-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathchange_point_detection.py
152 lines (117 loc) · 7.81 KB
/
change_point_detection.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
import matplotlib.pyplot as plt
import ruptures as rpt
import json
import numpy as np
import time
from sys import argv
import demjson
# No pre-set true change point
def predict_display(signal, computed_chg_pts=None):
if type(signal) != np.ndarray:
# Try to get array from Pandas dataframe
signal = signal.values
if signal.ndim == 1:
signal = signal.reshape(-1, 1) # Reshape the signal to n rows, 1 column
n_samples, n_features = signal.shape
# Set all options
figsize = (10, 2 * n_features) # figure size
color = "k" # color of the lines indicating the computed_chg_pts
linewidth = 3 # linewidth of the lines indicating the computed_chg_pts
linestyle = "--" # linestyle of the lines indicating the computed_chg_pts
fig, axarr = plt.subplots(n_features, figsize=figsize, sharex=True)
if n_features == 1:
axarr = [axarr]
for axe, sig in zip(axarr, signal.T):
# plot s
axe.plot(range(n_samples), sig)
# vertical lines to mark the computed_chg_pts
if computed_chg_pts is not None:
for bkp in computed_chg_pts:
if bkp != 0 and bkp < n_samples:
axe.axvline(x=bkp - 0.5,
color=color,
linewidth=linewidth,
linestyle=linestyle)
fig.tight_layout()
return fig, axarr
# Process json string to np.ndarray signal and record the time
def signal_process(json_str: str) -> (np.ndarray, list):
signal_dic = json.loads(json_str)
sort_dic = {}
for key in sorted(signal_dic.keys()):
sort_dic[key] = signal_dic[key]
s_v = list(sort_dic.values()) # signal value
signal = np.array(s_v)
t_v = [int(t) for t in list(sort_dic.keys())] # time value
time_list = []
for timestamp in t_v:
time_local = time.localtime(timestamp)
time_list.append(time.strftime("%Y-%m-%d %H:%M:%S", time_local))
return signal, time_list
# Choose an algorithm to predict the result of change point list
# n_bkps: number of change point
def bkp_predict(signal: np.ndarray, n_bkps: int) -> list:
algo = rpt.Dynp(model='l2').fit(signal)
result = algo.predict(n_bkps=n_bkps)
# algo = rpt.Window(width=10, model='l2').fit(signal)
# result = algo.predict(n_bkps=n_bkps)
# algo = rpt.Binseg(model='l2').fit(signal)
# result = algo.predict(n_bkps=n_bkps)
# algo = rpt.BottomUp(model='l2').fit(signal)
# result = algo.predict(n_bkps=n_bkps)
# algo = rpt.Omp(min_size=2).fit(signal)
# result = algo.predict(n_bkps=n_bkps)
return result
# The time of the change points
def bkp_time(result: list, time_list: list) -> list:
bkp_time_list = []
for i in result[:-1]:
bkp_time_list.append(time_list[i])
return bkp_time_list
# Find change point in one signal data flow
# [time_start, time_end] = getAbruptChangeRange(int timepoint, data[] )
def getAbruptChangeRange(data: str) -> list:
signal, time_list = signal_process(data)
result = bkp_predict(signal, 2)
# print(bkp_time(result, time_list))
predict_display(signal, result)
plt.show()
return bkp_time(result, time_list)
def computeDelta(signal: np.ndarray) -> float:
upper = 0.9 * signal.max()
below = 1.1 * signal.min()
delta = upper - below
return delta
def main(filename: str):
with open(filename) as file:
for line in file:
signal, time_list = signal_process(line)
result = bkp_predict(signal, 2)
print('异常点为：', bkp_time(result, time_list))
print('时间序列的 delta 为：', computeDelta(signal))
print('=' * 100)
predict_display(signal, result)
plt.show()
if __name__ == '__main__':
# # -------------- 若直接读取本地的json文件，用这段代码 --------------------------
# log_file = r"C:\Users\w50005335\Desktop\data.json"
# main(log_file)
# # ------------------------------------------------------------------------------
# ---------- 若从 java 程序里传递 json 文件路径参数，用这段代码 ----------------
log_file = argv[1]
main(log_file)
# ------------------------------------------------------------------------------
# # ------------ 若从 java 程序里传递单独的一段 json 字符串，用这段代码 ----------
# json_data = demjson.decode(argv[1])
# json_str = json.dumps(json_data)
#
# time_start, time_end = getAbruptChangeRange(json_str)
# print(time_start)
# print(time_end)
# # ------------------------------------------------------------------------------
# # ------------ 若直接分析一段单独的 json 字符串，用这段代码 --------------------
# json_str = '{"1561913400":4.4196485E9,"1561881600":4.6552146E9,"1561942800":7.1144233E9,"1561957740":8.0242033E9,"1561885800":5.4144666E9,"1561875600":4.3789404E9,"1561879800":4.5012844E9,"1561953540":1.20566057E10,"1561932600":5.0253384E9,"1561917600":4.4634163E9,"1561936800":3.82557338E9,"1561926600":4.14572186E9,"1561947540":1.50036183E10,"1561891800":4.6486825E9,"1561888800":4.6535875E9,"1561945800":1.50554573E10,"1561876800":5.11072E9,"1561922400":4.5116058E9,"1561941600":6.2651423E9,"1561903200":5.6121114E9,"1561907400":5.2482509E9,"1561956540":8.8836639E9,"1561880400":4.8173763E9,"1561884600":4.5765663E9,"1561952340":1.56685435E10,"1561933800":3.43927859E9,"1561912200":4.08181632E9,"1561916400":4.7291612E9,"1561927800":3.84530688E9,"1561894800":4.8031032E9,"1561890600":4.6913101E9,"1561930800":3.98551373E9,"1561887600":4.742913E9,"1561873800":4.6286981E9,"1561923600":4.6096256E9,"1561940400":4.5303542E9,"1561902000":4.9305027E9,"1561906200":4.4903306E9,"1561944600":1.33641196E10,"1561915800":4.7595223E9,"1561911600":4.7395077E9,"1561938000":4.14502912E9,"1561883400":4.5694152E9,"1561951140":1.41220925E10,"1561955340":9.1284716E9,"1561901400":4.4303498E9,"1561886400":5.1259244E9,"1561893600":4.3023104E9,"1561949340":1.47891354E10,"1561897800":4.3074744E9,"1561909800":4.7807718E9,"1561943400":9.4593096E9,"1561882200":5.43742E9,"1561920600":5.0325064E9,"1561924800":4.28210432E9,"1561905600":5.332137E9,"1561914600":6.3439882E9,"1561958340":7.2600468E9,"1561935000":3.8026921E9,"1561954140":1.14995845E10,"1561929000":4.5839877E9,"1561895400":4.7916211E9,"1561899600":4.4913316E9,"1561939200":3.74853146E9,"1561948140":1.50451845E10,"1561910400":4.05289779E9,"1561918800":4.8468285E9,"1561878000":4.7890181E9,"1561900200":4.8259589E9,"1561889400":5.0664131E9,"1561892400":4.3474191E9,"1561896600":6.0725443E9,"1561942200":6.0869704E9,"1561908600":4.03533338E9,"1561921800":5.9515044E9,"1561881000":4.8309176E9,"1561885200":5.4829824E9,"1561946400":1.48851507E10,"1561904400":4.6488059E9,"1561875000":4.6899021E9,"1561957140":8.3945047E9,"1561932000":3.85433523E9,"1561936200":4.3083776E9,"1561917000":5.080833E9,"1561894200":4.9764168E9,"1561898400":4.5028285E9,"1561879200":5.8956759E9,"1561903800":6.2376724E9,"1561888200":4.5363456E9,"1561945200":1.39520707E10,"1561891200":5.6361405E9,"1561926000":5.4089774E9,"1561941000":4.5894738E9,"1561884000":4.5772462E9,"1561876200":4.5364209E9,"1561937400":3.85036211E9,"1561912800":5.20975E9,"1561893000":4.5637059E9,"1561897200":5.7151227E9,"1561933200":4.3959757E9,"1561952940":1.46748672E10,"1561902600":4.525825E9,"1561887000":4.6441011E9,"1561923000":5.4042327E9,"1561890000":4.6868777E9,"1561906800":4.11002138E9,"1561944000":1.20529613E10,"1561927200":4.6994734E9,"1561946940":1.54137979E10,"1561938600":3.92263757E9}'
# time_start, time_end = getAbruptChangeRange(json_str)
# print(time_start)
# print(time_end)
# # -------------------------------------------------------------------------------