-
Notifications
You must be signed in to change notification settings - Fork 19
/
Copy pathexperiments.py
208 lines (170 loc) · 7.57 KB
/
experiments.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
#!/bin/python
"""
This file contains several functions testing ELMs in different configurations,
optimize them and save the results in data files and pickles
"""
import sys
import os
import numpy as np
from scipy.signal import convolve2d
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap
from pyrcn.util import new_logger, argument_parser, get_mnist, tud_colors
train_size = 60000
def images_filter(images, kernel, stride=1):
filtered = np.zeros(images.shape)
for idx in range(images.shape[0]):
filtered[idx, ...] = convolve2d(images[idx, ...], kernel, mode='same')
return filtered
def picture_gradient(directory):
self_name = 'picture_gradient'
logger = new_logger(self_name, directory=directory)
X, y = get_mnist(directory)
logger.info('Loaded MNIST successfully with {0} records'
.format(X.shape[0]))
# scale X so X in [0, 1]
X /= 255.
# reshape X
X_images = X.reshape((X.shape[0], 28, 28))
list_kernels = [
{'name': 'laplace', 'kernel': np.array([[-1., -1., -1.],
[-1., 8, -1.],
[-1., -1., -1.]])},
{'name': 'mexicanhat', 'kernel': np.array([[0., 0., -1., 0., 0.],
[0., -1., -2., -1., 0.],
[-1., -2., 16, -2., -1.],
[0., -1., -2., -1., 0.],
[0., 0., -1., 0., 0.]])},
{'name': 'v_prewitt', 'kernel': np.array([[-1., -1., -1.],
[0., 0., 0.],
[1., 1., 1.]])},
{'name': 'h_prewitt', 'kernel': np.array([[-1., -1., -1.],
[0., 0., 0.],
[1., 1., 1.]]).T},
{'name': 'v_sobel', 'kernel': np.array([[-1., -2., -1.],
[0., 0., 0.],
[1., 2., 1.]])},
{'name': 'h_sobel', 'kernel': np.array([[-1., -2., -1.],
[0., 0., 0.],
[1., 2., 1.]]).T}]
example_image_idx = 5
fig, axs = plt.subplots(1, 4, figsize=(6, 2))
axs[0].imshow(X_images[example_image_idx], cmap=plt.cm.gray_r,
interpolation='none')
axs[0].set_title('no filter')
axs[1].imshow(convolve2d(X_images[example_image_idx],
list_kernels[0]['kernel'], mode='same'),
cmap=plt.cm.gray_r, interpolation='none')
axs[1].set_title('laplace')
axs[2].imshow(convolve2d(X_images[example_image_idx],
list_kernels[2]['kernel'], mode='same'),
cmap=plt.cm.gray_r, interpolation='none')
axs[2].set_title('vertical\nprewitt')
axs[3].imshow(convolve2d(X_images[example_image_idx],
list_kernels[5]['kernel'], mode='same'),
cmap=plt.cm.gray_r, interpolation='none')
axs[3].set_title('horizontal\nsobel')
for ax in axs:
ax.set_xticks([0, 27])
ax.set_xticklabels([0, 27])
ax.set_yticks([0, 27])
ax.set_yticklabels([0, 27])
fig.tight_layout()
fig.savefig(
os.path.join(directory, 'mnist-image-filters.pdf'), format='pdf')
fig.savefig(
os.path.join(os.environ['PGFPATH'], 'mnist-image-filters.pgf'),
format='pgf')
def plot_confusion(directory):
filepath = os.path.join(os.environ['DATAPATH'], '/coates20210310/est_'
'coates-minibatch-pca50+kmeans16000_matrix-'
'predicted.npz')
npzfile = np.load(filepath, allow_pickle=True)
X_test = np.array(npzfile['X_test'])
y_test = np.array(npzfile['y_test']).astype(int)
y_pred = np.array(npzfile['y_pred']).astype(int)
conf_matrix = np.zeros((10, 10))
X_example = X_test[(y_pred == 5) & (y_test == 6), ...]
img_example = X_example[3, ...]
imgpath = os.path.join(directory, 'confused6for5.png')
plt.imsave(
imgpath, img_example.reshape(28, 28), format='png', cmap=plt.cm.gray_r)
for pred_idx in range(conf_matrix.shape[0]):
for test_idx in range(conf_matrix.shape[1]):
conf_matrix[pred_idx, test_idx] = \
int(np.sum((y_pred == pred_idx) & (y_test == test_idx)))
tpr = np.zeros(10)
tnr = np.zeros(10)
for idx in range(10):
tpr[idx] = conf_matrix[idx, idx] / np.sum(conf_matrix[idx, :])
tnr[idx] = conf_matrix[idx, idx] / np.sum(conf_matrix[:, idx])
conf_matrix_norm = np.zeros((10, 10))
# norm row by row! => TPR
for idx in range(10):
conf_matrix_norm[idx, :] =\
conf_matrix[idx, :] / np.sum(conf_matrix[idx, :])
conf_matrix_norm[idx, idx] = 1 - conf_matrix_norm[idx, idx]
# colormap
n_colorsteps = 255 # in promille
color_array = np.zeros((n_colorsteps, 4))
lower_margin = 255
color_array[:lower_margin, :] += np.linspace(
start=tud_colors['lightgreen'], stop=tud_colors['red'],
num=lower_margin)
cm = ListedColormap(color_array)
fig = plt.figure(figsize=(4, 3))
ax = fig.add_axes([.15, .15, .75, .75])
img = ax.imshow(conf_matrix_norm * 100, interpolation='none', cmap=cm,
origin='lower', alpha=.7)
ax.set_xticks(np.arange(10))
ax.set_xticklabels(['{0:.0f}'
.format(pred_idx) for pred_idx in np.arange(10)])
ax.set_xlabel('true')
ax.set_yticks(np.arange(10))
ax.set_yticklabels(['{0:.0f}'
.format(pred_idx) for pred_idx in np.arange(10)])
ax.set_ylabel('predicted')
plt.colorbar(img, ax=ax, shrink=1., label='deviation from ideal TPR [%]')
for pred_idx in range(conf_matrix.shape[0]):
for test_idx in range(conf_matrix.shape[1]):
ax.text(x=pred_idx, y=test_idx, s='{0:.0f}'
.format(conf_matrix.T[pred_idx][test_idx]),
fontsize='xx-small', verticalalignment='center',
horizontalalignment='center')
# plt.show()
plt.savefig(os.path.join('./experiments/', 'confusion_matrix.pdf'),
format='pdf')
plt.savefig(os.path.join(os.environ['PGFPATH'], 'confusion_matrix.pgf'),
format='pgf')
def main(directory, params):
if not os.path.isdir(directory):
try:
os.mkdir(directory)
except PermissionError as e:
print('mkdir failed due to missing privileges: {0}'.format(e))
exit(1)
# subfolder for results
file_dir = os.path.join(directory, 'experiments')
if not os.path.isdir(file_dir):
os.mkdir(file_dir)
logger = new_logger('main', directory=file_dir)
logger.info('Started main with directory={0} and params={1}'
.format(directory, params))
# register parameters
experiment_names = {
'picture_gradient': picture_gradient,
'plot_confusion': plot_confusion,
}
# run specified programs
for param in params:
if param in experiment_names:
experiment_names[param](file_dir)
else:
logger.warning('Parameter {0} invalid/not found.'.format(param))
if __name__ == '__main__':
parsed_args = argument_parser.parse_args(sys.argv[1:])
if os.path.isdir(parsed_args.out):
main(parsed_args.out, parsed_args.params)
else:
main(parsed_args.params)
exit(0)