In [1]:
import numpy as np
from keras.models import Model
from keras.layers import Input
from keras.layers.pooling import AveragePooling3D
from keras import backend as K
import json
from collections import OrderedDict
In [2]:
def format_decimal(arr, places=6):
return [round(x * 10**places) / 10**places for x in arr]
In [3]:
DATA = OrderedDict()
[pooling.AveragePooling3D.0] input 4x4x4x2, pool_size=(2, 2, 2), strides=None, padding='valid', data_format='channels_last'
In [4]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=None, padding='valid', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(290)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.0'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.1] input 4x4x4x2, pool_size=(2, 2, 2), strides=(1, 1, 1), padding='valid', data_format='channels_last'
In [5]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=(1, 1, 1), padding='valid', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(291)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.1'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.2] input 4x5x2x3, pool_size=(2, 2, 2), strides=(2, 1, 1), padding='valid', data_format='channels_last'
In [6]:
data_in_shape = (4, 5, 2, 3)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=(2, 1, 1), padding='valid', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(282)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.2'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.3] input 4x4x4x2, pool_size=(3, 3, 3), strides=None, padding='valid', data_format='channels_last'
In [7]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=None, padding='valid', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(283)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.3'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.4] input 4x4x4x2, pool_size=(3, 3, 3), strides=(3, 3, 3), padding='valid', data_format='channels_last'
In [8]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=(3, 3, 3), padding='valid', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(284)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.4'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.5] input 4x4x4x2, pool_size=(2, 2, 2), strides=None, padding='same', data_format='channels_last'
In [9]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=None, padding='same', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(285)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.5'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.6] input 4x4x4x2, pool_size=(2, 2, 2), strides=(1, 1, 1), padding='same', data_format='channels_last'
In [10]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=(1, 1, 1), padding='same', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(286)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.6'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.7] input 4x5x4x2, pool_size=(2, 2, 2), strides=(1, 2, 1), padding='same', data_format='channels_last'
In [11]:
data_in_shape = (4, 5, 4, 2)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=(1, 2, 1), padding='same', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(287)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.7'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.8] input 4x4x4x2, pool_size=(3, 3, 3), strides=None, padding='same', data_format='channels_last'
In [12]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=None, padding='same', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(288)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.8'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.9] input 4x4x4x2, pool_size=(3, 3, 3), strides=(3, 3, 3), padding='same', data_format='channels_last'
In [13]:
data_in_shape = (4, 4, 4, 2)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=(3, 3, 3), padding='same', data_format='channels_last')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(289)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.9'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.10] input 2x3x3x4, pool_size=(3, 3, 3), strides=(2, 2, 2), padding='valid', data_format='channels_first'
In [14]:
data_in_shape = (2, 3, 3, 4)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=(2, 2, 2), padding='valid', data_format='channels_first')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(290)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.10'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.11] input 2x3x3x4, pool_size=(3, 3, 3), strides=(1, 1, 1), padding='same', data_format='channels_first'
In [15]:
data_in_shape = (2, 3, 3, 4)
L = AveragePooling3D(pool_size=(3, 3, 3), strides=(1, 1, 1), padding='same', data_format='channels_first')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(291)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.11'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
[pooling.AveragePooling3D.12] input 3x4x4x3, pool_size=(2, 2, 2), strides=None, padding='valid', data_format='channels_first'
In [16]:
data_in_shape = (3, 4, 4, 3)
L = AveragePooling3D(pool_size=(2, 2, 2), strides=None, padding='valid', data_format='channels_first')
layer_0 = Input(shape=data_in_shape)
layer_1 = L(layer_0)
model = Model(inputs=layer_0, outputs=layer_1)
# set weights to random (use seed for reproducibility)
np.random.seed(292)
data_in = 2 * np.random.random(data_in_shape) - 1
result = model.predict(np.array([data_in]))
data_out_shape = result[0].shape
data_in_formatted = format_decimal(data_in.ravel().tolist())
data_out_formatted = format_decimal(result[0].ravel().tolist())
print('')
print('in shape:', data_in_shape)
print('in:', data_in_formatted)
print('out shape:', data_out_shape)
print('out:', data_out_formatted)
DATA['pooling.AveragePooling3D.12'] = {
'input': {'data': data_in_formatted, 'shape': data_in_shape},
'expected': {'data': data_out_formatted, 'shape': data_out_shape}
}
In [17]:
import os
filename = '../../../test/data/layers/pooling/AveragePooling3D.json'
if not os.path.exists(os.path.dirname(filename)):
os.makedirs(os.path.dirname(filename))
with open(filename, 'w') as f:
json.dump(DATA, f)
In [18]:
print(json.dumps(DATA))
In [ ]: