experiments/plot_results.py at master · iolfcv/experiments · 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
import matplotlib
matplotlib.use('agg')
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.markers import CARETDOWN
import argparse
import os

parser = argparse.ArgumentParser(description='Incremental learning')
parser.add_argument('-f', '--file', type=str,
                    help='CSV file name containing results')

savefig = True
TITLE_SIZE = 28
AXIS_SIZE = 20
matplotlib.rcParams['font.family'] = ['serif']

matplotlib.rc('axes', titlesize=TITLE_SIZE)
matplotlib.rc('axes', labelsize=AXIS_SIZE)
dotted_line_width = 2.5
title = 'CRIB-Toys Incremental Learning Performance'
anchor = (0.65, 0.43)


def plot_acc(datafile):
    fig, ax1 = plt.subplots()
    if not os.path.exists(datafile):
        raise Exception('Results file path not found')

    # Load results from files
    classes_file = '%s-classes.npz' % os.path.splitext(datafile)[0]
    matr_file = '%s-matr.npz' % os.path.splitext(datafile)[0]
    info_classes = np.load(classes_file)
    info_matr = np.load(matr_file)

    classes = info_classes['classes_seen']
    num_classes = info_matr['args'][()].total_classes
    num_le = len(classes) # Number of learning exposures

    os.makedirs('results', exist_ok=True)
    filename = 'results/plot.pdf'
    linestyle = '-'
    all_lines = []

    ax1.set_xlabel('Number of learning exposures')
    ax1.set_ylabel('% Test accuracy over seen objects')
    ax1.set_yticks(np.arange(0, 105, 10))
    ax1.set_ylim([0, 105])

    ax2 = ax1.twinx()
    ax2.set_ylim([0,1.05*num_classes])
    ax2.set_ylabel('Unique objects seen (UOS)')
    ax2.set_yticks(np.arange(0, num_classes+1, num_classes//10))

    plt.xticks(np.arange(0, num_le+1, num_le//10))
    ax1.set_yticks(np.arange(0, 101, 10))
    ax1.grid()

    acc_matr = info_matr['acc_matr']
    counter = []
    cnt = 0
    for i in range(len(classes)):
        if classes[i] in classes[:i]:
            counter.append(cnt)
        else:
            cnt += 1
            counter.append(cnt)

    test_acc = np.sum(acc_matr, axis=0) / np.array(counter)
    print(test_acc)

    all_lines.append(ax2.plot(np.arange(0, num_le), counter, color='purple',
                              linestyle=':', linewidth=dotted_line_width,
                              label='Ground-truth UOS')[0])
    all_lines.append(ax1.plot(np.arange(0, num_le), test_acc, color='blue',
                              label='Learner Performance', linestyle='-')[0])


    all_labels = [l.get_label() for l in all_lines]
    plt.title(title)
    ax1.legend(all_lines, all_labels, bbox_to_anchor=anchor,
               loc=2, borderaxespad=0., fancybox=True, framealpha=0.7,
               fontsize=18, numpoints=1)

    plt.gcf().set_size_inches(16, 8)
    plt.savefig(filename, dpi=300, bbox_inches='tight',
                pad_inches=0.01 ,transparent=True)

if __name__ == '__main__':
    args = parser.parse_args()
    plot_acc(args.file)