def add_unavailability(self, solver, unavailability_map):
"""
This constraint ensures that workers cannot work scheduled tasks within given time spans
unavailability_map : { [worker_id]: { range: { start_time, end_time } } }
"""
for worker in self.workers:
if str(worker.id) in unavailability_map:
range = utils.get_range(unavailability_map[str(worker.id)]['range'])
tasks_in_range = utils.get_tasks_in_range(self.tasks, range)
for task in tasks_in_range:
solver.Add(self.assignments[worker.index][task.index] == 0)
def get_work(factory_id, worker_id):
w_id = db.assign_work(factory_id, worker_id).get('work_id')
if w_id is None:
return {'isWork': 0}
f_info = db.get_factory_info(factory_id)
f_size = f_info['file_size']
w_size = f_info['work_size']
[a, b] = utils.get_range(f_size, w_size, w_id)
print('[INTERFACE GET WORK] done.')
return {'isWork': 1, 'work_id': w_id, 'range_start': a, 'range_end': b}
def matching_files_by_size(torrent, exist_files):
tor_name = torrent_utils.get_torrent_name(torrent)
files = []
if torrent_utils.is_single_file_torrent(torrent):
size = torrent_utils.get_payload_size(torrent)
files.append( TFile('', tor_name, size, 0) )
else:
tor_files = torrent['info']['files']
stream_offset = 0
for file in tor_files:
fsize = file['length']
matched_files = get_range(exist_files, fsize, lambda x: x['len'])
fpath = os.path.join(tor_name, *file['path'][:-1])
fname = file['path'][-1]
files.append(TFile(fpath, fname, fsize, stream_offset, matched_files))
stream_offset += fsize
return files
def matching_files_by_size(torrent, exist_files):
tor_name = torrent_utils.get_torrent_name(torrent)
files = []
if torrent_utils.is_single_file_torrent(torrent):
size = torrent_utils.get_payload_size(torrent)
files.append(TFile('', tor_name, size, 0))
else:
tor_files = torrent['info']['files']
stream_offset = 0
for file in tor_files:
fsize = file['length']
matched_files = get_range(exist_files, fsize, lambda x: x['len'])
fpath = os.path.join(tor_name, *file['path'][:-1])
fname = file['path'][-1]
files.append(
TFile(fpath, fname, fsize, stream_offset, matched_files))
stream_offset += fsize
return files
dataloader = DataLoader(dataset, batch_size=batch_size)
net = models.UNET_1D(input_channels, depth_step, kernel_size)
writer = SummaryWriter(log_dir=f'./logs/{time}')
weight = torch.tensor(weight)
loss_fn = nn.NLLLoss(weight)
optimizer = optim.Adam(net.parameters(), lr=lr)
max_acc = 0
l, v = dataset.get_length()
num_batches = get_range(l, mini_batch)
val_batches = get_range(v, mini_batch)
for i in range(num_epochs):
""" Set variables to zero """
batch_losses, counter, batch_accuracy = 0, 0, 0
recall, precision, val_recall, val_precision = 0, 0, 0, 0
epoch_losses, epoch_counter, epoch_accuracy = 0, 0, 0
for data in dataloader:
""" Training """
net.train()
targets = data['training']['labels']
inputs = data['training']['sequence']
for k in num_batches:
SYNAPSE_USER = os.environ["SYNAPSE_USER"]
SYNAPSE_PW = os.environ["SYNAPSE_PW"]
driver = "ODBC+Driver+17+for+SQL+Server"
db_name = "loan"
table_name = "cleanLoan"
port = "1433"
engine = create_engine(
f"mssql+pyodbc://{SYNAPSE_USER}:{SYNAPSE_PW}@{SYNAPSE_HOST}:{port}/{db_name}?driver={driver}"
)
connection = engine.connect()
# Get distincts and range
loan_min, loan_max = utils.get_range(connection, db_name, table_name,
"loan_amnt")
inc_min, inc_max = utils.get_range(connection, db_name, table_name,
"annual_inc")
app_types = utils.get_unique(connection, db_name, table_name,
"application_type")
purposes = utils.get_unique(connection, db_name, table_name, "purpose")
ownerships = utils.get_unique(connection, db_name, table_name,
"home_ownership")[1:]
# Close connection
connection.close()
# Make some calculations based on value range retrieved
loan_marks = loan_max // 4
loan_min //= loan_marks
import json
from os.path import dirname
#net = torch.load('./trained_models/best_model_weight_3.pt')
path = 'logs/2012012257/best_performance.pt'
root = dirname(path)
net = torch.load(path)
net.eval()
dataset = SyntheticData('./test_data', 1)
dataloader = DataLoader(dataset, batch_size=1)
mini_batch = 512
l, v = dataset.get_length()
num_batches = get_range(l, mini_batch)
N = len(dataset)
test_results = {'PNR': [], 'Counts': [], 'TP': [], 'TC': [], 'FP': []}
for i, data in enumerate(dataloader):
targets = data['training']['labels']
inputs = data['training']['sequence']
counts, ground = 0, 0
epoch_losses, epoch_counter, epoch_accuracy, recall, precision = 0, 0, 0, 0, 0
true_positive, total_count, false_positive, total_count_p = 0, 0, 0, 0
for k in num_batches:
with torch.no_grad():
x, t = get_chunks(inputs, targets, k, mini_batch)
def check_range(self, sub_fighter, obj_fighter):
if self.max_range > utils.get_range(
sub_fighter.fcs.platform,
obj_fighter.fcs.platform) > self.min_range:
self.increase_fitness(1)
def compute_range_error_score( self, blue_plant, red_plant ) :
distance_to_red = utils.get_range( blue_plant.platform, red_plant.platform )
# Safety distance is set to 500m
engagement_range = distance_to_red - 500.0
# "Optimal" range is set to 800m
return 1.0 - np.exp( -1.0*np.power(engagement_range - 800.0,2.0) )
def save(self, *args, **kwargs):
if self.ip:
self.range_from, self.range_to = get_range(self.ip)
self.is_network = (self.range_to - self.range_from) > 0
super(Address, self).save(*args, **kwargs)
def matching(self, ip):
ip_from, ip_to = get_range(ip)
where = "(ip = %s) OR (%s BETWEEN range_from AND range_to) OR (%s BETWEEN range_from AND range_to) OR (%s <= range_from AND %s >= range_to)"
params = [str(ip), ip_from, ip_to, ip_from, ip_to]
return self.get_query_set().extra(where=[where], params=params)
