# 기간 필터링
training_data = training_data[(training_data['date'] >= '2018-01-01') &
(training_data['date'] <= '2018-01-31')]
training_data = training_data.dropna()
# 차트 데이터 분리
features_chart_data = ['date', 'open', 'high', 'low', 'close', 'volume']
chart_data = training_data[features_chart_data]
# 학습 데이터 분리
features_training_data = [
'open_lastclose_ratio', 'high_close_ratio', 'low_close_ratio',
'close_lastclose_ratio', 'volume_lastvolume_ratio',
'close_ma5_ratio', 'volume_ma5_ratio',
'close_ma10_ratio', 'volume_ma10_ratio',
'close_ma20_ratio', 'volume_ma20_ratio',
'close_ma60_ratio', 'volume_ma60_ratio',
'close_ma120_ratio', 'volume_ma120_ratio'
]
training_data = training_data[features_training_data]
# 비 학습 투자 시뮬레이션 시작
policy_learner = PolicyLearner(
stock_code=stock_code, chart_data=chart_data, training_data=training_data,
min_trading_unit=1, max_trading_unit=3)
policy_learner.trade(balance=10000000,
model_path=os.path.join(
settings.BASE_DIR,
'models/{}/model_{}.h5'.format(stock_code, model_ver)))
parser.add_argument('COIN', type=str, help="coin type?")
parser.add_argument('BALANCE', type=int, help="initial balance?")
args = parser.parse_args()
MODEL = args.MODEL
COIN = args.COIN
BALANCE = args.BALANCE
log_dir = os.path.join(settings.BASE_DIR, 'logs/%s' % COIN)
timestr = settings.get_time_str()
file_handler = logging.FileHandler(filename=os.path.join(
log_dir, "%s_%s.log" % (COIN, timestr)),
encoding='utf-8')
stream_handler = logging.StreamHandler()
file_handler.setLevel(logging.DEBUG)
stream_handler.setLevel(logging.INFO)
logging.basicConfig(format="%(message)s",
handlers=[file_handler, stream_handler],
level=logging.DEBUG)
policy_learner = PolicyLearner(coin_code=COIN,
coin_chart=None,
training_data=None,
min_trading_unit=1,
max_trading_unit=3)
policy_learner.trade(balance=BALANCE,
model_path=os.path.join(
settings.BASE_DIR,
'models/{}/model_{}.h5'.format(COIN, MODEL)))
def learnFunc(self):
if self.code is None or self.df is None:
return
self.change_value.emit(ZERO)
# 데이터 전처리
code = self.code
chart_data = self.df
prep_data = data_manager.preprocess(chart_data)
training_data = data_manager.build_training_data(prep_data)
training_data = training_data.dropna()
# 차트데이터 분리
feature_chart_data = ['date', 'open', 'high', 'low', 'close', 'volume']
chart_data = training_data[feature_chart_data]
# emit
self.change_value.emit(TWENTY_FIVE)
# 학습데이터 분리
feature_chart_data = [
'open_lastclose_ratio',
'high_close_ratio',
'low_close_ratio',
'close_lastclose_ratio',
'volume_lastvolume_ratio',
'close_ma5_ratio',
'volume_ma5_ratio',
'close_ma10_ratio',
'volume_ma10_ratio',
'close_ma20_ratio',
'volume_ma20_ratio',
'close_ma60_ratio',
'volume_ma60_ratio',
'close_ma120_ratio',
'volume_ma120_ratio',
]
training_data = training_data[feature_chart_data]
# 정책 신경망을 파일로 저장
self.createFolder('model')
mdir = os.path.join(settings.BASE_DIR, 'model')
self.createFolder(os.path.join(mdir, code))
model_dir = os.path.join(mdir, code)
model_path = os.path.join(model_dir, 'model%s.h5' % code)
# model_path 경로가 없으면 학습모델을 해당 dir에 만들어서 학습
# model_path가 있으면 해당 모델 선택 후 예측
print(model_path)
# emit
self.change_value.emit(FIFTY)
if not os.path.isfile(model_path):
start_time = time.time()
policy_learner = PolicyLearner(stock_code=code,
chart_data=chart_data,
training_data=training_data,
fig=self.fig,
canvas=self.canvas,
min_trading_unit=1,
max_trading_unit=2,
delayed_reward_threshold=0.2,
lr=0.001)
policy_learner.fit(balance=10000000,
num_epoches=200,
discount_factor=0,
start_epsilon=0.5)
end_time = time.time()
policy_learner.policy_network.save_model(model_path)
print("LearningTime: {} sec".format(end_time - start_time))
else:
start_time = time.time()
policy_learner = PolicyLearner(stock_code=code,
chart_data=chart_data,
training_data=training_data,
fig=self.fig,
canvas=self.canvas,
min_trading_unit=1,
max_trading_unit=2)
end_time = time.time()
print("LearningTime: {} sec".format(end_time - start_time))
policy_learner.trade(balance=1000000,
model_path=os.path.join(
model_dir, 'model%s.h5' % (code)))
# emit
self.change_value.emit(A_HUNDRED)
if not os.path.isdir(model_dir):
os.makedirs(model_dir)
#model_path = os.path.join(model_dir, 'model_%s.h5' % timestr)
model_path = os.path.join(model_dir, 'model_%s.h5' % STOCK_CODE)
policy_learner.policy_network.save_model(model_path)
if SIMULATION:
chart_data, data = prepare_data(STOCK_CODE, MARKET_CODE, SIMULATION_START_DATE, SIMULATION_END_DATE)
policy_learner = PolicyLearner(
stock_code=STOCK_CODE, chart_data=chart_data, training_data=data,
min_trading_unit=MIN_TRADING_UNIT, max_trading_unit=MAX_TRADING_UNIT)
policy_learner.trade(balance=INITIAL_BALANCE, num_epoches=NUM_EPOCHS,max_memory=MAX_MEMORY,
discount_factor=DISCOUNT_FACTOR, start_epsilon=START_EPSILON,
model_path=os.path.join(
settings.BASE_DIR,
'models/{}/model_{}.h5'.format(STOCK_CODE, STOCK_CODE)))
# 'models/{}/model_{}.h5'.format(STOCK_CODE, model_ver)))
=======
from agent import DQNAgent
from environment import Environment
from keras import backend as K
K.clear_session()
def update_delayed_reward(episode_buffer, last_reward):
rewarded_buffer = []
for state, action, reward, next_state, done in episode_buffer:
reward = last_reward # 마지막 reward 로 앞선 action 의 reward 모두 change
rewarded_buffer.append((state, action, reward, next_state, done))
'close_ma10_ratio',
'volume_ma10_ratio',
'close_ma20_ratio',
'volume_ma20_ratio',
'close_ma60_ratio',
'volume_ma60_ratio',
'close_ma120_ratio',
'volume_ma120_ratio',
]
training_data = training_data[features_training_data]
# 비학습 투자 시뮬레이션 시작
policy_learner = PolicyLearner(stock_code=stock_code,
chart_data=chart_data,
training_data=training_data,
min_trading_unit=1,
max_trading_unit=10)
policy_learner.trade(balance=52777,
model_path=os.path.join(
settings.BASE_DIR, 'model\{}\model_{}.h5'.format(
stock_code, model_ver)),
cur_stock_value=9 * 27650,
init_stocks=9)
# 정책 신경망을 파일로 저장. 추가적인 학습을 수행하여 모델을 새로 저장하고 싶다면 코드 블록을 그대로 두면 된다
# model_dir = os.path.join(settings.BASE_DIR, 'model/%s' % stock_code)
# if not os.path.isdir(model_dir):
# os.makedirs(model_dir)
# model_path = os.path.join(model_dir, 'model_%s.h5' % timestr)
# policy_learner.policy_network.save_model(model_path)
