def controller(x: simulator.State, dt: float, time: float):
global active_parking_spot, recalculate
running = True
for event in pygame.event.get():
if event.type == pygame.QUIT: running = False
if event.type == pygame.KEYDOWN:
if event.key == pygame.K_0:
active_parking_spot = 0
elif event.key == pygame.K_1:
active_parking_spot = 1
elif event.key == pygame.K_2:
active_parking_spot = 2
elif event.key == pygame.K_3:
active_parking_spot = 3
elif event.key == pygame.K_q:
running = False
recalculate = True
print(f"time: {time:.1f}")
if car_is_in_box(x, p_list[active_parking_spot]):
if simulator.is_zero(x.speed):
active_parking_spot += 1
if active_parking_spot == len(p_list): active_parking_spot = 0
visualizer.update(x.position.x, x.position.y, x.orientation, active_parking_spot)
recalculate = True if ((not recalculate and simulator.is_zero(x.speed)) or time % 10 < 0.1) else False
cs = mpc.get_control_signals(x, p_list[active_parking_spot], 0.3, recalculate=recalculate)
if simulation_time < time or not running:
x.position.x, x.position.y, x.orientation = 0, 0, 0
simulator.run(replay)
cs_s.append(cs)
return cs
#Output file name
OUTFILENAME = 'results_example.data'
#number of trials to run
TRIALS = 100
#attempts to generate a predicate structure the first time, to see if it is possible at all in the current dimension
ATTEMPTS = 10
#dimensions over which to test as (name,levels) tuples
D = [
('predicates',[5,6,7,8,9,10]), #total number of predicates (expressions+functions+relations)
('types',[10,]), # total number of possible predicate types (with the same amount of arity-1 as arity-2 types)
('objects',[1,2,3,4]), #total number of objects
('height',[2,3,4]), #longest path to an object
('chance_function',[1.0,]), #probability of arity-1 predicates being functions
('preservation',[0.5]), #closeness of the graphs
('typeshape',['random']) #shape of predicate type distribution
('heightshape',['square']) #shape of height distribution
('max_arity',[2]) #max arity of predicates
('preservationdecay',[0]) #decay of preservation parameter per layer
('scaling',[1.0]) #size of the target graph compared to base
]
#
# END CONFIGURATION SECTION
#
simulator.run(A,D,TRIALS,ATTEMPTS,OUTFILENAME)
def main():
#env = gym.make(ENV_NAME)
agent = Agent(5)
if TRAIN: # Train mode
for _ in range(NUM_EPISODES):
terminal = False
time = 0
point_list = np.zeros(18)
reward_circle_list = np.zeros((8,3))
rx = 226
ry = 226
r_theta = 1
ex = 23
ey = 23
e_theta = 2
e_action = 0
#observation = env.reset()
observation = lim.learn_image_init()
"""
for _ in xrange(random.randint(1, NO_OP_STEPS)):
last_observation = observation
observation, _, _, _ = env.step(0) # Do nothing
state = agent.get_initial_state(observation, last_observation)
"""
last_observation = observation
is_init_state_random = random.randint(1,10)
if is_init_state_random <=6:
rx = random.randint(20,230)
ry = random.randint(20,230)
while(simulator.collision_check(rx,ry,ex,ey)):
rx = random.randint(20,230)
ry = random.randint(20,230)
observation = lim.learn_image_maker(rx,ry,r_theta,ex,ey,point_list[0],point_list[1],point_list[2],point_list[3],point_list[4],point_list[5],point_list[6],point_list[7],point_list[8],point_list[9],point_list[10],point_list[11],point_list[12],point_list[13],point_list[14],point_list[15],point_list[16],point_list[17])
state = agent.get_initial_state(observation, last_observation)
state = state.transpose(1,2,0)
while not terminal:
last_observation = observation
action = agent.get_action(state)+1
#observation, reward, terminal, _ = env.step(action)
reward,terminal,point_list,time,rx,ry,r_theta,ex,ey,e_theta,point_list,reward_circle_list=simulator.run(time,action,e_action,rx,ry,r_theta,ex,ey,e_theta,point_list,reward_circle)
observation = lim.learn_image_maker(rx,ry,r_theta,ex,ey,point_list[0],point_list[1],point_list[2],point_list[3],point_list[4],point_list[5],point_list[6],point_list[7],point_list[8],point_list[9],point_list[10],point_list[11],point_list[12],point_list[13],point_list[14],point_list[15],point_list[16],point_list[17])
# env.render()
#processed_observation = preprocess(observation, last_observation)
state = agent.run(state, action, reward, terminal, observation)
print((rx,ry,r_theta))
print(('reward:'+str(reward)))
else: # Test mode
# env.monitor.start(ENV_NAME + '-test')
for _ in range(NUM_EPISODES_AT_TEST):
terminal = False
observation = env.reset()
for _ in range(random.randint(1, NO_OP_STEPS)):
last_observation = observation
observation, _, _, _ = env.step(0) # Do nothing
state = agent.get_initial_state(observation, last_observation)
while not terminal:
last_observation = observation
action = agent.get_action_at_test(state)
observation, _, terminal, _ = env.step(action)
env.render()
processed_observation = preprocess(observation, last_observation)
state = np.append(state[1:, :, :], processed_observation, axis=0)
try:
start_address = compiler.parse_program(stream, load_address)
except SyntaxError as exception:
if name.endswith("fail.txt"):
print "failed as expected"
continue
else:
print "failed", str(exception)
sys.exit(1)
else:
print "passed"
if name in compile_only:
print "Not running", name
continue
reload(simulator)
simulator.load(generator.code, load_address)
result = simulator.run(start_address, step = False, verbose = False)
try:
expected = expected_results[name]
except KeyError:
print "Obtained:", result
raise
if result != expected:
print "Expected:", expected
print "Obtained:", result
ATTEMPTS = 10
#dimensions over which to test as (name,levels) tuples
D = [
('predicates',
[5, 6, 7, 8, 9,
10]), #total number of predicates (expressions+functions+relations)
(
'types', [
10,
]
), # total number of possible predicate types (with the same amount of arity-1 as arity-2 types)
('objects', [1, 2, 3, 4]), #total number of objects
('height', [2, 3, 4]), #longest path to an object
('chance_function', [
1.0,
]), #probability of arity-1 predicates being functions
('preservation', [0.5]), #closeness of the graphs
('typeshape', ['random']) #shape of predicate type distribution
('heightshape', ['square']) #shape of height distribution
('max_arity', [2]) #max arity of predicates
('preservationdecay', [0]) #decay of preservation parameter per layer
('scaling', [1.0]) #size of the target graph compared to base
]
#
# END CONFIGURATION SECTION
#
simulator.run(A, D, TRIALS, ATTEMPTS, OUTFILENAME)
import simulator
from conf import init_conf, Conf
if __name__ == '__main__':
for i in range(1000, 2000, 50):
Conf.CLIENT_NO = i
init_conf('conf')
_, accuracy = simulator.run()
if accuracy < 0.05:
result = "sufficient number of patients for 95% accuracy: {}".format(
i)
file = open('accuracy_result.txt', 'w')
file.write(result)
file.close()
break
import simulator
from utils import plot_history
init = {'victim': 0, 'predator': 2}
value_range = 3
def vp(config):
impacts = dict(victim=0, predator=0)
if config['victim'] < config['predator']:
impacts['victim'] = -1
impacts['predator'] = -1
elif config['victim'] > config['predator']:
impacts['victim'] = 1
impacts['predator'] = 1
else:
impacts['victim'] = 1
impacts['predator'] = -1
return impacts
response = simulator.run("report", value_range, init, vp, nsteps=150)
print(response)
response = plot_history("report")
print(response)
sys.stderr.write(str(exception) + "\n")
sys.exit(1)
print "Functions:"
pprint.pprint(compiler.functions)
print "Main variables:"
pprint.pprint(compiler.local_variables)
print "Main code:"
addr = program_address
for v in generator.code:
print "%04x: %03i (%02x)" % (addr, v, v)
addr += 1
print "Opcode usage:"
d = compiler.get_opcodes_used()
freq = map(lambda (k, v): (v, k), d.items())
freq.sort()
for v, k in freq:
print simulator.lookup[k], v
if run:
print "Loading"
simulator.load(generator.code, program_address)
print "Running"
print simulator.run(start_address)
if output:
compiler.save_opcodes(file_name)
if __name__ == '__main__':
# check for 2 rooms with 2 and 3 doctors
visit_rate = [1, 1, 1, 1, 1]
content = None
plot_array = []
while (True):
content = "2, 6, 20, 5\n{}, {}\n{}, {}, {}\n".format(
visit_rate[0], visit_rate[1], visit_rate[2], visit_rate[3],
visit_rate[3])
file = open("conf", "w")
file.write(content)
file.close()
init_conf('conf')
r, _ = simulator.run()
plot_array.append(sum(r[0]) + sum(r[1]))
if (sum(r[0]) + sum(r[1])) == 0:
break
else:
for i in range(len(visit_rate)):
visit_rate[i] += 1
print("ideal doctors visit rate:")
print(visit_rate)
labels = range(len(plot_array))
fig, ax = plt.subplots()
ax.plot(labels, plot_array)
ax.set(xlabel="doctors service rate",
ylabel='number of patients in rooms queue',
title='Check if queues are empty in rooms')
delta_y = state.position.y - ref[1]
if delta_x < 0.5 and delta_y < 0.5:
delta_fi = ref[2] - state.orientation
delta_fi = atan2(sin(delta_fi), cos(delta_fi))
hl_1 = half_length * cos(delta_fi) + half_width * sin(delta_fi)
hw_1 = half_width * cos(delta_fi) + half_length * sin(delta_fi)
R = np.array(((cos(-ref[2]), -sin(-ref[2])),
(sin(-ref[2]), cos(-ref[2]))))
delta_1 = R.dot([delta_x, delta_y])
if abs(delta_1[0]) < 0.9 - hw_1 and abs(delta_1[1]) < 1.175 - hl_1 and delta_fi < 0.411:
return True
return False
def list_from_file(file_in):
output_list = []
with open(file_in) as file:
for line in file:
p = line.split(':')
p_ = [float(i) for i in p]
output_list.append(p_)
return output_list
p_list = list_from_file("parking_spots.txt")
mpc = mpc2.ModelPredictiveControl()
visualizer.visualize(p_list)
simulator.run(controller)
## Experiment data -------------------------
initialBacklogSize = 15
initialCodedSize = 0
maxTime = 30.0 # days
meanDevTime = 10.5 # mean, days
varDevTime = 0.1 # variation (the less the better process)
quality = 1 # probability of reopen
meanTestTime = 6.0 # mean, days
meanUsArrival = 2.0 # mean, days
developerCount = 5
qaCount = 2
from simulator import run
data = run({"theseed": 99999, "initialBacklogSize": initialBacklogSize, "maxTime": maxTime, "meanDevTime": meanDevTime,
"varDevTime": varDevTime, "meanTestTime": meanTestTime, "meanUsArrival": meanUsArrival,
"developerCount": developerCount, "qaCount": qaCount, "initialCodedSize": initialCodedSize,
"quality": quality})
print data
input('press a key')
from request_stream import RequestStream
import simulator
import request_handler_fifo
req_stream = RequestStream("write")
simulator.run(100000)
#print simulator.time
#print simulator.required_request_count
print "#######################\n"
print "Done with insertions. Now will perform required operations"
print "#######################\n"
req_stream.update_type("mixed")
simulator.run(100000)
import simulator as sim
import numpy as np
import csv
num_simulations = 500
M = range(1, 31)
# Distance between Rx and Tx
# D = [5, 10, 20]
D = [5, 10, 20]
writer = csv.writer(open('geraf.csv', 'w'), delimiter=',')
writer.writerow(D)
writer.writerow(M)
for d in D:
n_hops = []
std_dev_on_hops = []
for m in M:
n_hops_m = []
for _ in range(num_simulations):
n_hops_m.append(sim.run(d, m))
n_hops.append(np.mean(n_hops_m))
std_dev_on_hops.append(np.std(n_hops_m))
writer.writerow(n_hops)
writer.writerow(std_dev_on_hops)
template_triangles = renderer.render_template_triangles()
# Generate or load agents
failed_to_load_agents = False
flocks = None
if cfg.reuse_agents:
flocks = agents.load_from_pickle(amaze)
if flocks is None:
failed_to_load_agents = True
if not cfg.reuse_agents or failed_to_load_agents:
flocks = agents.generate_first_flock(amaze)
global_map, buffers, completion_time, solver =\
simulator.run(context, device, queue, amaze, flocks, template_triangles)
t2 = time()
print("Calculations [s]:", t2 - t1)
if cfg.render_display_on:
animation = renderer.render_animation(amaze, flocks, template_triangles, global_map)
if cfg.output_video:
displayer.savevideo(animation)
if cfg.output_images:
displayer.saveimages(animation)
t3 = time()
print("Rendering [s]:", t3 - t2)
print("Total [s]:", t3 - t1)
