constant.py

import numpy

from chainer.backends import cuda
from chainer import initializer


class Identity(initializer.Initializer):

    """Initializes array with the identity matrix.

    It initializes the given array with the constant
    multiple of the identity matrix.
    Note that arrays to be passed must be 2D squared matrices.

    Attributes:
        ~Identity.scale (scalar): A constant to be multiplied to identity
        matrices.

    """

    def __init__(self, scale=1.0, dtype=None):
        self.scale = scale
        super(Identity, self).__init__(dtype)

    def __call__(self, array):
        if self.dtype is not None:
            assert array.dtype == self.dtype
        shape = array.shape
        if len(shape) != 2 or shape[0] != shape[1]:
            raise ValueError('Identity matrix initialization can only be used '
                             'for 2D squared matrices.')
        array[...] = 0
        xp = cuda.get_array_module(array)
        xp.fill_diagonal(array, self.scale)


class _Constant(initializer.Initializer):

    fill_value = None

    def __init__(self, dtype=None):
        if not (isinstance(self.fill_value, (numpy.ndarray, cuda.ndarray)) or
                numpy.isscalar(self.fill_value)):
            raise ValueError(
                'fill_value must be either scalar, numpy.ndarray or '
                'cupy.ndarray.')
        super(_Constant, self).__init__(dtype)

    def __call__(self, array):
        if self.dtype is not None:
            assert array.dtype == self.dtype
        xp = cuda.get_array_module(array)
        array[...] = xp.asarray(self.fill_value)


class Constant(_Constant):

    """Initializes array with constant value.

    Attributes:
        ~Constant.fill_value (scalar or numpy.ndarray or cupy.ndarray):
            A constant to be assigned to the initialized array.
            Broadcast is allowed on this assignment.
        ~Constant.dtype: Data type specifier.

    """

    def __init__(self, fill_value, dtype=None):
        self.fill_value = fill_value
        super(Constant, self).__init__(dtype)


class Zero(_Constant):
    """Initializes array to all-zero.

    Attributes:
        ~Zero.dtype: Data type specifier.
    """

    fill_value = 0.0


class One(_Constant):
    """Initializes array to all-one.

    Attributes:
        ~One.dtype: Data type specifier.
    """

    fill_value = 1.0


class NaN(_Constant):
    """Initializes array to all-NaN.

    Attributes:
        ~NaN.dtype: Data type specifier.
    """

    fill_value = numpy.nan