Lines Matching refs:total_count
41 sampling `self.total_count` draws from this Dirichlet-Multinomial distribution,
49 `tf.reduce_sum(value, -1) = self.total_count`. Its shape must be broadcastable
50 with `self.concentration` and `self.total_count`."""
58 length-`K` `concentration` vectors (`K > 1`) and a `total_count` number of
61 `tf.reduce_sum(counts, -1) = total_count`. The Dirichlet-Multinomial is
79 * `total_count = N`, `N` a positive integer,
95 `counts = [n_0,...,n_{K-1}] ~ Multinomial(total_count, probs)`
99 `concentration`, `total_count` and `counts` are broadcast to the same shape.
167 total_count,
175 total_count: Non-negative floating point tensor, whose dtype is the same
195 with ops.name_scope(name, values=[total_count, concentration]):
204 self._total_count = ops.convert_to_tensor(total_count, name="total_count")
225 def total_count(self):
253 n_draws = math_ops.cast(self.total_count, dtype=dtypes.int32)
278 self.total_count, counts)
285 return self.total_count * (self.concentration /
316 return x * (self.total_count * scale - x)
323 return math_ops.sqrt((1. + c0 / self.total_count) / (1. + c0))
344 self.total_count, math_ops.reduce_sum(counts, -1),
345 message="counts last-dimension must sum to `self.total_count`"),
