-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path3. AgeInfluence.py
141 lines (99 loc) · 4.67 KB
/
3. AgeInfluence.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
# %%
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from scipy.optimize import curve_fit
from utils import read_data, func, line_styles
from sklearn.metrics import r2_score, mean_squared_error
plt.style.use(['science', 'no-latex'])
plt.rcParams.update({
"figure.figsize": (6, 4)})
def get_failure_rate(s, pipe_data, year):
tt = pipe_data.loc[s['age']]
if tt['ASBUILTLENGTH'] < 5280*1:
s[f'failure_rate_{year}'] = np.nan
s[f'length_{year}'] = np.nan
else:
s[f'failure_rate_{year}'] = s[0] / tt['ASBUILTLENGTH'] * 528000
s[f'length_{year}'] = tt['ASBUILTLENGTH'] / 528000
s[f'failure_number_{year}'] = s[0]
return s
def yearly_failure_rate(break_record, pipe_record, year, considered_bins):
year_break = break_record[break_record['used_time'].dt.year == year]
pipe_record.loc[:, 'age'] = year - pipe_record.INSTALLDATE.dt.year
annual_pipe = pipe_record[pipe_record['age'] > 0]
failure_number = year_break.groupby(['break_age']).size().to_frame()
pipe_length = annual_pipe.groupby(['age']).sum()
pipe_length = pipe_length.reindex(considered_bins, fill_value=0)
failure_number = failure_number.reindex(considered_bins, fill_value=0)
failure_number['age'] = failure_number.index
failure_number = failure_number.apply(
get_failure_rate, pipe_data=pipe_length, year=year, axis=1)
failure_number.drop(columns=[0], inplace=True)
return failure_number
def fitting_curve(bins, failure_rates):
nan_idx = []
for i, arr in enumerate(failure_rates):
if ~np.isfinite(arr) or arr == 0:
nan_idx.append(i)
failure_ratio_used = np.delete(failure_rates, nan_idx)
bin_edges_used = np.delete(bins, nan_idx)
popt, pcov = curve_fit(func, bin_edges_used, failure_ratio_used)
fitted_value = func(bins, *popt)
fitting_evaluation = func(bin_edges_used, *popt)
r2_value = r2_score(failure_ratio_used, fitting_evaluation)
RMSE = np.sqrt(mean_squared_error(failure_ratio_used, fitting_evaluation))
print(f"The material is {material}")
print(f"The parameter is: {popt}")
print(f"The r2 score is: {r2_value}")
print(f"The MSE value: {RMSE}")
return fitted_value
def FR_Age(interested_material):
break_record, _, _, pipe_record = read_data()
break_record.loc[:, 'break_age'] = break_record['break_age'].astype(int)
break_record = break_record[break_record['MATERIAL']
== interested_material]
pipe_record = pipe_record[pipe_record['MATERIAL'] == interested_material]
failures = []
considered_bins = np.arange(0, 101)
for year in range(1990, 2020):
failure = yearly_failure_rate(
break_record, pipe_record, year, considered_bins)
failures.append(failure)
final_failure = pd.concat(failures, axis=1)
considered_name = [f'failure_rate_{i}' for i in range(1990, 2020)]
weighted_name = [f'length_{i}' for i in range(1990, 2020)]
kept_idex = final_failure.index[final_failure[considered_name].isna().sum(
1) < 28]
failure_rate = final_failure[considered_name].loc[kept_idex]
failure_length = final_failure[weighted_name].loc[kept_idex]
weighted_sum = False
if weighted_sum:
weights = failure_length.div(failure_length.sum(axis=1), axis=0)
averaged_failure = failure_rate.mul(np.array(weights))
averaged_failure = averaged_failure.sum(axis=1)
else:
averaged_failure = failure_rate.mean(axis=1, skipna=True)
# averaged_failure = averaged_failure[averaged_failure.values <= 0.25*365]
fitted_value = fitting_curve(
averaged_failure.index.values, averaged_failure.values)
return averaged_failure, fitted_value
if __name__ == '__main__':
line_style, marker_style = line_styles()
for count, material in enumerate(['Cast Iron', 'Ductile Iron', 'Unknown']):
averaged, fitted = FR_Age(material)
averaged.replace(0, np.nan, inplace=True)
plt.scatter(averaged.index[:100], averaged.values[:100],
alpha=0.8, marker=marker_style[count])
plt.plot(averaged.index[:100], fitted[:100],
label=material, linewidth=2, alpha=0.6, linestyle=line_style[count])
plt.ylim(0.02*365, 0.15*365)
plt.legend()
plt.xlabel("Pipe age (years)")
plt.ylabel("FR (No./year/100 miles)")
plt.tight_layout()
plt.savefig("../results/MonthlyPrediction/Cast-DuctileCohortAnalysis.tiff",
dpi=300, bbox_inches='tight')
plt.show()
# Let;s check if sync
# Myabe change it again