def make_obs_ph(name):
return U_b.BatchInput((64, 64), name=name) #64 64
print("Average Reward is %s" %
(env.portfolio.average_profit_per_trade))
if final_test:
env.generate_summary_stats()
with U.make_session(8):
# csv = "/home/adrian/Escritorio/polinex/EURUSD60.csv"
csv = "data/EURUSD60.csv"
env = gym.make('trading-v0')
env.initialise_simulator(csv, trade_period=50, train_split=0.7)
act, train, update_target, debug = deepq.build_train(
make_obs_ph=lambda name: UT.BatchInput(env.observation_space.shape,
name=name),
q_func=model,
num_actions=env.action_space.n,
optimizer=tf.train.AdamOptimizer(learning_rate=5e-4),
)
try:
UT.load_state('./test_model/test_model')
except:
print("nodata")
replay_buffer = ReplayBuffer(50000)
# Create the schedule for exploration starting from 1 (every action is random) down to
# 0.02 (98% of actions are selected according to values predicted by the model).
exploration = LinearSchedule(schedule_timesteps=10000,
initial_p=1.0,
final_p=0.02)
def make_obs_ph(name):
return U_b.BatchInput((32, 32), name=name)
def make_obs_ph(name):
# return U.BatchInput((64, 64), name=name)
return QU.BatchInput((64, 64), name=name)
def make_obs_ph(
name
): # Creates a placeholder for a batch of tensors of a given shape and dtype
return U_b.BatchInput((16, 16), name=name)
def make_obs_ph(name):
return U2.BatchInput((16, 16), name=name)
