-
Notifications
You must be signed in to change notification settings - Fork 4
/
RQ-2.py
213 lines (203 loc) · 9.01 KB
/
RQ-2.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
#!/usr/bin/env python
# coding: utf-8
import pandas as pd
import numpy as np
import sys
import time
import os
# get_ipython().run_line_magic('matplotlib', 'inline')
import matplotlib
import matplotlib.pyplot as plt
import scipy.stats as st
directories = os.listdir("Dataset")
algorithm_name = ["LogisticRegression", "TreeRegressor", "Decision_Tree", "Discriminant_Analysis", "SVM"]
dataset_name = ["census", "credit", "bank", "compas"]
sensitive_attr = ["gender", "race", "age"]
searches = ["random", "mutation", "coverage"]
key_num_inp = {}
key_max_AOD = {}
key_max_TPR = {}
search_name = {}
with open("Results/RQ2.csv", 'w') as f:
f.write("name,num_inputs_random,num_inputs_mutation,num_inputs_coverage,|range_AOD_random|,|range_AOD_mutation|,|range_AOD_coverage|,|range_TPR_random|,|range_TPR_mutation|,|range_TPR_coverage|\n")
for drs in directories:
if not drs.startswith("Run"):
continue
for filename in os.listdir("Dataset" + "/" + drs):
if filename.endswith("res.csv"):
search = ""
key = ""
found = False
for s in searches:
if s in filename:
search = s
for an in algorithm_name:
if an in filename:
for ds in dataset_name:
if ds in filename:
for sa in sensitive_attr:
if sa in filename:
key = an + "-" + ds + "-" + sa
found = True
if(found):
break
if(found):
break
if(found):
break
if(found):
break
if found:
df = pd.read_csv("Dataset" + "/" + drs + "/" + filename)
df = df[df["score"] <= 1.0]
df = df[df["AOD"] <= 1.0]
df = df[df["TPR"] <= 1.0]
accuracy = df["score"]
AOD = df["AOD"]
TPR = df["TPR"]
if key not in key_num_inp:
key_num_inp[key] = []
key_max_AOD[key] = []
key_max_TPR[key] = []
search_name[key] = []
key_num_inp[key].append(df.shape[0])
key_max_AOD[key].append(abs(AOD.max() - AOD.min()))
key_max_TPR[key].append(abs(TPR.max() - TPR.min()))
search_name[key].append(search)
for key in key_num_inp:
print(key)
f.write(key)
f.write(",")
num_inputs_data = {}
key_max_AOD_data = {}
key_max_TPR_data = {}
for i in range(len(search_name[key])):
if search_name[key][i] not in num_inputs_data:
num_inputs_data[search_name[key][i]] = []
num_inputs_data[search_name[key][i]].append(key_num_inp[key][i])
key_max_AOD_data[search_name[key][i]] = []
key_max_AOD_data[search_name[key][i]].append(key_max_AOD[key][i])
key_max_TPR_data[search_name[key][i]] = []
key_max_TPR_data[search_name[key][i]].append(key_max_TPR[key][i])
else:
num_inputs_data[search_name[key][i]].append(key_num_inp[key][i])
key_max_AOD_data[search_name[key][i]].append(key_max_AOD[key][i])
key_max_TPR_data[search_name[key][i]].append(key_max_TPR[key][i])
print(num_inputs_data[search])
for search in searches:
num_inputs_data_average = np.mean(num_inputs_data[search])
num_inputs_data_CI = st.t.interval(0.95, len(num_inputs_data[search])-1, loc=num_inputs_data_average, scale=st.sem(num_inputs_data[search]))
try:
f.write(str(round(num_inputs_data_average)) + " (+/- " + str(round(num_inputs_data_average - num_inputs_data_CI[0])) + ")")
except:
f.write("0 (+/- 0)")
f.write(",")
for search in searches:
key_max_AOD_data_average = np.mean(key_max_AOD_data[search])
key_max_AOD_data_CI = st.t.interval(0.95, len(key_max_AOD_data[search])-1, loc=key_max_AOD_data_average, scale=st.sem(key_max_AOD_data[search]))
try:
f.write(str(round(key_max_AOD_data_average * 100,1)) + "\% (+/- " + str(round((key_max_AOD_data_average - key_max_AOD_data_CI[0]) * 100, 1)) + "\%)")
except:
f.write("0 (+/- 0)")
f.write(",")
for search in searches:
key_max_TPR_data_average = np.mean(key_max_TPR_data[search])
key_max_TPR_data_CI = st.t.interval(0.95, len(key_max_TPR_data[search])-1, loc=key_max_TPR_data_average, scale=st.sem(key_max_TPR_data[search]))
try:
f.write(str(round(key_max_TPR_data_average * 100,1)) + "\% (+/- " + str(round((key_max_TPR_data_average - key_max_TPR_data_CI[0]) * 100, 1)) + "\%)")
except:
f.write("0 (+/- 0)")
f.write(",")
f.write("\n")
X_dict = {}
Y_dict = {}
search_dict = {}
# plots
for drs in directories:
if not drs.startswith("Run"):
continue
for filename in os.listdir("Dataset" + "/" + drs):
if filename.endswith("res.csv"):
print(filename)
search = ""
key = ""
found = False
for s in searches:
if s in filename:
search = s
for an in algorithm_name:
if an in filename:
for ds in dataset_name:
if ds in filename:
for sa in sensitive_attr:
if sa in filename:
key = an + "-" + ds + "-" + sa
found = True
if(found):
break
if(found):
break
if(found):
break
if(found):
break
if found:
df = pd.read_csv("Dataset" + "/" + drs + "/" + filename)
AOD = np.array(df["AOD"])
x = []
y = []
max_AOD_over_time = df["AOD"][0]
for a in range(0,14500,1):
try:
df1 = df[df["timer"] >= a]
df1 = df1[df1["timer"] < a + 1]
except TypeError as TE:
pass
if not df1.empty:
max_AOD_a = df1["AOD"].max()
if(max_AOD_over_time < max_AOD_a):
max_AOD_over_time = max_AOD_a
x.append(a+1)
y.append(max_AOD_over_time)
if key not in X_dict:
X_dict[key] = []
Y_dict[key] = []
search_dict[key] = []
X_dict[key].append(x)
Y_dict[key].append(y)
search_dict[key].append(search)
else:
X_dict[key].append(x)
Y_dict[key].append(y)
search_dict[key].append(search)
for key in X_dict:
Y_searach_AOD_data = {}
Y_search_AOD_average = {}
Y_search_AOD_CI = {}
# print(search_dict[key])
for i in range(len(search_dict[key])):
if search_dict[key][i] not in Y_searach_AOD_data:
Y_searach_AOD_data[search_dict[key][i]] = []
Y_searach_AOD_data[search_dict[key][i]].append(Y_dict[key][i])
else:
Y_searach_AOD_data[search_dict[key][i]].append(Y_dict[key][i])
for search in Y_searach_AOD_data:
data = np.array(Y_searach_AOD_data[search])
data_mean = np.mean(data, axis=0)
data_CI = np.array([st.t.interval(0.95, len(data[:,i])-1, loc=np.mean(data[:,i]), scale=st.sem(data[:,i])) for i in range(data.shape[1])])
Y_search_AOD_average[search] = data_mean
Y_search_AOD_CI[search] = np.nan_to_num(data_CI)
plt.figure(dpi=150)
X = X_dict[key]
name = search_dict[key]
for search in searches:
plt.plot(X[0], Y_search_AOD_average[search], label=search)
plt.fill_between(X[0],[y[0] for y in Y_search_AOD_CI[search]],[y[1] for y in Y_search_AOD_CI[search]], alpha=0.1)
plt.xlabel('Time (s)')
plt.ylabel('Average Odds Difference (AOD)')
key = key.replace("_","")
plt.title(key)
plt.legend(loc = "lower right")
print(key)
plt.savefig("Results/" + key + "_AOD_vs_time", dpi=150)
plt.close()