def prepare_data(data):
# for market(0, 3), ema(4, 7), macd(8, 9)
sigmoid = get_sigmoid_ration
sigm0 = sigmoid(data[:, 0])
sigm1 = sigmoid(data[:, 1])
sigm2 = sigmoid(data[:, 2])
sigm3 = sigmoid(data[:, 3])
delta_oc = get_delta(data, 0, 3)
diff1 = get_diff(data[:, 1])
diff2 = get_diff(data[:, 2])
diff3 = get_diff(data[:, 3])
logdiff1 = get_log_diff(data[:, 1])
logdiff2 = get_log_diff(data[:, 2])
logdiff3 = get_log_diff(data[:, 3])
detrend1 = get_delta(data, 3, 4) # close - ema13
detrend2 = get_delta(data, 3, 5) # close - ema26
diff_ema1 = get_diff(data[:, 4])
diff_ema2 = get_diff(data[:, 5])
delta_ema1 = get_delta(data, 4, 5)
delta_ema2 = get_delta(data, 6, 7)
#
return np.array(
np.column_stack(
(sigm0, sigm1, sigm2, sigm3, delta_oc, diff1, diff2, diff3,
logdiff1, logdiff2, logdiff3, detrend1, detrend2, diff_ema1,
diff_ema2, delta_ema1, delta_ema2, data[:, 8], data[:, 9])))
def prepare_data(data):
# for time(0, 6), market(7, 10), ema(11, 14), macd(15, 16)
# for atr(17), cci(18), rsi(19), usdx(20), eurx(21)
#
# delta = get_delta(data, 7, 10)
# sigmoid = get_sigmoid_to_zero
# sigmoid = get_sigmoid_ration
# sigm1 = sigmoid(data[:, 8])
# sigm2 = sigmoid(data[:, 9])
# sigm3 = sigmoid(data[:, 10])
diff1 = get_diff(data[:, 8])
diff2 = get_diff(data[:, 9])
diff3 = get_diff(data[:, 10])
# logdiff1 = get_log_diff(data[:, 8])
# logdiff2 = get_log_diff(data[:, 9])
# logdiff3 = get_log_diff(data[:, 10])
detrend1 = get_delta(data, 10, 11) # close - ema13
detrend2 = get_delta(data, 10, 12) # close - ema26
#
# edelta1 = get_delta(data, 11, 12)
# edelta2 = get_delta(data, 13, 14)
# ediff1 = get_diff(data[:, 11])
# ediff2 = get_diff(data[:, 12])
# elogdiff1 = get_log_diff(data[:, 11])
# elogdiff2 = get_log_diff(data[:, 12])
#
# xdelta = get_delta(data, 20, 21)
xdiff1 = get_diff(data[:, 20])
xdiff2 = get_diff(data[:, 21])
# xlogdiff1 = get_log_diff(data[:, 20])
# xlogdiff2 = get_log_diff(data[:, 21])
return np.array(
np.column_stack((
# data[:, 5:6], # hours and minutes
# data[:, 8:11], # prices (without open)
# delta,
# sigm1, sigm2, sigm3,
diff1,
diff2,
diff3,
# logdiff1, logdiff2, logdiff3,
detrend1,
detrend2,
# data[:, 11:15], # ema's
# edelta1, edelta2,
# ediff1, ediff2,
# elogdiff1, elogdiff2,
# data[:, 15:17], # macd
data[:, 17:20], # atr, cci, rsi
# data[:, 20:22], # usd and eur indexes
# xdelta,
# xdiff1, xdiff2,
# xlogdiff1, xlogdiff2,
)))
def prepare_data(data):
# for time(0, 6), market(7, 10), ema(11, 14), macd(15, 16)
# for atr(17), cci(18), rsi(19), usdx(20), eurx(21)
#----------------------------
# for market(0, 3), ema(4, 7),
# for atr(8), cci(9), rsi(10)
mrkt, ema = range(4), range(4, 8)
delta = get_delta(data, mrkt[0], mrkt[3])
diff1 = get_diff(data[:, mrkt[1]])
diff2 = get_diff(data[:, mrkt[2]])
diff3 = get_diff(data[:, mrkt[3]])
# logdiff1 = get_log_diff(data[:, mrkt[1]])
# logdiff2 = get_log_diff(data[:, mrkt[2]])
# logdiff3 = get_log_diff(data[:, mrkt[3]])
detrend1 = get_delta(data, mrkt[3], ema[0]) # close - ema13
detrend2 = get_delta(data, mrkt[3], ema[1]) # close - ema26
#
ediff1 = get_diff(data[:, ema[0]])
ediff2 = get_diff(data[:, ema[1]])
ediff3 = get_diff(data[:, ema[2]])
# elogdiff1 = get_log_diff(data[:, 11])
# elogdiff2 = get_log_diff(data[:, 12])
# elogdiff3 = get_log_diff(data[:, 13])
return np.array(
np.column_stack((
# data[:, 5:6], # hours and minutes
# data[:, 8:11], # prices (without open)
delta,
diff1,
diff2,
diff3,
# logdiff1, logdiff2, logdiff3,
detrend1,
detrend2,
ediff1,
ediff2,
ediff3,
# elogdiff1, elogdiff2, elogdiff3,
# data[:, 15:17], # macd
data[:, 8:10],
data[:, 10] - 50, # atr, cci, rsi
# data[:, 20:22], # usd and eur indexes
)))