def transform_by_kumaraswamy(x, feature_ndims, example_ndims):
"""Apply a Kumaraswamy bijector to features."""
concentration1 = util.pad_shape_with_ones(
self.concentration1,
example_ndims,
start=-(feature_ndims + 1))
concentration0 = util.pad_shape_with_ones(
self.concentration0,
example_ndims,
start=-(feature_ndims + 1))
bij = kumaraswamy_cdf.KumaraswamyCDF(
concentration1, concentration0, validate_args=validate_args)
return bij.forward(x)
def __init__(self,
concentration1=1.,
concentration0=1.,
validate_args=False,
allow_nan_stats=True,
name='Kumaraswamy'):
"""Initialize a batch of Kumaraswamy distributions.
Args:
concentration1: Positive floating-point `Tensor` indicating mean
number of successes; aka 'alpha'. Implies `self.dtype` and
`self.batch_shape`, i.e.,
`concentration1.shape = [N1, N2, ..., Nm] = self.batch_shape`.
concentration0: Positive floating-point `Tensor` indicating mean
number of failures; aka 'beta'. Otherwise has same semantics as
`concentration1`.
validate_args: Python `bool`, default `False`. When `True` distribution
parameters are checked for validity despite possibly degrading runtime
performance. When `False` invalid inputs may silently render incorrect
outputs.
allow_nan_stats: Python `bool`, default `True`. When `True`, statistics
(e.g., mean, mode, variance) use the value '`NaN`' to indicate the
result is undefined. When `False`, an exception is raised if one or
more of the statistic's batch members are undefined.
name: Python `str` name prefixed to Ops created by this class.
"""
parameters = dict(locals())
with tf.name_scope(name) as name:
dtype = dtype_util.common_dtype([concentration1, concentration0],
dtype_hint=tf.float32)
concentration1 = tensor_util.convert_nonref_to_tensor(
concentration1, name='concentration1', dtype=dtype)
concentration0 = tensor_util.convert_nonref_to_tensor(
concentration0, name='concentration0', dtype=dtype)
self._kumaraswamy_cdf = kumaraswamy_cdf.KumaraswamyCDF(
concentration1=concentration1,
concentration0=concentration0,
validate_args=validate_args)
batch_shape = distribution_util.get_broadcast_shape(
concentration1, concentration0)
super(Kumaraswamy, self).__init__(
# TODO(b/137665504): Use batch-adding meta-distribution to set the
# batch shape instead of tf.zeros.
distribution=uniform.Uniform(
low=tf.zeros(batch_shape, dtype=dtype),
high=tf.ones([], dtype=dtype),
allow_nan_stats=allow_nan_stats),
bijector=invert.Invert(
self._kumaraswamy_cdf, validate_args=validate_args),
parameters=parameters,
name=name)