# EpyNN/epynn/dropout/models.py
# Local application/library specific imports
from epynn.commons.models import Layer
from epynn.dropout.forward import dropout_forward
from epynn.dropout.backward import dropout_backward
from epynn.dropout.parameters import (
dropout_compute_shapes,
dropout_initialize_parameters,
dropout_compute_gradients,
dropout_update_parameters
)
[docs]class Dropout(Layer):
"""
Definition of a dropout layer prototype.
:param drop_prob: Probability to drop one data point from previous layer to next layer, defaults to 0.5.
:type drop_prob: float, optional
:param axis: Compute and apply dropout mask along defined axis, defaults to all axis.
:type axis: int or tuple[int], optional
"""
def __init__(self, drop_prob=0.5, axis=()):
"""Initialize instance variable attributes.
"""
super().__init__()
axis = axis if isinstance(axis, tuple) else (axis,)
self.d['d'] = drop_prob
self.d['a'] = axis
self.trainable = False
return None
[docs] def compute_shapes(self, A):
"""Wrapper for :func:`epynn.dropout.parameters.dropout_compute_shapes()`.
:param A: Output of forward propagation from previous layer.
:type A: :class:`numpy.ndarray`
"""
dropout_compute_shapes(self, A)
return None
[docs] def initialize_parameters(self):
"""Wrapper for :func:`epynn.dropout.parameters.dropout_initialize_parameters()`.
"""
dropout_initialize_parameters(self)
return None
[docs] def forward(self, A):
"""Wrapper for :func:`epynn.dropout.forward.dropout_forward()`.
:param A: Output of forward propagation from previous layer.
:type A: :class:`numpy.ndarray`
:return: Output of forward propagation for current layer.
:rtype: :class:`numpy.ndarray`
"""
self.compute_shapes(A)
A = self.fc['A'] = dropout_forward(self, A)
self.update_shapes(self.fc, self.fs)
return A
[docs] def backward(self, dX):
"""Wrapper for :func:`epynn.dropout.backward.dropout_backward()`.
:param dX: Output of backward propagation from next layer.
:type dX: :class:`numpy.ndarray`
:return: Output of backward propagation for current layer.
:rtype: :class:`numpy.ndarray`
"""
dX = dropout_backward(self, dX)
self.update_shapes(self.bc, self.bs)
return dX
[docs] def compute_gradients(self):
"""Wrapper for :func:`epynn.dropout.parameters.dropout_compute_gradients()`. Dummy method, there are no gradients to compute in layer.
"""
dropout_compute_gradients(self)
return None
[docs] def update_parameters(self):
"""Wrapper for :func:`epynn.dropout.parameters.dropout_update_parameters()`. Dummy method, there are no parameters to update in layer.
"""
if self.trainable:
dropout_update_parameters(self)
return None