def workflow(day_dir, args):
st = time.time()
mul_num = args.mul_num
expert = args.expert
subject_name = args.subject_name
output_dir = args.output_dir
post_process = args.post_process
svm = args.svm
if args.preprocess:
print('Start preprocess video')
preprocess(day_dir, expert, subject_name, output_dir, mul_num=mul_num)
print('Preprocess video completed.')
if args.predict:
date = get_path_leaf(day_dir)
test_name = '{}_{}_{}.txt'.format(expert, subject_name, date)
file_list_path = make_list(test_name, output_dir, date)
model = get_pretrained_model(args)
print("Post process: {}, svm: {}".format(post_process, svm))
predict(args,
model,
file_list_path,
output_dir,
date,
args.result_name,
post_process=post_process,
svm=svm)
if args.remove_intermediate:
shutil.rmtree(os.path.join(output_dir, 'intermediate'))
print('Directory {} complete processing!'.format(day_dir))
print("Cost {} minutes".format((time.time() - st) / 60))
def init(self, root_dir, pair, time_series, prices, labels, price_range, is_use_residual, is_train = True): if is_use_residual: file_ext = pair + "_use_resid" else: file_ext = pair if os.path.isfile(root_dir + "volatility41_model_test_predictions_" + file_ext) == False: if is_train == False: return None, None x = self.create_training_set(pair, time_series, prices, labels, price_range) start = 0 end = 700 predictions = [] current_prices = [] while end < len(labels): predictions.append(predict(start, end, x, labels, is_use_residual)) current_prices.append(prices[end - 1]) print start start += 12 end += 12 pickle.dump(predictions, open(root_dir + "volatility41_model_test_predictions_" + file_ext, 'wb')) pickle.dump(current_prices, open(root_dir + "volatility41_model_test_prices_" + file_ext, 'wb')) predictions = pickle.load(open(root_dir + "volatility41_model_test_predictions_" + file_ext, 'rb')) current_prices = pickle.load(open(root_dir + "volatility41_model_test_prices_" + file_ext, 'rb')) return predictions, current_prices
def make_prediction(self, pair, time_series, prices, labels, price_range, history_prices, is_use_residual): train_start = len(labels) - 700 train_end = len(labels) x = self.create_training_set(pair, time_series, prices, labels, price_range) return predict(train_start, train_end, x, labels, is_use_residual)
def put_text(self):
data = predict("image/test.png")
for digit in data:
k, m, n = digit
self.c.create_text(k,
m,
text=n,
fill="red",
font="Times 40 italic bold",
anchor="center")
def init(self,
root_dir,
pair,
time_series,
prices,
labels,
price_range,
is_use_residual,
is_train=True):
#0, 200, 300, 100, 200 in sample
if is_use_residual:
file_ext = pair + "_use_resid"
else:
file_ext = pair
if os.path.isfile(root_dir + "regime59_model_test_predictions_" +
file_ext) == False:
if is_train == False:
return None, None
#self.calendar = pd.DataFrame.from_records(download_calendar(31536000), columns=['currency', 'impact', 'actual', 'forecast', 'time', 'region'])
x = self.create_training_set(pair, time_series, prices, labels,
price_range, prices)
start = 0
end = 700
predictions = []
current_prices = []
while end < len(labels):
predictions.append(
predict(start, end, x, labels, is_use_residual))
current_prices.append(prices[end - 1])
print start
start += 12
end += 12
pickle.dump(
predictions,
open(root_dir + "regime59_model_test_predictions_" + file_ext,
'wb'))
pickle.dump(
current_prices,
open(root_dir + "regime59_model_test_prices_" + file_ext,
'wb'))
predictions = pickle.load(
open(root_dir + "regime59_model_test_predictions_" + file_ext,
'rb'))
current_prices = pickle.load(
open(root_dir + "regime59_model_test_prices_" + file_ext, 'rb'))
return predictions, current_prices
def back_test_recent(self, pair, time_series, prices, labels, price_range, history_prices): x = self.create_training_set(pair, time_series, prices, labels, price_range) end = len(labels) - (40 * 24) start = end - 700 predictions = [] current_prices = [] while end < len(labels): predictions.append(predict(start, end, x, labels)) current_prices.append(prices[end - 1]) print start start += 12 end += 12 return predictions, current_prices
def main():
path = 'J:/weibodata/a.csv'
#result = readfile()
result = openExcel(path)
print(result)
List = name(result)
print(List)
relations = (predict(List))
f = open('relation.csv','w',encoding='utf-8',newline='')
csv_writer = csv.writer(f)
csv_writer.writerow(["number","person1", "person2", "relation","content","time_created"])
for number in range(len(relations)):
relation = relations[number]
if relation[2] != '':
csv_writer.writerow([number]+relation+[result[number][1]])
def init(self, pair, time_series, prices, labels, price_range, lag):
if os.path.isfile("/tmp/barrier1_model_test_predictions_" +
pair) == False:
cluster_num = 4
self.kmeans = KMeans(n_clusters=cluster_num,
init='k-means++',
max_iter=100,
n_init=1,
random_state=42).fit(prices)
x = self.create_training_set(pair, time_series, prices, labels,
price_range, lag)
start = 0
end = 700
predictions = []
current_prices = []
while end < len(labels):
predictions.append(predict(start, end, x, labels))
current_prices.append(prices[end - 1])
print end
start += 12
end += 12
pickle.dump(
predictions,
open("/tmp/barrier1_model_test_predictions_" + pair, 'wb'))
pickle.dump(current_prices,
open("/tmp/barrier1_model_test_prices_" + pair, 'wb'))
predictions = pickle.load(
open("/tmp/barrier1_model_test_predictions_" + pair, 'rb'))
current_prices = pickle.load(
open("/tmp/barrier1_model_test_prices_" + pair, 'rb'))
return predictions, current_prices
help='Print debug info')
subparsers = parser.add_subparsers(dest='command')
training = subparsers.add_parser(
'train', help="Trains the model, '-nc' skip cleaning & '-ni' skip IDF")
training.add_argument('-nc',
action='store_true',
help="Skips the Data Cleaning Process")
training.add_argument('-d', type=str, help="Enter the name of the dataset")
training.add_argument('-t',
type=int,
help="Enter the rows to be truncated")
prediction = subparsers.add_parser(
'predict', help="Predicts the output, 'main.py -P number_of_tags url'")
prediction.add_argument('n', type=int)
prediction.add_argument('url', type=str)
args = parser.parse_args()
if args.command == 'train':
if args.t and args.d is not None:
train(args.nc, args.d, args.t)
elif args.t is None and args.d is None:
train(args.nc, 'articles.csv', -1)
elif args.t is None:
train(args.nc, args.d, -1)
elif args.d is None:
train(args.nc, 'articles.csv', args.t)
elif args.command == 'predict':
predict(args.n, args.url)
def test_predict(self):
no_of_predicted_calls, shape = predict('srkw_cnn.h5',
'../datasets/test_srkw/calls')
self.assertEqual(no_of_predicted_calls, 3)
self.assertEqual(shape, (3, 1))
def main():
List = name()
print(List)
predict(List)
def index():
oeng = str(request.get_json(force=True)["eng"])
eng = clean(oeng)
prediction = predict(eng)
print(prediction)
intent = prediction["intent"]
slots = proc_slots(prediction["slots"], eng)
if "slots" in request.get_json(force=True):
for k in request.get_json(force=True)["slots"]:
slots[k] = request.get_json(force=True)["slots"][k]
if intent == "mkdir":
if "dirname" in slots:
if "dirloc" in slots:
return jsonify({
"cmd":
["cd " + slots["dirloc"], "mkdir " + slots["dirname"]]
})
else:
return jsonify({"cmd": ["mkdir " + slots["dirname"]]})
if "dirname" not in slots:
return jsonify({
"res": "What do you want the directory to be named?",
"slot": "dirname"
})
elif intent == "rmdir":
if "dirname" in slots:
if "dirloc" in slots:
return jsonify({
"cmd":
["cd " + slots["dirloc"], "rmdir " + slots["dirname"]]
})
else:
return jsonify({"cmd": ["rmdir " + slots["dirname"]]})
if "dirname" not in slots:
return jsonify({
"res": "What is the name of the folder?",
"slot": "dirname"
})
elif intent == "cd":
if "dirname" not in slots:
return jsonify({
"res": "Which directory do you want to move to?",
"slot": "dirname"
})
else:
return jsonify({"cmd": ["cd " + slots["dirname"]]})
elif intent == "ls":
if "dirname" not in slots:
return jsonify({"cmd": ["ls"]})
else:
return jsonify({"cmd": ["cd " + slots["dirname"], "ls"]})
elif intent == "touch":
if "filname" not in slots:
return jsonify({
"res": "What should be the name of this new file?",
"slot": "filname"
})
else:
if "dirloc" not in slots:
return jsonify({"cmd": ["touch " + slots["filname"]]})
else:
return jsonify({
"cmd":
["cd " + slots["dirloc"], "touch " + slots["filname"]]
})
elif intent == "man":
if "comname" not in slots:
return jsonify({
"res": "Which command do you want to know about?",
"slot": "comname"
})
else:
return jsonify({"cmd": ["man " + slots["comname"]]})
elif intent == "cdback":
return jsonify({"cmd": ["cd .."]})
# elif intent == "mv":
# if "mv_source" in slots and "mv_dest" in slots:
# return jsonify({
# "cmd": ["mv " + slots["mv_source"] + " " + slots["mv_dest"]]
# })
# elif "mv_source" not in slots:
# return jsonify({
# "res": "Which file do you wish to move?"
# })
# elif "mv_dest" not in slots:
# return jsonify({
# "res": "Where do you want this file to be moved?"
# })
return "no intent detected"
def test_predict(self):
no_of_predicted_calls, shape = model_predict.predict("preprocess_mag_scipy_Srkws.h5", "content/datasets/val_srkw/calls")
self.assertEqual(no_of_predicted_calls, 44)
self.assertEqual(shape, (44, 1))
