def step(self, action):
# 0: Sit
# 1: But
# 2: Sell
assert (action in (0, 1, 2))
# State transition
next_state = getState(self.data, self.t + 1, self.window_size + 1)
# Reward
if action == 0:
reward = 0
elif action == 1:
reward = 0
self.agent.inventory.append(self.data[self.t])
print("Buy: " + formatPrice(self.data[self.t]))
else:
if len(self.agent.inventory) > 0:
bought_price = self.agent.inventory.pop(0)
profit = self.data[self.t] - bought_price
reward = max(profit, 0)
self.total_profit += profit
print("Sell: " + formatPrice(self.data[self.t]) +
" | Profit: " + formatPrice(profit))
else:
reward = 0 # try to sell, but con't do
done = True if self.t == len(self.data) - 2 else False
self.t += 1
return next_state, reward, done, {}
def render(self, mode='human', close=False):
if self.current_step >= self.price_history.get_length():
return
if close:
if self.viewer is not None:
os.kill(self.viewer.pid, signal.SIGKILL)
else:
close_price = self.trading_days[self.current_step].get_close()
print(self.trading_days[self.current_step].get_date().strftime(
'%m/%d/%Y') + " Cash " + formatPrice(self.cash) +
" | Holdings: " +
formatPriceOfHoldings(self.inventory, close_price) +
" | Balance: " + formatPrice(self.balance))
agent.inventory = []
for t in range(l):
action = agent.ActRbot(state)
print('****************************')
print('Action is {}'.format(f.decipheraction(action)))
# hold
next_state = f.getState(data, t + 1, window_size + 1)
reward = 0
if action == 1: # buy
agent.inventory.append(data[t + 1])
print("Buy: " + f.formatPrice(data[t + 1]))
print('****************************')
elif action == 2 and len(agent.inventory) > 0: # sell
bought_price = agent.inventory.pop(0)
reward = max(data[t + 1] - bought_price, 0)
total_profit += data[t + 1] - bought_price
print("Sell: " + f.formatPrice(data[t + 1]) + " | Profit: " +
f.formatPrice(data[t + 1] - bought_price))
print('****************************')
done = True if t == l - 1 else False
agent.MemoryRbot(state, action, reward, next_state, done)
state = next_state
if done:
data_size = len(data) - 1
batch_size = 100
gain = 0
broker.trades_list = []
reward = 0
state = getState(data, 0, state_size + 1)
for t in range(data_size):
action = broker.act(state)
if action == 1 and broker.portfolio >= data[t]: # buy
broker.trades_list.append(data[t])
broker.portfolio = broker.portfolio - data[t]
print("Bought: " + formatPrice(data[t]) + "| Portfolio Value: " + formatPrice(broker.portfolio))
reward = 0
elif action == 2 and len(broker.trades_list) > 0: # sell
buying_price = broker.trades_list.pop(0)
gain += data[t] - buying_price
if (gain > 0):
reward = 1
else:
reward = 0
broker.portfolio = broker.portfolio + data[t]
print("Sold: " + formatPrice(data[t]) + " | Financial Gain: " + formatPrice(
data[t] - buying_price) + "| Portfolio Value: " + formatPrice(broker.portfolio))
for e in range(nb_epochs + 1):
print("Epoch Number:" + str(e) + "/" + str(nb_epochs))
state = getState(data, 0, state_size + 1)
reward = 0
gain = 0
broker.trades_list = []
broker.portfolio = 100000
for t in range(data_size - state_size + 1):
action = broker.act(state)
if action == 1 and broker.portfolio >= data[t]: # buy
broker.trades_list.append(data[t])
broker.portfolio = broker.portfolio - data[t]
print("Buy: " + formatPrice(data[t]) + "| Portfolio Value: " +
formatPrice(broker.portfolio) + "| Inventory Size:" +
str(len(broker.trades_list)))
reward = 0
elif action == 2 and len(broker.trades_list) > 0: # sell
buying_price = broker.trades_list.pop(0)
broker.portfolio = broker.portfolio + data[t]
#reward = max(data[t] - buying_price, 0)
#gain += data[t] - buying_price
if (gain > 0):
reward = 1
else:
reward = 0
print("Sell: " + formatPrice(data[t]) + " | Profit: " +