def main():
try:
parser = get_args_parser()
args = parser.parse_args()
if args.command == "predict_proba":
predict_proba(args.input_path, args.model_path, args.device)
elif args.command == "predict":
print(predict(args.input_path, args.model_path, args.device))
elif args.command == "learn":
learn(args.location, args.num_samples, args.device)
elif args.command == "crossval":
crossval(path=args.model_path)
elif args.command in ["locations", "ls"]:
locations(args.model_path)
elif args.command == "rename":
rename_label(args.label, args.new_label)
print("Retraining model...")
train_model()
elif args.command == "train":
train_model(args.model_path)
else:
parser.print_help()
parser.exit(1)
except (KeyboardInterrupt, SystemExit):
exit()
def learn(label, n=100):
path = ensure_whereami_path()
label_path = get_label_file(path, label + ".txt")
for _ in tqdm(range(n)):
try:
new_sample = sample()
if new_sample:
write_data(label_path, new_sample)
except KeyboardInterrupt: # pragma: no cover
break
train_model()
def learn(label, n=100):
path = ensure_whereami_path()
label_path = get_label_file(path, label + ".txt")
file_exists = os.path.isfile(label_path)
for _ in tqdm(range(n)):
try:
data = sample()
write_data(label_path, data, file_exists)
file_exists = True
except KeyboardInterrupt:
break
train_model()
def learn(label, n=1, device=""):
path = ensure_whereami_path()
label_path = get_label_file(path, label + ".txt")
for i in tqdm(range(n)):
if i != 0:
time.sleep(15)
try:
new_sample = sample(device)
if new_sample:
write_data(label_path, new_sample)
except KeyboardInterrupt: # pragma: no cover
break
train_model()
def main():
parser = get_args_parser()
args = parser.parse_args()
if args.command == "predict_proba":
predict_proba()
elif args.command == "predict":
print(predict())
elif args.command == "learn":
learn(args.location, args.num_samples)
elif args.command == "crossval":
crossval()
elif args.command in ["locations", "ls"]:
locations()
elif args.command == "rename":
rename_label(args.label, args.new_label)
print("Retraining model...")
train_model()
else:
parser.print_help()
parser.exit(1)
for i in tqdm(range(len(predictions))):
print(i, predictions[i], labels[i])
if predictions[i] == labels[i]:
num_correct = num_correct + 1
return num_correct / len(predictions)
exp1_source_files = [
'/home/pi/.whereami/test_data/test_circle_table_west.txt',
'/home/pi/.whereami/test_data/test_square_table_east.txt',
'/home/pi/.whereami/test_data/test_thermo_red_square.txt',
'/home/pi/.whereami/test_data/test_table_1_north_outlet.txt'
]
exp4_source_files = [
'/home/pi/.whereami/test_data/test_near_base.txt',
'/home/pi/.whereami/test_data/test_near_two.txt',
'/home/pi/.whereami/test_data/test_near_one.txt'
]
exp3_source_files = [
'/home/pi/.whereami/test_data/test_m.txt',
'/home/pi/.whereami/test_data/test_ms.txt',
'/home/pi/.whereami/test_data/test_mss.txt'
]
path = '/home/pi/.whereami/test_data'
source_files = [
os.path.join(path, f) for f in os.listdir(path)
if os.path.isfile(os.path.join(path, f))
]
train_model(model_type='nn')
print(evaluate_test_acc(source_files))