def params_size(event_shape=(),
dispersion='full',
name="NegativeBinomialDispLayer_params_size"):
r"""The number of `params` needed to create a single distribution."""
if dispersion == 'full':
return 2 * _event_size(event_shape, name=name)
return _event_size(event_shape, name=name)
def params_size(event_shape=(), name=None):
"""The number of `params` needed to create a single distribution."""
with tf.compat.v1.name_scope(name, 'Gamma_params_size',
[event_shape]):
event_shape = tf.convert_to_tensor(
value=event_shape, name='event_shape', dtype=tf.int32)
return 2 * _event_size(event_shape, name=name or 'Gamma_params_size')
def params_size(event_shape=(), name=None):
"""The number of `params` needed to create a single distribution."""
with tf.compat.v1.name_scope(name, 'Gamma_params_size', [event_shape]):
event_shape = tf.convert_to_tensor(value=event_shape,
name='event_shape',
dtype=tf.int32)
return 2 * _event_size(event_shape,
name=name or 'Gamma_params_size')
def params_size(event_shape=(), tied_inflation_rate=False,
name=None):
"""The number of `params` needed to create a single distribution."""
with tf.compat.v1.name_scope(name,
'ZeroInflatedNegativeBinomial_params_size',
[event_shape]):
event_shape = tf.convert_to_tensor(
value=event_shape, name='event_shape', dtype=tf.int32)
return 3 * _event_size(event_shape,
name=name or 'ZeroInflatedNegativeBinomial_params_size')
def params_size(event_shape=(), tied_inflation_rate=False, name=None):
"""The number of `params` needed to create a single distribution."""
with tf.compat.v1.name_scope(
name, 'ZeroInflatedNegativeBinomial_params_size',
[event_shape]):
event_shape = tf.convert_to_tensor(value=event_shape,
name='event_shape',
dtype=tf.int32)
return 3 * _event_size(
event_shape,
name=name or 'ZeroInflatedNegativeBinomial_params_size')
def params_size(event_shape=(),
dispersion='full',
inflation='full',
name="ZINegativeBinomialDisp_params_size"):
r"""The number of `params` needed to create a single distribution."""
size = _event_size(event_shape, name=name)
total = 3 * size
if dispersion != 'full':
total -= size
if inflation != 'full':
total -= size
return total
def params_size(event_shape=(), name='BinomialLayer_params_size'):
r"""The number of `params` needed to create a single distribution."""
return 2 * _event_size(event_shape, name=name)
def params_size(event_shape=(), name="ZeroInflatedPoisson_params_size"):
r"""The number of `params` needed to create a single distribution."""
return 2 * _event_size(event_shape, name=name)
def params_size(event_shape=(), name="LogNormal_params_size"): r"""The number of `params` needed to create a single distribution.""" return 2 * _event_size(event_shape, name=name)
def params_size(event_shape, name='VectorDeterministicLayer_params_size'):
r""" The number of `params` needed to create a single distribution. """
return _event_size(event_shape, name)
def params_size(event_shape=(), name='RelaxedBernoulliLayer_params_size'): r"""The number of `params` needed to create a single distribution.""" return _event_size(event_shape, name=name)
def params_size(event_shape, name='RelaxedSoftmaxLayer_params_size'): """The number of `params` needed to create a single distribution.""" return _event_size(event_shape, name=name)
def params_size(event_shape, name='OneHotCategoricalLayer_params_size'): """The number of `params` needed to create a single distribution.""" return _event_size(event_shape, name=name)
def params_size(event_shape=(), name='ZeroInflatedBernoulli_params_size'): r"""The number of `params` needed to create a single distribution.""" return 2 * _event_size(event_shape, name=name)
def params_size(event_shape=(), name='DirichletMultinomial_params_size'): r"""The number of `params` needed to create a single distribution.""" return _event_size(event_shape, name=name) + 1.
def params_size(event_shape=(), name="ZINegativeBinomialDisp_params_size"): """The number of `params` needed to create a single distribution.""" return 3 * _event_size(event_shape, name=name)
