def __init__(self, arg):
self.arg = arg
self.simMode = arg.get('simMode', 1)
self.config = arg.get('config', None)
self.breakpoint = arg.get('breakpoint', None)
self.plot = arg.get('plot', None)
self.log = logging.getLogger('trader')
self.tick_btce = api_btce.publicapi()
self.tick_stamp = api_stamp.publicapi()
if self.simMode != 0:
self.tapi_btce = api_btce.tradeapi(self.config.apikey_btce,
self.config.apisecret_btce)
self.tapi_stamp = api_stamp.tradeapi(self.config.apikey_stamp,
self.config.apisecret_stamp,
self.config.apiid_stamp)
self.hr_btce = historyRecorder(self.arg, "btce")
self.hr_stamp = historyRecorder(self.arg, "stamp")
else:
self.hr_btce = simulator.simHistoryRecorder(self.arg, "btce")
self.hr_stamp = simulator.simHistoryRecorder(self.arg, "stamp")
self.sim_btce = simulator.simulator("btce", {
"usd": 0,
"btc": 1
}, self.arg, self.hr_btce, 0.002)
self.sim_stamp = simulator.simulator("stamp", {
"usd": 500,
"btc": 0
}, self.arg, self.hr_stamp, 0.005)
def make_random(k):
vtx = random_vertex()
theta, phi = random_direction()
direct = unit_vector(theta, phi)
pts, deps = track_points(vtx, direct)
print(len(deps))
x, y, z = unzip(pts)
scatter3d_module(x, y, z, deps, k)
zero = geo.point_3d(0., 0., 0.)
#box = geo.box(714*cm, 300*cm, 510*cm, zero, "ar_cube")
module_box = geo.box(100.0 * cm, 300.0 * cm, 100 * cm, zero,
"rand" + str(k) + "_box")
geometry = geo.geometry([module_box], zero)
simul = sim.simulator(geometry)
simul.supervisor._outfile = OUT_DIR + "rand_event" + str(k) + ".root"
geom = simul.project_event(pts, deps, vtx)
simul.set_minimal()
simul.supervisor._truth.SetBinContent(1, vtx[0])
simul.supervisor._truth.SetBinContent(2, vtx[1])
simul.supervisor._truth.SetBinContent(3, vtx[2])
simul.supervisor._truth.SetBinContent(4, theta)
simul.supervisor._truth.SetBinContent(5, phi)
simul.supervisor.write()
def learningTrial(self):
for _ in range(self.N):
episode = []
cstate = self.initialState()
sim = simulator(self.mdp, currentS=cstate, terminalS=self.ts)
T = 0
visited = set()
while True:
a = self.exploreOrExploitPolicy(cstate)
res = sim.takeAction(a)
#print(res)
if res == "end":
# print(111)
break
r, _ = res
episode.append((cstate, a, r))
_, cstate = res
T += 1
# print(episode)
for t in range(T):
cstate, a, _ = episode[t]
if (cstate, a) in visited:
continue
visited.add((cstate, a))
gt = 0
for tt in range(t, T):
gt += episode[tt][2] * (self.beta**(tt - t))
self.totalq[cstate][a] += gt
self.countq[cstate][a] += 1
self.q[cstate][
a] = self.totalq[cstate][a] / self.countq[cstate][a]
def create_temp_simulator(items, agents, main_agent):
local_map = []
row = [0] * 10
for i in range(10):
local_map.append(list(row))
local_items = []
for i in range(len(items)):
(item_x, item_y) = items[i].get_position()
local_item = item.item(item_x, item_y, items[i].level, i)
local_item.loaded = items[i].loaded
local_items.append(local_item)
if not local_item.loaded:
local_map[item_y][item_x] = 1
local_agents = list()
(a_agent_x, a_agent_y) = agents[0].get_position()
local_map[a_agent_y][a_agent_x] = 8
local_agent = agent.Agent(a_agent_x, a_agent_y, 'l1', 0)
local_agents.append(local_agent)
(m_agent_x, m_agent_y) = main_agent.get_position()
local_map[m_agent_y][m_agent_x] = 9
local_main_agent = agent.Agent(m_agent_x, m_agent_y, 'l1', 1)
local_main_agent.set_level(main_agent.level)
tmp_sim = simulator.simulator(local_map, local_items, local_agents,
local_main_agent, 10, 10)
return tmp_sim
def gen_ckpt(netname):
model = ''
sys.path.append("input")
if 'vgg' in netname:
from vggnet import vgg
model = vgg(num_classes=1000)
elif 'resnet18' in netname:
from resnet import resnet18
model = resnet18()
elif 'resnet34' in netname:
from resnet import resnet34
model = resnet34()
elif 'resnet50' in netname:
from resnet import resnet50
model = resnet50()
elif 'simulator' in netname:
from simulator import simulator
model = simulator()
import torch
for m in model.modules():
if isinstance(m, (torch.nn.Conv2d, torch.nn.Linear)):
m.register_buffer(f'scale', torch.tensor(1.0))
torch.save({
'state_dict': model.state_dict(),
'hyper_parameters': {
'conv.scale' : 0.0,
'act_quant_bit' : 8,
'weight_quant_bit' : 8,
},
}, f'input/{netname}.ckpt')
sys.path.append("..")
def generate_ppm_file(image, pi_estimator, nb_points_par_image, i, decimale):
'''
Cette fonction permet de générer une image au format ppm. Elle appelle
la fonction simulation lui permettant de générer des points aléatoirement.
Chaque estimation receuillie permet grâce à la méthode add_new_pi_estimate
d'obtenir des valeurs convergentes vers pi. La valeur est ensuite écrite
sur l'image créée et sauvegardée dans le dossier appelée out.
'''
# 1 - Obtention de l'estimation de pi
list_blue, list_pink = [], []
pi_estime = simulator.simulator(nb_points_par_image, list_blue, list_pink)
# 2 - Coloration des pixels de l'image en fonction de leur distance au centre.
color_image_with_points(image, list_blue, list_pink)
if ARGUMENTS.p:
pourcentage = validation_points(image, i, nb_points_par_image)
print(f'Image {i+1} : {int(pourcentage)} %')
# 3 - Nouvelle estimation ajoutée pour actualiser la moyenne globale
pi_val = pi_estimator.add_new_pi_estimate(pi_estime)
# 4 - Affichage de l'estimation de pi sur l'image
# La copie entière de l'image permet d'écrire le nombre pi
# sur celle-ci avant de la sauvegarder au format ppm.
copie_image = copy.deepcopy(image)
nb_decimal = f".{decimale}f"
write_pi_on_image(copie_image, pi_val, nb_decimal)
# 5 - Sauvegarde de l'image sur le disque dur
partie_decimale = int((pi_val - int(pi_val)) * (10**decimale))
imageio.imwrite(f"out/img{i}_{int(pi_val)}-{partie_decimale}.ppm",
copie_image)
def launch():
t = time.time()
network = q_network.DeepQLearner(input_width, input_height, num_actions,
phi_length,
discount,
learning_rate,
rms_decay,
rms_epsilon,
momentum,
clip_delta,
freeze_interval,
batch_size,
network_type,
update_rule,
batch_accumulator,
rng)
print 'compile network .. done' , time.time()-t
agt = agent.NeuralAgent(network,
epsilon_start,
epsilon_min,
epsilon_decay,
replay_memory_size,
experiment_prefix,
replay_start_size,
update_frequency,
rng)
print 'create agent & simulator .. done'
env = simulator.simulator()
exp = experiment(agt,env)
exp.run_episode(50)
def generate_data_labels(img_size, x0, y0, SNR, N):
prev_time = time.time()
sim = simulator.simulator(img_size, 1)
data_tr_for_different_snr = []
labels_tr_for_different_snr = []
for snr in SNR:
data_tr = []
labels_tr = []
for i in range(N):
r = np.random.rand()
if r > 0.5:
labels_tr.append(1)
data = sim.create_simulation_from_SNR(x0, y0, snr)
data_tr.append(data)
else:
labels_tr.append(0)
data = sim.create_simulation_from_SNR(x0,
y0,
snr,
no_atom=True)
data_tr.append(data)
data_tr_for_different_snr.append(np.array(data_tr))
labels_tr_for_different_snr.append(np.array(labels_tr))
print(f"time used: {time.time() - prev_time}")
return np.array(data_tr_for_different_snr), np.array(
labels_tr_for_different_snr)
def update_models2(belief, state, expert_models, p, num_models, state_rewards):
_rewards = np.zeros(num_models)
next_state = np.zeros(num_models)
for i in range(0, num_models):
next_state[i], _rewards[i] = sim.simulator(state, p, expert_models[i],
state_rewards)
return int(np.random.choice(next_state, 1, p=belief)[0])
def toggleSimulator(self):
if self.sim == None:
enemy = "goudsimu"
self.zeppelins.append(AbstractZeppelin("red", enemy))
mq.initEnemy(enemy)
self.sim = simulator.simulator(100,100, enemy)
return
self.sim.toggleActive()
def __init__(self, global_assignment, name, trainer, cfg):
self.name = str(name) + "_worker_" + str(global_assignment)
self.global_assignment = self.name[-1] #get global assignment index
self.trainer = trainer #get initialized Adam optimizer
self.cfg = cfg
self.local_AC = AC_Network(self.cfg.s_size, self.cfg.a_size, self.name,
trainer) #set up worker
self.env = simulator(a_size)
self.experience = experience(self.cfg)
def update_models(state, expert_models, p, num_models, state_rewards,
real_world_expert):
_rewards = np.zeros(num_models)
next_state = np.zeros(num_models)
for i in range(0, num_models):
next_state[i], _rewards[i] = sim.simulator(state, p, expert_models[i],
state_rewards)
return int(next_state[real_world_expert]), _rewards
def getConfigurationMatrix():
powerSimulator = simulator()
heatMatrix = [[
powerSimulator.runSimulation(
[batteryCapacity / 100.0, panelArea / 100.0])
for batteryCapacity in xrange(0, 1001, 50)
] for panelArea in xrange(0, 2001, 100)]
with open('results.txt', 'w') as outFile:
outFile.write(str(heatMatrix))
def __init__(self, name, trainer, cfg):
self.name = "worker_" + str(name)
self.number = name
self.trainer = trainer
self.cfg = cfg
self.local_AC = AC_Network(self.cfg.s_size, self.cfg.a_size, self.name,
trainer)
self.update_local_ops = update_target_graph('global', self.name)
self.env = simulator(a_size)
self.experience = experience(self.cfg)
def test_redraw(config):
transformer = image_normalize('background')
db = coco(config, 'train', transform=transformer)
output_dir = osp.join(config.model_dir, 'test_redraw')
maybe_create(output_dir)
env = simulator(db, config.batch_size)
env.reset()
loader = DataLoader(db,
batch_size=config.batch_size,
shuffle=True,
num_workers=config.num_workers)
for cnt, batched in enumerate(loader):
out_inds = batched['out_inds'].long().numpy()
out_vecs = batched['out_vecs'].float().numpy()
# sequences = []
for i in range(out_inds.shape[1]):
frames = env.batch_render_to_pytorch(out_inds[:, i], out_vecs[:,
i])
# sequences.append(frames)
# sequences = torch.stack(sequences, dim=1)
sequences = env.batch_redraw(True)
# sequences = [tensors_to_imgs(x) for x in sequences]
for i in range(len(sequences)):
sequence = sequences[i]
image_idx = batched['image_index'][i]
name = '%03d_' % i + str(image_idx).zfill(12)
out_path = osp.join(output_dir, name + '.png')
color = cv2.imread(batched['color_path'][i], cv2.IMREAD_COLOR)
color, _, _ = create_squared_image(color)
fig = plt.figure(figsize=(32, 16))
plt.suptitle(batched['sentence'][i], fontsize=30)
for j in range(min(len(sequence), 14)):
plt.subplot(3, 5, j + 1)
partially_completed_img = clamp_array(sequence[j], 0,
255).astype(np.uint8)
partially_completed_img = partially_completed_img[:, :, ::-1]
plt.imshow(partially_completed_img)
plt.axis('off')
plt.subplot(3, 5, 15)
plt.imshow(color[:, :, ::-1])
plt.axis('off')
fig.savefig(out_path, bbox_inches='tight')
plt.close(fig)
break
def second_thread(self):
while not self.stop.is_set():
self.reset_serial_flag = False # if true, will close out the serial
self.uart_tab.set_port_info(self.serialPort,'disconnected') # update the name of the serial port
if self.serialPort is not 'simulator':
self.connect_serial()
else:
self.sim = simulator()
self.execute_simulator()
else:
return
def getOptimalConfiguration():
powerSimulator = simulator()
results = la.getBestParameters(powerSimulator.runSimulation, 2, 0, 300,
0.01)
with open('results.txt', 'w') as file:
file.write(str(results))
print(results[1][0])
pd.Series(results[0]).plot()
def __init__(self, arg):
self.arg = arg
self.simMode = arg.get('simMode', 1)
self.config = arg.get('config', None)
self.breakpoint = arg.get('breakpoint', None)
self.plot = arg.get('plot', None)
self.log = logging.getLogger('trader')
self.tick_btce = api_btce.publicapi()
self.tick_stamp = api_stamp.publicapi()
if self.simMode != 0:
self.tapi_btce = api_btce.tradeapi(self.config.apikey_btce, self.config.apisecret_btce)
self.tapi_stamp = api_stamp.tradeapi(self.config.apikey_stamp, self.config.apisecret_stamp, self.config.apiid_stamp)
self.hr_btce = historyRecorder(self.arg, "btce")
self.hr_stamp = historyRecorder(self.arg, "stamp")
else:
self.hr_btce = simulator.simHistoryRecorder(self.arg, "btce")
self.hr_stamp = simulator.simHistoryRecorder(self.arg, "stamp")
self.sim_btce = simulator.simulator("btce", {"usd": 0, "btc": 1}, self.arg, self.hr_btce, 0.002)
self.sim_stamp = simulator.simulator("stamp", {"usd": 500, "btc": 0}, self.arg, self.hr_stamp, 0.005)
def statrunsTimeDone():
num_runs = 20
output_file = "../results/tLimited_auction.txt"
file = open(output_file, 'w')
num_rows = 10
row_len = 8
num_bins = 100
num_workers = 15
num_bots = 8
# 0 = greedy, 1 = auction based, 2= replanning
coord_method = 2
writeHeader(file, num_rows, row_len, num_bins, num_workers, num_bots,
coord_method, '0')
file.write("percentage picked, ")
file.write("time wasted\n")
time_done = []
for i in range(num_runs):
sim = simulator(num_rows, row_len, num_bots, num_bins, num_workers)
coord = coordinator(coord_method)
plnr = planner(sim)
time_taken = 0
while (sim.applesLeft() != 0 and time_taken < 10000):
plan = coord.cordStep(sim)
plan = plnr.getPlan(plan, sim)
sim.step()
#sim.drawSimulator()
time_taken += 1
time_done.append(time_taken)
print "done running iteration: ", i
print "mean percent picked: ", np.mean(time_done)
print "std percent picked: ", np.std(time_done)
def learningTrial(self):
for _ in range(self.N):
cs = self.initialState()
sim = simulator(self.mdp, currentS=cs, terminalS=self.ts)
while True:
# print(cs, sim.cs)
a = self.exploreOrExploitPolicy(cs)
res = sim.takeAction(a)
if res == "end":
break
r, ns = res
# print(cs, ns)
ga = self.greedyPolicy(ns)
self.q[cs][a] = self.q[cs][a] + self.alpha * (
r + (self.beta * self.q[ns][ga]) - self.q[cs][a])
cs = ns
def eachSimulation(mdp, p, beta, maxSteps, pOfParkingRandom, terminalState):
_, A, _, _ = mdp
steps = 0
cs = 0
reward = 0
sim = simulator(mdp, currentS=cs, terminalS=terminalState)
while steps < maxSteps:
res = sim.takeAction(p(mdp, cs, p=pOfParkingRandom))
if res == "end":
break
elif res == "illegal":
steps -= 1
else:
r, cs = res
reward += (beta**steps) * r
steps += 1
return reward
def __init__(self, port):
self.port = port
self.simulator = simulator.simulator()
self.gui = simulator_gui.simulator_gui(self.simulator)
self.led_yellow = output_dummy(self.gui.set_yellow_led)
self.led_green = output_dummy(self.gui.set_green_led)
self.heater = output_dummy(self.simulator.set_heater)
self.sirup_pump = output_dummy(self.simulator.set_sirup)
self.water_pump = output_dummy(self.simulator.set_water)
self.sirup_valve = output_dummy(self.simulator.set_sirup_valve)
self.water_valve = output_dummy(self.simulator.set_water_valve)
self.keypad = sensor_dummy(self.gui.get_keypad)
self.reflex = sensor_dummy(self.simulator.get_cup)
self.temperature = sensor_dummy(self.simulator.read_temp)
self.distance = sensor_dummy(self.simulator.read_mm)
def upload_files():
sm = simulator()
sm.admin_login(name, pw)
file_index = get_file_index()
pass_filename = ['README.md', '_config.yml']
for file_name in file_index:
file_name = file_name.strip()
if file_name not in pass_filename:
file_content = open(get_file_path() + file_name.strip(),
'r').read()
data = {
'title': file_name,
'content': file_content,
'event_time':
common_tools.time_to_str(int(time.time())).split()[0],
'post_items': []
}
res = sm.upload(data)
print(res)
def eachSimulation(self, maxSteps):
_, A, _, _ = self.mdp
steps = 0
cs = self.initialState()
reward = 0
sim = simulator(self.mdp, currentS=cs, terminalS=self.ts)
while steps < maxSteps:
# print(self.greedyPolicy(cs))
res = sim.takeAction(self.greedyPolicy(cs))
if res == "end":
break
elif res == "illegal":
steps -= 1
else:
r, cs = res
reward += (self.beta**steps) * r
steps += 1
# print(steps)
return reward
def main():
num_rows = 10
row_len = 8
num_bins = 100
num_workers = 8
num_bots = 5
# 0 = greedy, 1 = auction based, 2= replanning
coord_method = 0
num_timestep = 100
sim = simulator(num_rows, row_len, num_bots, num_bins, num_workers)
coord = coordinator(coord_method)
plnr = planner(sim)
sim.drawSimulator()
for timestep in range(num_timestep):
plan = coord.cordStep(sim)
# print sim.getIdleBots()
# for item in plan:
# print "Robot: ", item.robot_id
# print "Goals: ", item.locations
# print "here"
plan = plnr.getPlan(plan, sim)
# print "Planner Plan", plan
sim.drawSimulator()
# print "apples picked: ",sim.apples_picked
# for key in sim.bots.keys():
# print key, sim.bots[key].plan
#raw_input()
print timestep
sim.step()
print "done running"
print "total num apples: ", sim.total_apples
print "num apples picked: ", sim.apples_picked
print "percentage picked: ", sim.apples_picked / sim.total_apples * 100
print "wasted time: ", sim.wasted_time
plt.show()
return 0
def action(self):
# self.qTable = {}
lactions = self.legalAction(self.gb.board)
# print(lactions)
for i in range(self.nos):
sim = simulator(board=numpy.copy(self.gb.board),
stepsLimited=self.sl,
beta=self.beta,
discounted=self.d)
sboard = sim.sgb.board.tostring()
unactions = []
backQ = []
for a in lactions:
if (sboard, a) not in self.qTable:
unactions.append(a)
la = lactions
# print(len(unactions))
while len(unactions) == 0 and not sim.sgb.islost:
ucbA = self.UCBPolicy(sboard, la)
backQ.append((sboard, ucbA))
sim.takeAction(ucbA)
sboard = sim.sgb.board.tostring()
la = self.legalAction(sim.sgb.board)
unactions = []
for a in la:
if (sboard, a) not in self.qTable:
unactions.append(a)
if len(unactions) != 0:
simA = unactions[randint(0, len(unactions) - 1)]
backQ.append((sim.sgb.board.tostring(), simA))
sim.takeAction(simA)
r = sim.simulation(sim.randomPolicy)
# print(r)
# print(backQ)
for backS, backA in backQ[::-1]:
if (backS, backA) not in self.qTable:
self.qTable[backS, backA] = [0, 0]
self.qTable[backS, backA][0] += r
self.qTable[backS, backA][1] += 1
return self.greedyPolicy(self.gb.board.tostring(), lactions)
def conversation(bot, update):
input_text = update.message.text
chat_id = update.message.chat_id
print(chat_id, input_text)
if input_text == 'exit':
send_message(bot, chat_id, 'bye')
return
dbm = DBManager()
code = dbm.get_code(input_text)
if code is None:
send_message(bot, chat_id, 'not found.')
return
poten_datas = dbm.get_target_forecast(code)
forecast_msg = get_forecast_explain(poten_datas)
simul_msg = simulator.simulator(code=code)
print(forecast_msg)
print(simul_msg)
send_message(bot, chat_id, forecast_msg)
send_message(bot, chat_id, simul_msg)
def __main__():
geometry = sim.default_geo()
data_file = root.TFile.Open("~/hex/berk/nersc/data/jackBig.root")
tree = data_file.Get("tree")
dummy = rd.supervisor()
d = rd.data(tree, dummy)
d.stack()
for k, event_number in enumerate(events):
print("SIMULATING EVENT: " + str(event_number))
simul = sim.simulator(geometry)
d = rd.data(tree, simul.supervisor)
p, e, vtx = d.get_event_w_shift(event_number, module_box._xwalls,
module_box._ywalls, module_box._zwalls)
geom = simul.project_event(p, e, vtx)
simul.get_stats()
simul.supervisor.write()
def effect_k_scenario():
logger = logging.getLogger('upgrade_product_logger')
logger.debug('Effect of K simulation')
kvalues = [6, 7, 8, 9, 10]
dims = np.full(1, defaultDim, dtype='int32')
datasizes = np.full(1, defaultDataSize, dtype='int32')
dists = np.array(['corr', 'uni', 'anti'])
algNames = np.array(['Upgrade_Algorithm', 'New_Upgrade_Algorithm'])
simSettings = generate_simulation_settings(algNames, datasizes, dims, dists, kvalues)
for setting in simSettings:
ProductFileName = getProductFileName(int(setting[dimIndex]))
ProductBuffer = np.loadtxt(ProductFileName, delimiter=',', dtype='int32')
ProductIndex = 0
for iter in range(numIteration):
DataFileName = getDataFileName(setting[distIndex], int(setting[datasizeIndex]), int(setting[dimIndex]), iter)
logger.debug('[info]Dist = %s, dataSize = %d, dim = %d, k = %d, iter = %d data_file = %s' %
(setting[distIndex], int(setting[datasizeIndex]), int(setting[dimIndex]),
int(setting[kValueIndex]), iter, DataFileName))
aSim = sim.simulator(kValue=int(setting[kValueIndex]), fileName=DataFileName, algName=setting[algIndex],
product=ProductBuffer[ProductIndex])
aSim.run()
ProductIndex += 1
def policy_switching(start_state, experts, num_models, w, h, policy,
state_rewards, _belief):
num_policies = policy.shape[0]
reward_pol_ex = np.zeros((num_policies, num_models))
for pol in range(0, num_policies): # iterate through each policy
for m in range(0, num_models): # iterate through each expert
ave_reward = 0 # holds average reward of expert after w policy evaluations
for _h in range(0, w): # run each policy on expert w times
val, val_temp, state = 0, 0, start_state
for j in range(0, h): # perform h steps of policy rollout
next_s, val_temp = sim.simulator(state, policy[pol],
experts[m], state_rewards)
val += val_temp
state = next_s
ave_reward += val
reward_pol_ex[pol][m] = ave_reward / w # ave reward after w calls
# determine best policy based on p(m1)*R11 + p(m2)*R12 > or < p(m1)*R21 + p(m2)*R22
scaled_val = np.zeros(num_policies)
for i in range(0, num_policies):
scaled_val[i] = np.sum((reward_pol_ex[i] * _belief))
return np.argmax(scaled_val)
def action(self):
lactions = self.legalAction()
self.rewardOfActions = {}
for _ in range(self.nos):
sim = simulator(board = numpy.copy(self.gb.board), stepsLimited = self.sl)
unactions = []
if len(self.rewardOfActions) != len(lactions) * 2:
for a in lactions:
if (a, 't') not in self.rewardOfActions:
unactions.append(a)
a = unactions[randint(0, len(unactions) - 1)]
else:
a = self.exploreOrExploitPolicy(lactions)
sim.takeAction(a)
r = sim.simulation(sim.randomPolicy)
if (a, 't') not in self.rewardOfActions:
self.rewardOfActions[a, 't'] = 0
self.rewardOfActions[a, 'c'] = 0
self.rewardOfActions[a, 't'] += r
self.rewardOfActions[a, 'c'] += 1
return self.greedyPolicy(lactions)
import matplotlib.image as mpimg
import numpy as np
from matplotlib.patches import Rectangle
import vot
import sys
import time
import cv2
import numpy
import collections
from camShift import camShift
sim = True
if sim:
#handle = simulator.simulator("/home/boka/arp/david/")
handle = simulator.simulator("/home/boka/arp/vot-toolkit/workspace/sequences/cup/")
#handle = simulator.simulator("/home/boka/arp/vot-toolkit/workspace/sequences/woman/")
#handle = simulator.simulator("/home/boka/arp/vot-toolkit/workspace/sequences/juice/")
#handle = simulator.simulator("/home/boka/arp/vot-toolkit/workspace/sequences/jump/")
else:
handle = vot.VOT("rectangle")
selection = handle.region()
imagefile = handle.frame()
print("prvo")
if not imagefile:
sys.exit(0)
# image = cv2.imread(imagefile, cv2.IMREAD_GRAYSCALE)
image = cv2.imread(imagefile, cv2.IMREAD_COLOR)
print(imagefile)
args.timeshape, args.njobs, args.est_factor, seed)
final_results = shelve.open(os.path.join(args.dirname, fname))
for name, scheduler, errfunc, args.iterations in instances:
print(name, end='')
if args.iterations is None:
# if no. of iterations is None, it means that a single pass is
# enough (no randomness there)
if name in final_results:
continue
else:
args.iterations = 1
scheduler_results = final_results.get(name, [])
for i in range(args.iterations - len(scheduler_results)):
results = list(simulator.simulator(jobs, scheduler, errfunc))
sojourns = numpy.zeros(n_jobs)
for compl, jobid in results:
sojourns[job_idxs[jobid]] = compl - job_start[jobid]
scheduler_results.append(sojourns)
print('', sojourns.mean(), end='')
sys.stdout.flush()
print()
final_results[name] = scheduler_results
final_results.close()
#def getCurrentMap from the simulator, which builds a map of 1s and 0s
#from the state of the map
# def getCurrentMap(self,sim):
#Returns a map of 1s and 0s based on obstacles in the env
#Build 1s and 0s for the terrain type
# terrain=np.zeros((sim.orchard_map.shape[0],sim.orchard_map.shape[1]))
if __name__ == '__main__':
l = [[1, [1, 1, 'G']], [2, [10, 11, 'G'], [7, 5, 'P']], [3, [4, 6]]]
# p=planner()
# p.getPlan(l)
# print p.robotIDs
# print p.robotTargets
s = simulator(10, 20, 5, 30, 5)
num_rows = 5
row_size = 8
p = planner(s)
p.getPlan(l, s)
# print s.orchard_map[0][1].terrain
# astar2d()
# OLD MAP INITIALIZATION CODE
# m=np.zeros((row_size+2,num_rows*2-1))
#make every other row full of trees
# m[:,0::2]=np.ones((m[:,0::2].shape[0],m[:,0::2].shape[1]))
# orch_module=np.array([1,1,0])
# orch=np.array([])
# for k in range(num_rows):
# orch=np.concatenate((orch,orch_module))
print(name, end='')
sojourns_per_priority = {pri: [] for pri in range(1, 6)}
if args.iterations is None:
# if no. of iterations is None, it means that a single pass is
# enough (no randomness there)
if name in final_results:
continue
else:
args.iterations = 1
scheduler_results = final_results.get(name, [])
for i in range(args.iterations - len(scheduler_results)):
results = list(simulator.simulator(jobs, scheduler, errfunc,
weights))
sojourns = numpy.zeros(args.njobs)
for compl, jobid in results:
sojourn = compl - job_start[jobid]
sojourns[jobid] = sojourn
sojourns_per_priority[priorities[jobid]].append(sojourn)
scheduler_results.append(sojourns)
print('', sojourns.mean(), end='')
sys.stdout.flush()
final_results[name] = scheduler_results
print({pri: numpy.array(s).mean()
for pri, s in sojourns_per_priority.items()})
print()
final_results.close()
('LIFO', schedulers.LIFO),
('LIFO_SR', schedulers.LIFO_SR),
('PS', schedulers.PS),
('SRPT', schedulers.SRPT),
('SRPTPS', schedulers.SRPT_plus_PS),
('FSP', schedulers.FSP),
('FSP+PS', schedulers.FSP_plus_PS),
('LAS', schedulers.LAS),
('FSP+LAS', schedulers.FSP_plus_LAS),
('SRPTLAS', schedulers.SRPT_plus_LAS),
('WFQEGPS', schedulers.WFQE_GPS),
]
results = {}
for name, scheduler in instances:
sim = simulator.simulator(jobs, scheduler,
simulator.fixed_estimations(estimations))
results[name] = {jobid: t for t, jobid in sim}
head_fmt = '\t'.join(['{}'] * (len(instances) + 4))
fmt = '\t'.join(['{}'] + ['{:.2f}'] * (len(instances) + 3))
scheduler_names = [n for n, _ in instances]
header = head_fmt.format('Job', 'Arr.', 'Size', 'Est.', *scheduler_names)
print(header)
print('=' * len(header.expandtabs()))
for (jobid, arrival, d), e in zip(jobs, estimations):
print(fmt.format(jobid, arrival, d, e,
*(results[n][jobid] for n in scheduler_names)))
def run(self, verbose=False, **options):
simu = simulator(self)
simu.run(verbose=verbose, **options)
self.raw = simu.raw
self.assertTrue(kdom_util.kDomByPointsPy(self.buf[4], self.buf, 5))
self.assertTrue(kdom_util.kDomByPointsPy(self.buf[4], self.buf, 6))
self.assertTrue(kdom_util.kDomByPointsPy(self.buf[4], self.buf, 4))
self.assertTrue(kdom_util.kDomByPointsPy(self.buf[0], self.buf, 3))
self.assertFalse(kdom_util.kDomByPointsPy(self.buf[0], self.buf, 6))
self.assertFalse(kdom_util.kDomByPointsPy(self.buf[0], self.buf, 5))
self.assertFalse(kdom_util.kDomByPointsPy(self.buf[3], self.buf, 6))
self.assertTrue(kdom_util.kDomByPointsPy(self.buf[3], self.buf, 5))
def test_retrieve_KSkylines(self):
kdomSky = kdom_util.retrieveKDomSkylinePy(self.buf, 4)
self.assertEqual(2, len(kdomSky))
self.assertTrue(np.array_equal(np.array([[1,1,1,3,3,3], [3,3,3,1,1,1]]), kdomSky))
kdomSky = kdom_util.retrieveKDomSkylinePy(self.buf, 5)
self.assertEqual(3, len(kdomSky))
self.assertTrue(np.array_equal(np.array([[1,1,1,3,3,3], [3,3,3,1,1,1],[2,2,2,4,2,2]]), kdomSky))
if __name__ == '__main__':
#unittest.main()
pathName = os.path.dirname(os.path.abspath(__file__))
ProductBuffer = np.loadtxt(pathName + "\data\product_uni_1_5_0.db", delimiter=",")
aSim = sim.simulator(kValue=4, fileName=os.path.dirname(os.path.abspath(__file__)) + '\data\data_corr_100_5_0.db',
algName='Upgrade_Algorithm',
product=ProductBuffer[0])
start_time = time.time()
aSim.run()
end_time = time.time()
print(end_time-start_time)
#handle = vot.VOT("rectangle")
import vot
import sys
import time
import cv2
import numpy
import collections
import flow
import simulator
import ORF
from ncc import NCCTracker
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
import numpy as np
from matplotlib.patches import Rectangle
handle = simulator.simulator("/home/boka/arp/david/")
selection = handle.region()
imagefile = handle.frame()
print("prvo")
if not imagefile:
sys.exit(0)
#image = cv2.imread(imagefile, cv2.IMREAD_GRAYSCALE)
image = cv2.imread(imagefile, cv2.IMREAD_COLOR)
print(imagefile)
tracker = NCCTracker(image, selection)
tracker_flow = flow.flow(image, selection)
tracker_OT = ORF.flow(image, selection)
print("do tukej")
plt.ion()
def run_jobs(self, jobs, error=simulator.identity):
return list(simulator.simulator(jobs, self.scheduler, error))
import simulator #sim=simulator(z=3,V=3,m=3) #sim.make_W(a_active_W=10,b_active_W=1e5) #sim.make_H(a_active_H=10,b_active_H=1,n=100) #sim.make_X() #dir_save="/global/data/tcga/metadata/data/" #prefix="simulated_data" #sim.save(dir_save,prefix) a_active_W=10 b_active_W=1e5 a_active_H=10 b_active_H=1 sim=simulator.simulator(z=3,V=3,m=10) sim.make_W(a_active_W=10,b_active_W=1e5) sim.make_H(a_active_H=10,b_active_H=1,n=1000) sim.make_X() # save dir_save="/home/yifeng/research/mf/mvmf_v1_1/data/" prefix="simulated_data" + "_a_active_W="+str(a_active_W) + "_b_active_W="+str(b_active_W) + "_a_active_H="+str(a_active_H) + "_b_active_H="+str(b_active_H) sim.save(dir_save,prefix)
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Apr 8 10:08:26 2019
@author: tamird
"""
import simulator
# Test the simulator.py
coord = "/home/data/mashas/tamir/adpy/testing/lig.pdb"
param = "/home/data/mashas/tamir/adpy/testing/pf.mdp"
sim = simulator.simulator(coord, param)
sim.prepare_topology("prepare")
sim.run("run01")
final_results = shelve.open(result_fname)
for name, scheduler, errfunc, args.iterations in instances:
print("scheduler:", name)
if args.iterations is None:
# if no. of iterations is None, it means that a single pass is
# enough (no randomness there)
if name in final_results:
continue
else:
args.iterations = 1
scheduler_results = final_results.get(name, [])
for i in range(args.iterations - len(scheduler_results)):
results = list(simulator.simulator(jobs, scheduler, errfunc))
sojourns = zeros(n_jobs)
for compl, jobid in results:
sojourns[job_idxs[jobid]] = compl - job_start[jobid]
scheduler_results.append(sojourns)
print(sojourns.mean(), end=' ')
sys.stdout.flush()
print()
final_results[name] = scheduler_results
print()
final_results.close()