def predict_json():
content = request.json
try:
params = content['sample']
except KeyError:
return abort(
400,
"Key 'sample' was not found in the request",
)
if isinstance(params, str):
params = process_str_params(params)
elif isinstance(params, (tuple, list)):
params = process_collection_params(params)
else:
abort(
400,
"Invalid data for prediction. Expected 13 numerical objects",
)
model_path = "./model/trained_model.pkl"
if not os.path.exists(model_path):
train_model()
model = pickle.load(open(model_path, 'rb'))
predict = model.predict(params)
predict = {'class': str(predict[0])}
return jsonify(predict)
def heart(params):
params = params.split(',')
params = [float(num) for num in params]
model_path = "./model/trained_model.pkl"
if not os.path.exists(model_path):
train_model()
model = pickle.load(open(model_path, 'rb'))
params = np.array(params).reshape(1, -1)
predict = model.predict(params)
return str(predict)
def multi_nativebayes_train(model_data):
#
class_eachtoken_likelihood = {}
vocabulary = model_data.get_vocabulary()
for class_label in model_data.get_class_labels():
class_eachtoken_likelihood[class_label] = {}
for voc in vocabulary:
class_eachtoken_likelihood[class_label][voc] = 0
logprior = {}
vocabularyCount = model_data.get_vocabularyCount()
class_eachtoken_count = model_data.get_class_eachtoken_count()
for class_label in model_data.get_class_labels():
total_class_token = model_data.get_total_class_token()
logprior[class_label] = math.log(total_class_token[class_label] /
vocabularyCount)
for word in vocabulary:
if (class_eachtoken_count[class_label][word] == 0):
class_eachtoken_likelihood[class_label][word] = 0
else:
class_eachtoken_likelihood[class_label][word] = math.log(
class_eachtoken_count[class_label][word] /
total_class_token[class_label])
train_model_data = train_model(logprior, class_eachtoken_likelihood,
vocabulary, model_data.get_class_labels())
return train_model_data
def main():
# ============================ Setup ===================================
configure_logger()
config = get_config()
# ========================== Load & prepare input data ==============================
logging.info("Processing base data...")
base_data_df = process_data(base_data_dir=config.base_data_dir,
input_data_file=config.input_data_file)
# ========================= Train model ===============================
logging.info("Training model and evaluating...")
model = train_model(base_data_df, config.population_tests_dir, grid_search)
# ========================== Export Results ===================================
logging.info("Saving model...")
save_model(model, config.out_file)
def main():
final_model = train_model()
predictions_df = predict(final_model)
print(predictions_df)
driver_alerts = set(
predictions_df[predictions_df['alert'] == True]['driver_id'])
for driver in driver_alerts:
send_sms(driver)
#if __name__ == '__main__':
# main()
# read datasets ---------------------------
imdb, roidb, valroidb, output_dir = read_datasets(args)
# Remove roidb entries that have no usable RoIs.
roidb = filter_roidb(roidb)
valroidb = filter_roidb(valroidb)
# config gpu
tfconfig = tf.ConfigProto(allow_soft_placement=True)
tfconfig.gpu_options.allow_growth = True
# -------------------build model-----------------
# load networks ---------------------------
model_net = load_network(args.net)
# pre-training weights
pre_net_file_path = load_net_weights(args.net)
model_train, model_vol = build_model(imdb, model_net, pre_net_file_path)
# -------------------build model-----------------
# --------------------training-------------------
train_model(imdb,
roidb,
model_train,
valroidb,
model_vol,
output_dir,
max_iters=args.max_iters)
# --------------------training-------------------
if not os.path.exists('data/raw_data.csv'):
process = CrawlerProcess(get_project_settings())
process.crawl(LyricsSpider)
process.start()
else:
print(
'\033[33m' +
"WARNING: There is already data saved in the directory, if you want to collect new ones delete the raw_data.csv file in the data folder."
+ '\033[0;0m')
print('----------------------------')
print(' TRAINING THE MODEL ')
print('----------------------------')
train_df, test_df = prepare_data()
train_model(API_KEY, train_df)
print('----------------------------')
print(' TESTING THE MODEL ')
print('----------------------------')
test_model(API_KEY, test_df)
print('----------------------------')
print(' PIPELINE FINISHED ')
print('----------------------------')
while True:
answer = input('Do you want to launch the application? [Y/N] ')
if answer.lower().strip() == 'y':
os.system('cmd /k "streamlit run main.py"')
elif answer.lower().strip() == 'n':