-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCreateIndicatorData.py
182 lines (151 loc) · 7.36 KB
/
CreateIndicatorData.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
from Indicators import *
import logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger("INDICATOR GENERATOR")
def generate_technical_indicators(
df:pd.DataFrame,
price_column:str,
volume_column:str,
volume_seed:float=10000.0):
logger.info("Generating metadata.")
price_data = df[price_column].values
volume_data = df[volume_column].values
# Sice we have only total traded volume in a 24 hour period. We take a
# first difference and use a seed first volume to get by second volume
volume_data = np.cumsum(np.insert(np.diff(volume_data),0,volume_seed))
column_names = ["volume"]
output_data = np.array(volume_data).reshape(1,-1)
# Any time the kline period is multiplied this is to generate some type of
# moving average or a calculation period. This is because our data is
# by second.
kline_period = 60
#------------------------------------------------------------------------#
logger.info("Generate Average True Rating")
atr = average_true_rating(
data=price_data,
window=kline_period,
period=14)
atr_sma = simple_moving_average(atr,60*3)
column_names.extend(["atr","atr_sma"])
output_data = np.append(output_data,atr.reshape(1,-1),axis=0)
output_data = np.append(output_data,atr_sma.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
# Balance of power, its SMA, and the derivative of the SMA
logger.info("Generate Balance of Power")
bop = balance_of_power(
data=price_data,
period=kline_period,
smoothing_period=kline_period*14)
bop_sma = simple_moving_average(bop,60*3)
ddx_degree = 3
ddx_bop = np.subtract(bop_sma,array_shift(bop_sma,ddx_degree))/ddx_degree
column_names.extend(["bop","bop_sma","bop_prime"])
output_data = np.append(output_data,bop.reshape(1,-1),axis=0)
output_data = np.append(output_data,bop_sma.reshape(1,-1),axis=0)
output_data = np.append(output_data,ddx_bop.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Accumulatrion Distribution Line")
adl = accumulation_distribution_line(
price_data=price_data,
volume_data=volume_data,
period=30)
adl = rescale(arr=adl,low=price_data.min(),high=price_data.max())
column_names.append("adl")
output_data = np.append(output_data,adl.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Ulcer Index")
ulcer = ulcer_index(data=price_data,period=kline_period*2)
ulcer_normal = simple_moving_average(ulcer,kline_period*2*8)
ulcer_diff = ulcer_normal-ulcer
column_names.extend(["ulcer","ulcer_normal","ulcer_diff"])
output_data = np.append(output_data,ulcer.reshape(1,-1),axis=0)
output_data = np.append(output_data,ulcer_normal.reshape(1,-1),axis=0)
output_data = np.append(output_data,ulcer_diff.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Stochastic Oscillator")
stochastic = stochastic_oscillator(
data=price_data,
period=kline_period*10)
stochastic_signal = simple_moving_average(stochastic,kline_period*2)
stochastic_diff = stochastic_signal-stochastic
column_names.extend(["stochastic","stochastic_signal","stochastic_diff"])
output_data = np.append(output_data,stochastic.reshape(1,-1),axis=0)
output_data = np.append(output_data,stochastic_signal.reshape(1,-1),axis=0)
output_data = np.append(output_data,stochastic_diff.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Relative Strength Index")
rsi = relative_strength_index(data=price_data,periods=28)
rsi_oversold = np.repeat(30,len(rsi)) - rsi
rsi_overbought = rsi - np.repeat(70,len(rsi))
column_names.extend(["rsi","rsi_oversold","rsi_overbought"])
output_data = np.append(output_data,rsi.reshape(1,-1),axis=0)
output_data = np.append(output_data,rsi_oversold.reshape(1,-1),axis=0)
output_data = np.append(output_data,rsi_overbought.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate MACD")
macd_line,signal,histogram = mean_average_convergance_divergance(
data=price_data,
fast_window=kline_period*4,
slow_window=kline_period*9,
signal_window=kline_period*3)
column_names.extend(["macd_4_9_3","macd_4_9_3_signal"])
output_data = np.append(output_data,macd_line.reshape(1,-1),axis=0)
output_data = np.append(output_data,signal.reshape(1,-1),axis=0)
macd_line,signal,histogram = mean_average_convergance_divergance(
data=price_data,
fast_window=kline_period*12,
slow_window=kline_period*26,
signal_window=kline_period*9)
column_names.extend(["macd_12_26_9","macd_12_26_9_signal"])
output_data = np.append(output_data,macd_line.reshape(1,-1),axis=0)
output_data = np.append(output_data,signal.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Bollinger Bands")
upper,middle,lower,bandwidth,percentB = bollinger_bands(
data=price_data,
window=kline_period*20)
column_names.extend(["bollinger_upper","bollinger_lower","percent_b"])
output_data = np.append(output_data,upper.reshape(1,-1),axis=0)
output_data = np.append(output_data,lower.reshape(1,-1),axis=0)
output_data = np.append(output_data,percentB.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate On Balance Volume")
obv = on_balance_volume(data=volume_data,period=10)
column_names.append("obv")
output_data = np.append(output_data,obv.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
logger.info("Generate Adaptive Moving Average")
kama_short = kaufman_adaptive_moving_average(
data=price_data,
er_period=kline_period*10,
fast_ema=2,
slow_ema=30)
kama_mid = kaufman_adaptive_moving_average(
data=price_data,
er_period=kline_period*10,
fast_ema=2*3,
slow_ema=30*3)
kama_long = kaufman_adaptive_moving_average(
data=price_data,
er_period=kline_period*10,
fast_ema=2*6,
slow_ema=30*6)
column_names.extend(["kama_short","kama_mid","kama_long"])
output_data = np.append(output_data,kama_short.reshape(1,-1),axis=0)
output_data = np.append(output_data,kama_mid.reshape(1,-1),axis=0)
output_data = np.append(output_data,kama_long.reshape(1,-1),axis=0)
#------------------------------------------------------------------------#
return pd.DataFrame(output_data.T,columns=column_names)
def main():
raw_data = pd.read_csv(f"./ticker.csv")
indicators = generate_technical_indicators(
df=raw_data,
price_column="best_ask",
volume_column="total_traded_asset")
work_data = pd.concat(
[raw_data[["timestamp","best_ask","best_bid"]],
indicators],axis=1)
work_data.dropna(how='any',inplace=True)
work_data.to_csv("./indicator_data.csv")
if __name__ == "__main__":
main()