def test_fprop(self):
conv_layer = blobs.Conv2D(W=self.conv_W,
b=self.conv_b,
strides=(1, 1),
border_mode=B.BorderMode.valid,
channels_come_last=False)
self.create_small_net_with_conv_layer(conv_layer,
outputs_per_channel=9)
func = B.function([self.input_layer.get_activation_vars()],
self.conv_layer.get_activation_vars())
np.testing.assert_almost_equal(
func(self.inp),
np.array([[
[[439, 467, 495], [551, 579, 607], [663, 691, 719]],
[[-439, -467, -495], [-551, -579, -607], [-663, -691, -719]],
],
[[
[1335, 1363, 1391],
[1447, 1475, 1503],
[1559, 1587, 1615],
],
[[-1335, -1363, -1391], [-1447, -1475, -1503],
[-1559, -1587, -1615]]]]))
def test_dense_backprop(self):
conv_layer = blobs.Conv2D(W=self.conv_W,
b=self.conv_b,
strides=(1, 1),
border_mode=B.BorderMode.valid,
channels_come_last=False,
conv_mxts_mode=ConvMxtsMode.Linear)
self.create_small_net_with_conv_layer(conv_layer,
outputs_per_channel=9)
self.dense_layer.update_task_index(task_index=0)
func = B.function([
self.input_layer.get_activation_vars(),
self.input_layer.get_reference_vars()
], self.input_layer.get_mxts())
np.testing.assert_almost_equal(
func(self.inp, np.zeros_like(self.inp)),
np.array([[[[0, 2, 2, 2], [4, 12, 12, 8], [4, 12, 12, 8],
[4, 10, 10, 6]],
[[8, 18, 18, 10], [20, 44, 44, 24], [20, 44, 44, 24],
[12, 26, 26, 14]]],
[[[0, 2, 2, 2], [4, 12, 12, 8], [4, 12, 12, 8],
[4, 10, 10, 6]],
[[8, 18, 18, 10], [20, 44, 44, 24], [20, 44, 44, 24],
[12, 26, 26, 14]]]]))
def conv2d_conversion(layer, name, mxts_mode):
#mxts_mode not used
converted_activation = activation_conversion(layer,
name,
mxts_mode=mxts_mode)
to_return = [
blobs.Conv2D(
name=("preact_" if len(converted_activation) > 0 else "") + name,
W=layer.get_weights()[0],
b=layer.get_weights()[1],
strides=layer.get_config()[KerasKeys.subsample],
border_mode=layer.get_config()[KerasKeys.border_mode])
]
to_return.extend(converted_activation)
return to_return
def test_relu_after_conv2d(self):
input_layer = blobs.Input(num_dims=None, shape=(None, 2, 2, 2))
conv_layer = blobs.Conv2D(W=np.random.random((2, 2, 2, 2)),
b=np.random.random((2, )),
conv_mxts_mode=ConvMxtsMode.Linear,
strides=(1, 1),
border_mode=PaddingMode.valid,
channels_come_last=True)
conv_layer.set_inputs(input_layer)
relu_after_conv = blobs.ReLU(
nonlinear_mxts_mode=NonlinearMxtsMode.DeepLIFT_GenomicsDefault)
relu_after_conv.set_inputs(conv_layer)
relu_after_conv.build_fwd_pass_vars()
self.assertEqual(relu_after_conv.nonlinear_mxts_mode,
NonlinearMxtsMode.Rescale)
def conv2d_conversion(layer, name, verbose, nonlinear_mxts_mode,
conv_mxts_mode, **kwargs):
print(layer.get_config())
#nonlinear_mxts_mode only used for activation
converted_activation = activation_conversion(
layer, name, verbose, nonlinear_mxts_mode=nonlinear_mxts_mode)
to_return = [
blobs.Conv2D(
name=("preact_" if len(converted_activation) > 0 else "") + name,
W=layer.get_weights()[0],
b=layer.get_weights()[1],
strides=layer.get_config()[KerasKeys.strides],
padding_mode=layer.get_config()["padding"].upper(),
conv_mxts_mode=conv_mxts_mode)
]
to_return.extend(converted_activation)
return to_return
def test_fprop_stride(self):
conv_layer = blobs.Conv2D(W=self.conv_W,
b=self.conv_b,
strides=(2, 2),
border_mode=B.BorderMode.valid,
channels_come_last=False,
conv_mxts_mode=ConvMxtsMode.Linear)
self.create_small_net_with_conv_layer(conv_layer,
outputs_per_channel=9)
func = B.function([self.input_layer.get_activation_vars()],
self.conv_layer.get_activation_vars())
np.testing.assert_almost_equal(
func(self.inp),
np.array([[[[439, 495], [663, 719]], [[-439, -495], [-663, -719]]],
[[[1335, 1391], [1559, 1615]],
[[-1335, -1391], [-1559, -1615]]]]))
def conv2d_conversion(layer, name, verbose, nonlinear_mxts_mode, **kwargs):
#nonlinear_mxts_mode only used for activation
converted_activation = activation_conversion(
layer, name, verbose, nonlinear_mxts_mode=nonlinear_mxts_mode)
W = layer.get_weights()[0]
channels_come_last = False
if KerasKeys.dim_ordering in layer.get_config():
dim_ordering = layer.get_config()[KerasKeys.dim_ordering]
if (dim_ordering == 'tf'):
W = W.transpose(3, 2, 0, 1)
channels_come_last = True
to_return = [
blobs.Conv2D(
name=("preact_" if len(converted_activation) > 0 else "") + name,
W=W,
b=layer.get_weights()[1],
strides=layer.get_config()[KerasKeys.subsample],
border_mode=layer.get_config()[KerasKeys.border_mode],
channels_come_last=channels_come_last)
]
to_return.extend(converted_activation)
return to_return
def test_relu_after_conv2d_batchnorm(self):
input_layer = blobs.Input(num_dims=None, shape=(None, 2, 2, 2))
conv_layer = blobs.Conv2D(W=np.random.random((2, 2, 2, 2)),
b=np.random.random((2, )),
conv_mxts_mode=ConvMxtsMode.Linear,
strides=(1, 1),
border_mode=PaddingMode.valid,
channels_come_last=True)
conv_layer.set_inputs(input_layer)
batch_norm = blobs.BatchNormalization(gamma=np.array([1.0, 1.0]),
beta=np.array([-0.5, 0.5]),
axis=-1,
mean=np.array([-0.5, 0.5]),
std=np.array([1.0, 1.0]),
epsilon=0.001)
batch_norm.set_inputs(conv_layer)
relu_after_bn = blobs.ReLU(
nonlinear_mxts_mode=NonlinearMxtsMode.DeepLIFT_GenomicsDefault)
relu_after_bn.set_inputs(batch_norm)
relu_after_bn.build_fwd_pass_vars()
self.assertEqual(relu_after_bn.nonlinear_mxts_mode,
NonlinearMxtsMode.Rescale)
