-
Notifications
You must be signed in to change notification settings - Fork 37
/
rsi_divergence_finder.py
323 lines (270 loc) · 16.6 KB
/
rsi_divergence_finder.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
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
import pandas as pd
from helpers.calculus_helper import *
import logging
from datetime import datetime
from scipy import stats
logger = logging.getLogger(__name__)
RSI_COLUMN = 'rsi'
BASE_COLUMN = 'C'
TIME_COLUMN = 'T'
ANGLE_LIMIT = 45.0 # Limit for angle of divergence lines
def calc_percentage_increase(original, new):
increase = (new - original) / original
return increase * 100
# cur_candle_idx - index of the candle to which we compare candles in the past to find divergences
def get_rsi_divergences(df, tf, cur_candle_idx=-1):
divergences = []
cur_candle = df.iloc[cur_candle_idx]
cur_rsi_change = calc_percentage_increase(df.iloc[-2][RSI_COLUMN],
cur_candle[RSI_COLUMN])
# 'cur_base_value' is the close price here
cur_base_value_time = cur_candle[TIME_COLUMN]
cur_base_value = cur_candle[BASE_COLUMN]
cur_base_value_rsi = cur_candle[RSI_COLUMN]
# 'candles_to_compare' - Candles in the past to which we compare 'cur_candle' and look for divergences
# We skip the most recent 21 candles because divergence signals formed among 21 (or less) candles are not that strong
# We get the other 55 candles before that
candles_to_compare = df[df[TIME_COLUMN] < cur_base_value_time - pd.Timedelta(minutes=tf.value[0] * 21)]
candles_to_compare = candles_to_compare.tail(55)
candles_to_compare_len = candles_to_compare.shape[0]
if candles_to_compare is None:
return divergences
# The rest is RSI divergence detection part
# Some things are hardcoded there, those are the numbers that I find to be more accurate
# Feel free to play around with those numbers
# In the following block, we check if there is bullish divergence
if cur_base_value_rsi <= 37 and cur_rsi_change < 0:
bullish_divs = pd.DataFrame()
for idx, (past_candle_idx, past_candle) in enumerate(candles_to_compare.iterrows()):
try:
past_base_value = past_candle[BASE_COLUMN]
past_base_value_rsi = past_candle[RSI_COLUMN]
past_base_value_time = past_candle[TIME_COLUMN]
if past_base_value_rsi > 32:
continue
is_bullish = False
base_value_change = calc_percentage_increase(past_base_value,
cur_base_value)
rsi_change = calc_percentage_increase(past_base_value_rsi,
cur_base_value_rsi)
df_in_period = df[(past_base_value_time <= df[TIME_COLUMN]) & (df[TIME_COLUMN] <= cur_base_value_time)]
seconds = (df_in_period[TIME_COLUMN] - datetime(1970, 1, 1)).dt.total_seconds()
slope, intercept, r_value, p_value, std_err = stats.linregress(seconds,
df_in_period[
BASE_COLUMN])
if rsi_change >= 6 and base_value_change <= 0 and slope < 0 and pow(r_value, 2) > 0.3:
is_bullish = True
if is_bullish \
and does_any_value_cross_down(df,
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time,
diff=1.05,
value_column=RSI_COLUMN) is False \
and does_any_value_cross_down(df,
past_base_value,
past_base_value_time,
cur_base_value,
cur_base_value_time,
diff=1.03,
value_column=BASE_COLUMN) is False \
and get_angle(
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time,
tf=tf) <= ANGLE_LIMIT:
bullish_divs = bullish_divs.append(past_candle)
except Exception as e:
logging.exception(str(e))
for index, div in bullish_divs.iterrows():
divergences.append({'start_dtm': div[TIME_COLUMN],
'end_dtm': cur_base_value_time,
'rsi_start': div[RSI_COLUMN],
'rsi_end': cur_base_value_rsi,
'price_start': div[BASE_COLUMN],
'price_end': cur_base_value,
'type': 'bullish'})
# In the following block, we check if there is bearish divergence
elif cur_base_value_rsi >= 63 and 0 < cur_rsi_change:
bearish_divs = pd.DataFrame()
for idx, (past_candle_idx, past_candle) in enumerate(candles_to_compare.iterrows()):
try:
past_base_value_rsi = past_candle[RSI_COLUMN]
if past_base_value_rsi < 68:
continue
past_base_value = past_candle[BASE_COLUMN]
past_base_value_time = past_candle[TIME_COLUMN]
is_bearish = False
base_value_change = calc_percentage_increase(past_base_value,
cur_base_value)
rsi_change = calc_percentage_increase(past_base_value_rsi, cur_base_value_rsi)
df_in_period = df[(past_base_value_time <= df[TIME_COLUMN]) & (df[TIME_COLUMN] <= cur_base_value_time)]
seconds = (df_in_period[TIME_COLUMN] - datetime(1970, 1, 1)).dt.total_seconds()
slope, intercept, r_value, p_value, std_err = stats.linregress(seconds,
df_in_period[
BASE_COLUMN])
if rsi_change <= -6 and 0 <= base_value_change and slope > 0 and pow(r_value, 2) > 0.3:
is_bearish = True
if is_bearish \
and does_any_value_cross_up(df,
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time,
diff=1.05,
value_column=RSI_COLUMN) is False \
and does_any_value_cross_up(df,
past_base_value,
past_base_value_time,
cur_base_value,
cur_base_value_time,
diff=1.03,
value_column=BASE_COLUMN) is False \
and get_angle(
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time, tf=tf) <= ANGLE_LIMIT:
bearish_divs = bearish_divs.append(past_candle)
except Exception as e:
logging.exception(str(e))
for index, div in bearish_divs.iterrows():
divergences.append({'start_dtm': div[TIME_COLUMN],
'end_dtm': cur_base_value_time,
'rsi_start': div[RSI_COLUMN],
'rsi_end': cur_base_value_rsi,
'price_start': div[BASE_COLUMN],
'price_end': cur_base_value,
'type': 'bearish'})
# In the following block, we check if there is hidden bearish divergence
if 50 < cur_base_value_rsi <= 70 and cur_rsi_change > 0:
h_bearish_divs = pd.DataFrame()
for idx_lcl, (past_candle_idx, past_candle) in enumerate(candles_to_compare.iterrows()):
try:
if idx_lcl in [0, candles_to_compare_len - 1]:
continue
past_base_value = past_candle[BASE_COLUMN]
past_base_value_rsi = past_candle[RSI_COLUMN]
if candles_to_compare.iloc[idx_lcl - 1][RSI_COLUMN] < \
past_base_value_rsi > \
candles_to_compare.iloc[idx_lcl + 1][RSI_COLUMN]:
if not (50 < past_base_value_rsi < 65):
continue
past_base_value_time = past_candle[TIME_COLUMN]
is_bearish = False
base_value_change = calc_percentage_increase(past_base_value,
cur_base_value)
rsi_change = calc_percentage_increase(past_base_value_rsi,
cur_base_value_rsi)
df_in_period = df[
(past_base_value_time <= df[TIME_COLUMN]) & (df[TIME_COLUMN] <= cur_base_value_time)]
seconds = (df_in_period[TIME_COLUMN] - datetime(1970, 1, 1)).dt.total_seconds()
slope, intercept, r_value, p_value, std_err = stats.linregress(seconds,
df_in_period[BASE_COLUMN])
slope2, intercept2, r_value2, p_value2, std_err2 = stats.linregress(seconds,
df_in_period[
RSI_COLUMN])
if rsi_change >= 6 and base_value_change < 0 and slope < 0 < slope2 and pow(r_value, 2) > 0.3:
is_bearish = True
if is_bearish \
and does_any_value_cross_up(df,
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time,
diff=1.05,
value_column=RSI_COLUMN) is False \
and does_any_value_cross_up(df,
past_base_value,
past_base_value_time,
cur_base_value,
cur_base_value_time,
diff=1.03,
value_column=BASE_COLUMN) is False \
and get_angle(
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time, tf=tf) <= ANGLE_LIMIT:
h_bearish_divs = h_bearish_divs.append(past_candle)
except Exception as e:
logging.exception(str(e))
continue
for index, div in h_bearish_divs.iterrows():
divergences.append({'start_dtm': div[TIME_COLUMN],
'end_dtm': cur_base_value_time,
'rsi_start': div[RSI_COLUMN],
'rsi_end': cur_base_value_rsi,
'price_start': div[BASE_COLUMN],
'price_end': cur_base_value,
'type': 'h_bearish'})
# In the following block, we check if there is hidden bullish divergence
elif 30 < cur_base_value_rsi <= 50 and cur_rsi_change < 0:
h_bullish_divs = pd.DataFrame()
for idx_lcl, (past_candle_idx, past_candle) in enumerate(candles_to_compare.iterrows()):
try:
if idx_lcl in [0, candles_to_compare_len - 1]:
continue
past_base_value = past_candle[BASE_COLUMN]
past_base_value_rsi = past_candle[RSI_COLUMN]
if candles_to_compare.iloc[idx_lcl - 1][RSI_COLUMN] > \
past_base_value_rsi < \
candles_to_compare.iloc[idx_lcl + 1][RSI_COLUMN]:
if not (40 < past_base_value_rsi < 55):
continue
past_base_value_time = past_candle[TIME_COLUMN]
is_bullish = False
base_value_change = calc_percentage_increase(past_base_value,
cur_base_value)
rsi_change = calc_percentage_increase(past_base_value_rsi,
cur_base_value_rsi)
df_in_period = df[
(past_base_value_time <= df[TIME_COLUMN]) & (df[TIME_COLUMN] <= cur_base_value_time)]
seconds = (df_in_period[TIME_COLUMN] - datetime(1970, 1, 1)).dt.total_seconds()
slope, intercept, r_value, p_value, std_err = stats.linregress(seconds,
df_in_period[BASE_COLUMN])
slope2, intercept2, r_value2, p_value2, std_err2 = stats.linregress(seconds,
df_in_period[RSI_COLUMN])
if rsi_change <= -6 and 0 < base_value_change and slope > 0 > slope2 and pow(r_value,
2) > 0.3:
is_bullish = True
if is_bullish \
and does_any_value_cross_down(df,
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time,
diff=1.05,
value_column=RSI_COLUMN) is False \
and does_any_value_cross_down(df,
past_base_value,
past_base_value_time,
cur_base_value,
cur_base_value_time,
diff=1.03,
value_column=BASE_COLUMN) is False \
and get_angle(
past_base_value_rsi,
past_base_value_time,
cur_base_value_rsi,
cur_base_value_time, tf=tf) <= ANGLE_LIMIT:
h_bullish_divs = h_bullish_divs.append(past_candle)
except Exception as e:
logging.exception(str(e))
continue
for index, div in h_bullish_divs.iterrows():
divergences.append({'start_dtm': div[TIME_COLUMN],
'end_dtm': cur_base_value_time,
'rsi_start': div[RSI_COLUMN],
'rsi_end': cur_base_value_rsi,
'price_start': div[BASE_COLUMN],
'price_end': cur_base_value,
'type': 'h_bullish'})
return divergences
def get_all_rsi_divergences(df, tf):
all_divergences = []
for idx in range(df.shape[0]):
all_divergences += get_rsi_divergences(df, tf, idx)
return all_divergences