-
Notifications
You must be signed in to change notification settings - Fork 3
/
result.py
executable file
·149 lines (132 loc) · 5.75 KB
/
result.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
from sklearn.model_selection import train_test_split
from sklearn.svm import SVC
from sklearn.neighbors import KNeighborsClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.naive_bayes import GaussianNB
from sklearn.datasets import make_classification
from sklearn.model_selection import StratifiedKFold
from sklearn.metrics import plot_roc_curve
from sklearn.metrics import auc
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import os
import sys
import warnings
warnings.filterwarnings("ignore")
#Process ROC AUC Score with KFold
def raScore(datasetPath, ax, axTitle):
df = pd.read_csv(datasetPath)
#df = df.sample(frac=1, random_state=30) #Shuffle rows
data = df.values
X = data[:,1:]
y = data[:,0]
cv = StratifiedKFold(n_splits=3)
gnb = GaussianNB()
lr = LogisticRegression(random_state=0, multi_class='auto')
knn = KNeighborsClassifier(n_neighbors=3)
svc = SVC(random_state=42)
classifiers = [gnb,lr,knn,svc]
names = ["GNB","LR","KNN","SVC"]
colors = ["red","blue","green","brown"]
lmean_tpr = []
lmean_auc = []
mean_fpr = np.linspace(0, 1, 100)
for classifier, name, color in zip(classifiers, names, colors):
tprs = []
aucs = []
for i, (train, test) in enumerate(cv.split(X, y)):
classifier.fit(X[train], y[train])
viz = plot_roc_curve(classifier, X[test], y[test],
name='ROC fold {}{}{}'.format(i," ",name),
alpha=0.5, lw=1, ax=ax, color=color, linestyle=':')
interp_tpr = np.interp(mean_fpr, viz.fpr, viz.tpr)
interp_tpr[0] = 0.0
tprs.append(interp_tpr)
aucs.append(viz.roc_auc)
mean_tpr = np.mean(tprs, axis=0)
mean_tpr[-1] = 1.0
mean_auc = np.round(auc(mean_fpr, mean_tpr),2)
lmean_tpr.append(mean_tpr)
lmean_auc.append(mean_auc)
std_auc = np.round(np.std(aucs),2)
ax.plot(mean_fpr, mean_tpr, color=color,
label=r'Mean ROC '+name+' (AUC = %0.2f $\pm$ %0.2f)' % (mean_auc, std_auc),
lw=2, alpha=.8)
ax.plot([0, 1], [0, 1], linestyle='--', lw=2, color='purple',label='Chance', alpha=.8)
gmean_tpr = np.round(np.mean(lmean_tpr, axis=0),2)
gmean_auc = np.round(np.mean(lmean_auc),2)
ax.plot(mean_fpr, gmean_tpr, color="black",
label=r'Mean ROC (AUC = %0.2f)' % (gmean_auc),
lw=2, alpha=.8)
ax.set(xlim=[-0.05, 1.05], ylim=[-0.05, 1.05],
title=axTitle)
ax.legend(loc="lower right")
return gmean_auc
#plt.show()
#Process and plot individually
def individualRaScore(testName, datasetName):
datasetInfo1 = "datasets-input/"+datasetName
datasetInfo2 = "datasets-test/fss/"+datasetName
datasetInfo3 = "datasets-test/fsx/"+datasetName
f, ax = plt.subplots(1, 3)
f.set_size_inches(30, 10.5, forward=True)
plt.suptitle("ROC curve comparison "+datasetName)
fullAuc = raScore(datasetInfo1,ax[0],"Full features set "+datasetName) # full
fsAuc = raScore(datasetInfo2,ax[1],"Top 10 features "+datasetName) # fs
fs2Auc = raScore(datasetInfo3,ax[2],"Stepwise feature selection set "+datasetName) # fs2
#plt.show()
plt.savefig("results/"+testName+"/"+testName+"-"+datasetName+".png")
#plt.savefig("results/"+testName+"/"+testName+"-"+datasetName+".svg")
return [round(fullAuc,2), round(fsAuc,2), round(fs2Auc,2)]
#Process and plot by groups
def groupedRaScore(testName, datasetName, ax):
datasetInfo1 = "datasets-input/"+datasetName
datasetInfo2 = "datasets-test/fss/"+datasetName
datasetInfo3 = "datasets-test/fsx/"+datasetName
fullAuc = raScore(datasetInfo1,ax[0],"Full features set "+datasetName) # full
fsAuc = raScore(datasetInfo2,ax[1],"Top 10 features "+datasetName) # fs
fs2Auc = raScore(datasetInfo3,ax[2],"Stepwise feature selection set "+datasetName) # fs2
return [round(fullAuc,2), round(fsAuc,2), round(fs2Auc,2)]
#Process and plot results
def processResults(datasets, pGrouped=True):
fullAucList = []
fsAucList = []
fs2AucList = []
i = 0
size = 4
for datasetName in datasets:
print(datasetName)
if(pGrouped): #Plot grouped
if(i%size==0):
f, axs = plt.subplots(size, 3)
plt.subplots_adjust(left=0.02, bottom=0.01, right=0.99, top=0.99, wspace=0.07, hspace=0.11)
f.set_size_inches(26.4, 8*6)
[fullAuc, fsAuc, fs2Auc] = groupedRaScore(testName, datasetName, axs[i%size])
if(i%size==(size-1)): #Save figure
plt.savefig("results/"+testName+"/"+"full"+str(i+1)+".png")
#plt.savefig("results/"+testName+"/"+"full"+str(i+1)+".svg")
plt.savefig("results/"+testName+"/"+"full"+str(i+1)+".pdf")
else: #Plot individual
[fullAuc, fsAuc, fs2Auc] = individualRaScore(testName, datasetName)
#Apend results
fullAucList.append(fullAuc)
fsAucList.append(fsAuc)
fs2AucList.append(fs2Auc)
i+=1
return [fullAucList, fsAucList, fs2AucList]
if __name__ == "__main__":
testName = "top10-20"
datasets = ["ds1.csv", "ds2.csv", "ds3.csv", "ds4.csv", "ds5.csv", "ds6.csv",
"ds7.csv", "ds8.csv", "ds9.csv", "ds10.csv", "ds11.csv", "ds12.csv"]#,
#"ds13.csv", "ds14.csv", "ds15.csv"]
#Create result folder
if(not os.path.exists("results")):
os.mkdir("results/")
if(not os.path.exists("results/"+testName)):
os.mkdir("results/"+testName)
#Process and plot
results = processResults(datasets)
#Save results
df = pd.DataFrame(np.array(results), columns=datasets)
df.to_csv ("results/"+testName+".csv", index = False, header=True)