forked from omarmmousa/DiabeticRetinopathy
-
Notifications
You must be signed in to change notification settings - Fork 0
/
plots.py
127 lines (104 loc) · 4.83 KB
/
plots.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
import matplotlib.pyplot as plt
import json
import numpy as np
data = {'full': {'svm': {}, 'knn': {}, 'xgb': {}},
'limited': {'svm': {}, 'knn': {}, 'xgb': {}},
'binary': {'svm': {}, 'knn': {}, 'xgb': {}}}
with open('../data/results.dat', 'r') as f:
lines = [l.strip('\n') for l in f.readlines()]
for i, l in enumerate(lines):
if i % 2 == 0:
header = l.split(',')
header[1] = ','.join(header[1:])
next_line = [float(l) for l in lines[i+1].split(',')[:-1]]
d = json.loads(header[1].replace("'", "\""))
if 'kernel' in d:
data[header[0]]['svm'][str(d)] = next_line
elif 'n_neighbors' in d:
data[header[0]]['knn'][str(d)] = next_line
else:
data[header[0]]['xgb'][str(d)] = next_line
def keyWithMaxValue(d, i):
k = list(d.keys())
v = list(d.values())
return k[v.index(max(v, key=lambda x: x[i]))]
def iterateVariations():
print("Printing model performance by metric for each model for each data subset...\nFormat: metric name \t parameters \t metric value\n")
for s in ['full', 'limited', 'binary']:
print("'{0}'".format(s))
for m in ['KNN', 'SVM', 'XGB']:
print("\t\t{0}:".format(m))
for metric, index in [('Accuracy', 2), ('Precision', 3), ('Recall', 4),
('F-score', 5)]:
key = keyWithMaxValue(data[s][m.lower()], index)
value = data[s][m.lower()][str(key)][index]
print("\t\t\t{0}\t{1}\t{2}".format(metric, key, value))
iterateVariations()
# Plotting training / prediction time for all models
"""
bin_knn_time_x = [i[0] for i in data['binary']['knn'].values()]
bin_knn_time_y = [i[1] for i in data['binary']['knn'].values()]
limited_knn_time_x = [i[0] for i in data['limited']['knn'].values()]
limited_knn_time_y = [i[1] for i in data['limited']['knn'].values()]
full_knn_time_x = [i[0] for i in data['full']['knn'].values()]
full_knn_time_y = [i[1] for i in data['full']['knn'].values()]
bin_svm_time_x = [i[0] for i in data['binary']['svm'].values()]
bin_svm_time_y = [i[1] for i in data['binary']['svm'].values()]
limited_svm_time_x = [i[0] for i in data['limited']['svm'].values()]
limited_svm_time_y = [i[1] for i in data['limited']['svm'].values()]
full_svm_time_x = [i[0] for i in data['full']['svm'].values()]
full_svm_time_y = [i[1] for i in data['full']['svm'].values()]
bin_xgb_time_x = [i[0] for i in data['binary']['xgb'].values()]
bin_xgb_time_y = [i[1] for i in data['binary']['xgb'].values()]
limited_xgb_time_x = [i[0] for i in data['limited']['xgb'].values()]
limited_xgb_time_y = [i[1] for i in data['limited']['xgb'].values()]
full_xgb_time_x = [i[0] for i in data['full']['xgb'].values()]
full_xgb_time_y = [i[1] for i in data['full']['xgb'].values()]
fig, ax = plt.subplots()
ax.scatter(bin_knn_time_x, bin_knn_time_y, color='b')
ax.scatter(limited_knn_time_x, limited_knn_time_y, color='g')
ax.scatter(full_knn_time_x, full_knn_time_y, color='r')
ax.scatter(bin_svm_time_x, bin_svm_time_y, color='b', marker='s')
ax.scatter(limited_svm_time_x, limited_svm_time_y, color='g', marker='s')
ax.scatter(full_svm_time_x, full_svm_time_y, color='r', marker='s')
ax.scatter(bin_xgb_time_x, bin_xgb_time_y, color='b', marker='^')
ax.scatter(limited_xgb_time_x, limited_xgb_time_y, color='g', marker='^')
ax.scatter(full_xgb_time_x, full_xgb_time_y, color='r', marker='^')
import matplotlib.patches as patches
from matplotlib.lines import Line2D
red = patches.Patch(color='red', label='Full Dataset')
green = patches.Patch(color='green', label='Limited Dataset')
blue = patches.Patch(color='blue', label='Binary Dataset')
knn_dot = Line2D([], [], markersize='15', color='white', label='KNN', marker='o')
svm_square = Line2D([], [], markersize='15', color='white', label='SVM', marker='s')
xgb_carrot = Line2D([], [], markersize='15', color='white', label='XGBoost', marker='^')
plt.legend(handles=[red, green, blue, knn_dot, svm_square, xgb_carrot])
plt.xlabel("Train Time")
plt.ylabel("Predict Time")
plt.title("Training/Predicting Time by Data Subset and Model Type")
plt.show()
"""
"""
### KNN results sorted by n_neighbors -- BINARY
# BINARY
s = sorted(data['binary']['knn'], key=lambda k: list(json.loads(k.replace("'", "\"")).values())[0])
print("\nKNN results sorted by n_neighbors -- BINARY")
for k in s:
print(data['binary']['knn'][k])
bin_knn_time_x = [i[0] for i in s]
bin_knn_time_y = [i[1] for i in s]
# LIMITED
s = sorted(data['limited']['knn'], key=lambda k: list(json.loads(k.replace("'", "\"")).values())[0])
print("\nKNN results sorted by n_neighbors -- LIMITED")
for k in s:
print(data['limited']['knn'][k])
limited_knn_time_x = [i[0] for i in s]
limited_knn_time_y = [i[1] for i in s]
# FULL
s = sorted(data['full']['knn'], key=lambda k: list(json.loads(k.replace("'", "\"")).values())[0])
print("\nKNN results sorted by n_neighbors -- FULL")
for k in s:
print(data['full']['knn'][k])
full_knn_time_x = [i[0] for i in s]
full_knn_time_y = [i[1] for i in s]
"""