def label_data():
p = excel_reader.get_data(DATA_FROM, DATA_TO)
np.random.seed(int(time.time()))
#plt.plot(p)
b, s = init_b_s_lognormal(p)
index = comp_index_matrix(p)
comp_loss(b, s, index)
plt.plot(p / 4000 - 1)
db, ds = grad_b_s(b, s, index)
#plt.plot(ds * 10)
#plt.show()
lb, ls = gradient_descent(b, s, index, STEPS, LEARNING_RATE)
plt.plot(lb)
plt.show()
def label_data():
p = excel_reader.get_data(DATA_FROM, DATA_TO,
'D:\python\projdata\data\\1m.xlsx')
log_price = np.log(p)
#plt.plot(p)
#plt.show()
#plt.plot(log_price)
minutes1d = range(DATA_LENGTH)
minutes = np.array([minutes1d]).reshape(-1, 1)
am, bm = linear_regression(minutes, log_price)
a = am[0, 0]
b = bm[0]
minute_line = range(DATA_LENGTH)
minute_line = np.array([minutes1d]).reshape(-1, 1)
line = minute_line * a + b
#plt.plot(line)
#plt.show()
lin_reg = log_price - line
lin_reg = lin_reg[:, 0]
plt.plot(lin_reg)
#conv_array = np.ones((43200,), dtype=float)/43200.
conv_array = np.ones((40960, ), dtype=float) / 40960.
ma = np.convolve(lin_reg, conv_array, 'same')
plt.plot(ma)
plt.show()
without_month_avg = lin_reg - ma
plt.plot(without_month_avg)
plt.show()
convolutions_long(without_month_avg)
conv_array_10k = np.ones((10240, ), dtype=float) / 10240.
ma10k = np.convolve(without_month_avg, conv_array_10k, 'same')
without_10d_avg = without_month_avg - ma10k
plt.plot(without_10d_avg)
plt.show()
convolutions_short(without_10d_avg)
#plt.plot(ma)
#convolutions(lin_reg)
#min - 7000 - 0.16
#min - 22000 - 0.16
#0 - 38000
#+ - 80000
#0 - 118000
#+ - 70000
#0 - 188000
#+ - 60000
#0 - 248000
#semi-period - 70000
period = 125000.0
amplitude = 0.18
zero = 238000.0
def generate():
p = excel_reader.get_data(DATA_FROM, DATA_TO,
'D:\python\projdata\data\\1m.xlsx')
log_price = np.log(p)
plt.plot(p)
plt.show()
plt.plot(log_price)
plt.show()
features, out = generate_features(log_price, DATA_FROM + 1921,
DATA_TO - 60)
features_mean = np.mean(features, axis=0)
print('means:', features_mean)
print(features)
print(out)
plt.plot(features)
plt.show()
plt.plot(out)
plt.show()
def label_data():
p = excel_reader.get_data(DATA_FROM, DATA_TO, 'D:\python\projdata\data\\1m.xlsx')
log_price = np.log(p)
#plt.plot(p)
topology = [14, 100, 100, 50, 20, 2]
nn = NeuralNet(topology)
#nn = nn_factory.read('net_11_7d')
#index = comp_index_matrix(p)
#b, s = comp_b_s(nn, p, index)
#plt.plot(index[0,:])
#comp_loss(b, s, index)
#plt.plot(p / 4000 - 1)
#db, ds = grad_b_s(b, s, index)
#plt.plot(ds * 10)
#plt.show()
lb, ls = gradient_descent(nn, p, log_price, STEPS, LEARNING_RATE)
plt.plot(lb)
plt.show()
nn.save('net_final')
ma_320,
ma_640,
ma_1280,
ma_2560,
ma_5120,
))
features = (features - log_price[i, 0]) * 400.0
return features
DATA_FROM, DATA_TO = 0, 100000
TRAINING_LENGTH = DATA_TO - DATA_FROM
nn = nn_factory.read('net_batch_32')
#nn = nn_factory.read('D://python/TorgovecNets/net_w_c_loss_-40')
#p = excel_reader.get_data(DATA_FROM, DATA_TO, 'D:\python\projdata\data\\btc30m.xlsx')
p = excel_reader.get_data(DATA_FROM, DATA_TO,
'D:\python\projdata\data\\1m_short.xlsx')
#index = comp_index_matrix(p)
log_price = np.log(p)
inf_prices = 181
print('TEST')
wallet_btc = 1.0
wallet_usd = 0.0
sell_flag = True
trades_btc = 0
trades_usd = 0
activations = np.zeros((TRAINING_LENGTH, 2))
hold_btc = 0
hold_usd = 0
sell_btc = np.zeros((TRAINING_LENGTH, 1))
buy_btc = np.zeros((TRAINING_LENGTH, 1))
#treshhold = random.random()