velorama/evaluate.py at main · amudide/velorama · GitHub

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
import numpy as np
import pandas as pd
import sklearn
import torch
import matplotlib.pyplot as plt
from matplotlib.pyplot import figure
from sklearn import metrics
import os
import glob
import math
import pathlib

def lor(x, y):
    return [a + b for a, b in zip(x, y)]

def evaluate(trial, velo=False, proba=False, dyna=False, log=False, gstd=True, penalty='H', hidden=[32], lr=0.01, lag=5, max_iter=10000, lam_ridge=0):

    folder = trial + '-' + str(velo) + '-' + str(proba) + '-' + str(dyna) + '-' + str(log) + '-' + str(gstd) + '-' + str(max_iter) + '-' + str(hidden) + '-' + str(lag) + '-' + str(penalty) + '-' + str(lam_ridge) + '-' + str(lr)

    lag = int(folder.split('-')[-4])

    lams = [x[0].split('/')[-1] for x in os.walk('img/' + folder)][1:]
    lams = [float(x) for x in lams]

    mat1 = pd.read_csv('img/' + folder + "/1.0/preds.csv", index_col = 0)
    gt = mat1['y_true'].tolist()

    gc = []
    g = 0

    for lam in lams:
        gclam = []
        try:
            mat = torch.load('img/' + folder + "/" + str(lam) + '/lag.pt')
            mat = mat.detach()
            for l in range(lag):
                g = len(mat[:,:,l])
                gclam.append(mat[:,:,l].numpy().flatten())
            gc.append(gclam)
        except:
            pass


    ## produce matrices for each lag

    rgc = []

    for i in range(lag):
        rmats = []
        for j in range(len(gc)):
            nonzero = 0
            for val in gc[j][i]:
                if val > 1e-8:
                    nonzero += 1
            pnzero = nonzero / len(gc[j][i])

            if pnzero >= 0.01 and pnzero <= 0.95:
                rmats.append(gc[j][i])
        rgc.append(rmats)


    for l in range(lag):
        for i in range(len(rgc[l]) - 2, -1, -1):
            rgc[l][i] = lor(rgc[l][i], rgc[l][i + 1])

    pathlib.Path('eval/' + folder).mkdir(parents=True, exist_ok=True)

    for l in range(lag):
        for i in range(g):
            rgc[l][0][g*i + i] = 0

    for l in range(lag):

        np.savetxt('eval/' + folder + '/gc_' + str(lag - l) + '.csv', np.array(rgc[l][0]).reshape(100, 100), delimiter=",")

        fpr, tpr, threshold = metrics.roc_curve(gt, rgc[l][0])
        roc_auc = metrics.auc(fpr, tpr)

        plt.title('Receiver Operating Characteristic')
        plt.plot(fpr, tpr, 'b', label = 'AUC = %0.3f' % roc_auc)
        plt.legend(loc = 'lower right')
        plt.xlim([-0.05, 1.05])
        plt.ylim([-0.05, 1.05])
        plt.ylabel('True Positive Rate')
        plt.xlabel('False Positive Rate')

        plt.savefig('eval/' + folder + '/roc_' + str(lag - l) + '.png', bbox_inches='tight')
        plt.clf()

    for l in range(lag):
        precision, recall, threshold = metrics.precision_recall_curve(gt, rgc[l][0])
        prc = metrics.average_precision_score(gt, rgc[l][0])

        plt.title('Precision Recall Curve')
        plt.plot(recall, precision, 'b', label = 'AUC = %0.3f' % prc)
        plt.legend(loc = 'lower right')
        plt.xlim([-0.05, 1.05])
        plt.ylim([-0.05, 1.05])
        plt.ylabel('Precision')
        plt.xlabel('Recall')

        plt.savefig('eval/' + folder + '/prc_' + str(lag - l) + '.png', bbox_inches='tight')
        plt.clf()

    fgc = [0] * len(gt)

    for l in range(lag):
        fgc = lor(fgc, rgc[l][0])

    np.savetxt('eval/' + folder + '/gc_full.csv', np.array(fgc).reshape(100, 100), delimiter=",")

    fpr, tpr, threshold = metrics.roc_curve(gt, fgc)
    roc_auc = metrics.auc(fpr, tpr)

    plt.title('Receiver Operating Characteristic')
    plt.plot(fpr, tpr, 'b', label = 'AUC = %0.3f' % roc_auc)
    plt.legend(loc = 'lower right')
    plt.xlim([-0.05, 1.05])
    plt.ylim([-0.05, 1.05])
    plt.ylabel('True Positive Rate')
    plt.xlabel('False Positive Rate')

    plt.savefig('eval/' + folder + '/roc_full.png', bbox_inches='tight')
    plt.clf()

    precision, recall, threshold = metrics.precision_recall_curve(gt, fgc)
    prc = metrics.average_precision_score(gt, fgc)

    plt.title('Precision Recall Curve')
    plt.plot(recall, precision, 'b', label = 'AUC = %0.3f' % prc)
    plt.legend(loc = 'lower right')
    plt.xlim([-0.05, 1.05])
    plt.ylim([-0.05, 1.05])
    plt.ylabel('Precision')
    plt.xlabel('Recall')

    plt.savefig('eval/' + folder + '/prc_full.png', bbox_inches='tight')
    plt.clf()

trials = ['De-noised_100G_3T_300cPerT_dynamics_9_DS4',
          'De-noised_100G_4T_300cPerT_dynamics_10_DS5',
          'De-noised_100G_6T_300cPerT_dynamics_7_DS6',
          'De-noised_100G_7T_300cPerT_dynamics_11_DS7']


for trial in trials:
    for velo in [True]:
        for proba in [False, True]:
            for log in [False]:
                for gstd in [True]:
                    try:
                        evaluate(trial=trial, velo=velo, proba=proba, log=log, gstd=gstd)
                    except:
                        pass

'''
done = next(os.walk('img/'))[1]

for folder in done:
    if (not os.path.exists('eval/' + folder + '/roc_full.png')):
        try:
            evaluate(folder)
        except:
            pass
'''