVacuumAgent()))
self.vacuumEnvironment.run(10)
def main():
vacSim=vacuumSim()
vacSim.startSim()
#test_compare_agents() #used to compare multiple agents in the same environment conditions
if __name__ == '__main__': # if we're running file directly and not importing it
main()
def test_compare_agents():
environment = TrivialVacuumEnvironment
agents = [ModelBasedVacuumAgent, ReflexVacuumAgent]
result = compare_agents(environment, agents)
performance_ModelBasedVacummAgent = result[0][1]
performance_ReflexVacummAgent = result[1][1]
# The performance of ModelBasedVacuumAgent will be at least as good as that of
# ReflexVacuumAgent, since ModelBasedVacuumAgent can identify when it has
# reached the terminal state (both locations being clean) and will perform
# NoOp leading to 0 performance change, whereas ReflexVacuumAgent cannot
# identify the terminal state and thus will keep moving, leading to worse
# performance compared to ModelBasedVacuumAgent.
assert performance_ReflexVacummAgent <= performance_ModelBasedVacummAgent
def test_compare_agents() :
environment = TrivialVacuumEnvironment
agents = [ModelBasedVacuumAgent, ReflexVacuumAgent]
result = compare_agents(environment, agents)
performance_ModelBasedVacummAgent = result[0][1]
performance_ReflexVacummAgent = result[1][1]
# The performance of ModelBasedVacuumAgent will be at least as good as that of
# ReflexVacuumAgent, since ModelBasedVacuumAgent can identify when it has
# reached the terminal state (both locations being clean) and will perform
# NoOp leading to 0 performance change, whereas ReflexVacuumAgent cannot
# identify the terminal state and thus will keep moving, leading to worse
# performance compared to ModelBasedVacuumAgent.
assert performance_ReflexVacummAgent <= performance_ModelBasedVacummAgent
loc_A,
loc_B,
ReflexVacuumAgent,
TrivialVacuumEnvironment,
)
from utils import (
mean,
)
# Exercise 2.7
print "\n\nExercise 2.7"
n=100
steps=4
print "Compare ReflexVacuumAgent against {n} TrivialVacuumEnvironment instances and {steps} steps".format(**locals())
print compare_agents(TrivialVacuumEnvironment, [ReflexVacuumAgent], n=n, steps=4)
# Exercise 2.8
# Run the environment simulator with a simple reflex agent for all possible
# initial dirt configurations and agent locations
print "\n\nExercise 2.8"
scores = []
for loc_A_status in ('Clean', 'Dirty'):
for loc_B_status in ('Clean', 'Dirty'):
for agent_loc in (loc_A, loc_B):
agent = ReflexVacuumAgent()
env = TrivialVacuumEnvironment(loc_A_status, loc_B_status)
env.add_thing(agent, agent_loc)
env.run(steps)
scores.append(agent.performance)
print "Environment=({loc_A_status}, {loc_B_status}) and agent at {agent_loc}, agent scored {agent.performance} in {steps} steps".format(**locals())