def test_data():
n = 100
t = np.arange(n) # [0 .. n-1]
# triangle
# x = np.concatenate((np.arange(0,n/2,1), np.arange(n/2,0,-1)))
# step
x = ti.step(t - n/2) * 10
# ramp
# x = ramp(t - n/2)
# normalized random
# x = np.cumsum(np.random.randn(n))
# x = np.sin(8 * np.pi/n * t) + (.1 * t)
# x = 20 * np.sin(2 * 2*np.pi/n * t)
# Add noise:
x = x + 1.2 * np.random.randn(len(x))
assert len(ti.sma(x, 10)) == len(x)
assert len(ti.ema(x, 10)) == len(x)
assert len(ti.linear_fit(x, 10)) == len(x)
assert len(ti.iir_lowpass(x, 2, 10)) == len(x)
assert len(ti.moving_min(x, 10)) == len(x)
assert len(ti.moving_max(x, 10)) == len(x)
assert len(ti.aema(x, 10)) == len(x)
assert len(ti.rate_of_change(x, 20)) == len(x)
assert len(ti.acceleration(x, 20)) == len(x)
def make_current_data(binance, symbol, day_start, day_end):
df = pd.DataFrame(columns=[
"Time", "Close", "High", "Low", "Volume", "QuoteVolume", "TakerVolume",
"TakerQuoteVolume", "TradeCount", "WMA7", "EMA25", "MA99", "BOLL_UP",
"BOLL_DOWN", "SAR", "MACD", "MACD_SIGNAL", "RSI12"
])
#df = pd.DataFrame(columns=["Time","Open","Close","High","Low","Volume","QuoteVolume","TakerVolume","TakerQuoteVolume","TradeCount","MA7","MA25","MA99","EMA7","EMA25","EMA99","WMA7","WMA25","WMA99","BOLL_UP","BOLL_DOWN","VWAP","TEMA","SAR","MACD","MACD_SIGNAL","RSI6","RSI12","RSI24","K","D","J","OBV","WR","DI+","DI-","ADX","MTM","EMV"])
df.loc[:, "Time"] = pd.to_datetime(df["Time"])
df = df.set_index("Time")
#to match the minimum size
pad_multiplier = 300 // const.TIME_LENGTH
if day_start == 0 and day_end == 0:
klines = binance.get_klines(symbol, const.TIME_LENGTH * pad_multiplier)
df = df.append(klines)
else:
for i in range(day_start, day_end, -1):
klines = binance.get_historical_klines(
symbol,
str(i + 1) + " days ago UTC",
str(i) + " days ago UTC")
df = df.append(klines)
try:
df = df.astype("float64")
#mpf.plot(df, type="candle")
#df.loc[:, "MA7"] = technical_indicators.ma(df,7)
#df.loc[:, "MA25"] = technical_indicators.ma(df,25)
df.loc[:, "MA99"] = technical_indicators.ma(df, 99)
#df.loc[:, "EMA7"] = technical_indicators.ema(df,7)
df.loc[:, "EMA25"] = technical_indicators.ema(df, 25)
#df.loc[:, "EMA99"] = technical_indicators.ema(df,99)
df.loc[:, "WMA7"] = technical_indicators.wma(df, 7)
#df.loc[:, "WMA25"] = technical_indicators.wma(df,25)
#df.loc[:, "WMA99"] = technical_indicators.wma(df,99)
df.loc[:, "BOLL_UP"], df.loc[:,
"BOLL_DOWN"] = technical_indicators.boll(
df, 21)
#df.loc[:, "VWAP"] = technical_indicators.vwap(df,14)
#df.loc[:, "TEMA"] = technical_indicators.tema(df,9)
df.loc[:, "SAR"] = technical_indicators.sar(df)
df.loc[:,
"MACD"], df.loc[:,
"MACD_SIGNAL"] = technical_indicators.macd(df)
#df.loc[:, "RSI6"] = technical_indicators.rsi(df,6)
df.loc[:, "RSI12"] = technical_indicators.rsi(df, 12)
#df.loc[:, "RSI24"] = technical_indicators.rsi(df,24)
#df.loc[:, "K"],df.loc[:, "D"],df.loc[:, "J"] = technical_indicators.kdj(df)
#df.loc[:, "OBV"] = technical_indicators.obv(df)
#df.loc[:, "WR"] = technical_indicators.wr(df,14)
#df.loc[:, "DI+"],df.loc[:, "DI-"],df.loc[:, "ADX"] = technical_indicators.dmi(df)
#df.loc[:, "MTM"] = technical_indicators.mtm(df)
#df.loc[:, "EMV"] = technical_indicators.emv(df)
#df.loc[:, "PL"] = technical_indicators.pl(df,12)
#df = df.append(klines)
except Exception as e:
print('Exception : {}'.format(e))
df = None
return df
def _main():
print("test_indicator = {}".format(ti.test_indicator('XBB.TO')))
t = np.arange(-5, 5, 1)
s = ti.step(t)
r = ti.ramp(t)
plt.plot(t, s, marker='x', linestyle='None', label='step')
plt.plot(t,
r,
marker='o',
markerfacecolor='None',
linestyle='None',
label='ramp')
plt.legend()
plt.show()
t = np.linspace(0, 4 * np.pi, 50)
s = np.sin(t)
c = np.cos(t)
o = ti.cross_over(s, c)
u = ti.cross_under(s, c)
fig = plt.figure()
ax = fig.add_subplot(211)
ax.plot(t, s, marker='.', label='sin')
ax.plot(t, c, marker='x', label='cos')
ax.legend()
ax2 = fig.add_subplot(212)
ax2.plot(t,
o,
marker='^',
markerfacecolor='None',
linestyle='None',
label='cross over')
ax2.plot(t,
u,
marker='v',
markerfacecolor='None',
linestyle='None',
label='cross under')
ax2.legend()
plt.show()
n = 100
t = np.arange(n) # [0 .. n-1]
# triangle
# x = np.concatenate((np.arange(0,n/2,1), np.arange(n/2,0,-1)))
# step
x = ti.step(t - n / 2) * 10
# ramp
#x = ramp(t - n/2)
# normalized random
# x = np.cumsum(np.random.randn(n))
# x = np.sin(8 * np.pi/n * t) + (.1 * t)
# x = 20 * np.sin(2 * 2*np.pi/n * t)
# Add noise:
x = x + 1.2 * np.random.randn(len(x))
fig = plt.figure()
ax = fig.add_subplot(211)
ax.plot(t, x, 'o', label='raw')
ax.plot(t, ti.sma(x, 10), label='sma')
ax.plot(t, ti.ema(x, 10), label='ema')
ax.plot(t, ti.linear_fit(x, 10), label='linear_fit')
ax.plot(t, ti.iir_lowpass(x, 1, 10), label='iir_lowpass')
ax.plot(t, ti.moving_min(x, 10), label='moving_min')
ax.plot(t, ti.moving_max(x, 10), label='moving_max')
ax.plot(t, ti.aema(x, 10), label='aema')
ax.grid(True)
ax.legend()
ax2 = fig.add_subplot(212)
ax2.plot(t, ti.rate_of_change(x, 20), 'x', label='rate_of_change')
ax2.plot(t, ti.acceleration(x, 20), 'o--', label='acceleration')
ax2.grid(True)
ax2.legend()
plt.show()
def make_current_data(binance, symbol, day_start, day_end):
sub_pair_symbols = const.PAIR_SYMBOLS.copy()
sub_pair_symbols.remove(symbol)
df = pd.DataFrame(columns=[
"Time", "Close", "High", "Low", "Volume", "QuoteVolume", "TakerVolume",
"TakerQuoteVolume", "TradeCount", "WMA7", "EMA25", "MA99", "BOLL_UP",
"BOLL_DOWN", "SAR", "MACD", "MACD_SIGNAL", "RSI12"
])
#df = pd.DataFrame(columns=["Time","Open","Close","High","Low","Volume","QuoteVolume","TakerVolume","TakerQuoteVolume","TradeCount","MA7","MA25","MA99","EMA7","EMA25","EMA99","WMA7","WMA25","WMA99","BOLL_UP","BOLL_DOWN","VWAP","TEMA","SAR","MACD","MACD_SIGNAL","RSI6","RSI12","RSI24","K","D","J","OBV","WR","DI+","DI-","ADX","MTM","EMV"])
df.loc[:, "Time"] = pd.to_datetime(df["Time"])
df = df.set_index("Time")
sub_pair_dict = {}
for i in sub_pair_symbols:
sub_pair_dict[i] = pd.DataFrame(columns=[
"Close", "High", "Low", "Volume", "QuoteVolume", "TakerVolume",
"TakerQuoteVolume", "TradeCount"
])
#to match the minimum size
pad_multiplier = 128 // const.TIME_LENGTH
if day_start == 0 and day_end == 0:
klines = binance.get_klines(symbol, const.TIME_LENGTH * pad_multiplier)
df = df.append(klines)
for i in sub_pair_symbols:
sub_klines = binance.get_klines(i,
const.TIME_LENGTH * pad_multiplier)
sub_pair_dict[i] = sub_pair_dict[i].append(sub_klines)
else:
for i in range(day_start, day_end, -64):
klines = binance.get_historical_klines(
symbol,
str(i + 64) + " days ago UTC",
str(i) + " days ago UTC")
df = df.append(klines)
for j in sub_pair_symbols:
sub_klines = binance.get_historical_klines(
j,
str(i + 64) + " days ago UTC",
str(i) + " days ago UTC")
sub_pair_dict[j] = sub_pair_dict[j].append(sub_klines)
df = df.astype("float64")
#mpf.plot(df, type="candle")
#df.loc[:, "MA7"] = technical_indicators.ma(df,7)
#df.loc[:, "MA25"] = technical_indicators.ma(df,25)
df.loc[:, "MA99"] = technical_indicators.ma(df, 99)
#df.loc[:, "EMA7"] = technical_indicators.ema(df,7)
df.loc[:, "EMA25"] = technical_indicators.ema(df, 25)
#df.loc[:, "EMA99"] = technical_indicators.ema(df,99)
df.loc[:, "WMA7"] = technical_indicators.wma(df, 7)
#df.loc[:, "WMA25"] = technical_indicators.wma(df,25)
#df.loc[:, "WMA99"] = technical_indicators.wma(df,99)
df.loc[:,
"BOLL_UP"], df.loc[:,
"BOLL_DOWN"] = technical_indicators.boll(df, 21)
#df.loc[:, "VWAP"] = technical_indicators.vwap(df,14)
#df.loc[:, "TEMA"] = technical_indicators.tema(df,9)
df.loc[:, "SAR"] = technical_indicators.sar(df)
df.loc[:, "MACD"], df.loc[:, "MACD_SIGNAL"] = technical_indicators.macd(df)
#df.loc[:, "RSI6"] = technical_indicators.rsi(df,6)
df.loc[:, "RSI12"] = technical_indicators.rsi(df, 12)
#df.loc[:, "RSI24"] = technical_indicators.rsi(df,24)
#df.loc[:, "K"],df.loc[:, "D"],df.loc[:, "J"] = technical_indicators.kdj(df)
#df.loc[:, "OBV"] = technical_indicators.obv(df)
#df.loc[:, "WR"] = technical_indicators.wr(df,14)
#df.loc[:, "DI+"],df.loc[:, "DI-"],df.loc[:, "ADX"] = technical_indicators.dmi(df)
#df.loc[:, "MTM"] = technical_indicators.mtm(df)
#df.loc[:, "EMV"] = technical_indicators.emv(df)
#df.loc[:, "PL"] = technical_indicators.pl(df,12)
#df = df.append(klines)
for i in sub_pair_symbols:
df.insert(len(df.columns), i, sub_pair_dict[i]["Close"])
df = df.fillna(0.1)
df = df.astype("float64")
return df