Lines Matching refs:num_components
110 def make_univariate_mixture(batch_shape, num_components, use_static_graph):
113 array_ops.concat((batch_shape, [num_components]), axis=0),
119 for _ in range(num_components)
125 def make_multivariate_mixture(batch_shape, num_components, event_shape,
131 array_ops.concat((batch_shape_tensor, [num_components]), 0),
134 tensor_shape.TensorShape(batch_shape).concatenate(num_components))
146 components = [create_component() for _ in range(num_components)]
157 dist = make_univariate_mixture(batch_shape, num_components=10,
166 batch_shape, num_components=10, event_shape=event_shape,
252 batch_shape=batch_shape, num_components=2,
274 batch_shape=batch_shape, num_components=2, event_shape=(4,),
297 num_components = 2
302 batch_shape=batch_shape, num_components=num_components,
326 np.reshape(cat_probs_values, [-1, num_components]),
327 np.reshape(stacked_mean_res, [-1, num_components]),
328 np.reshape(stacked_dev_res, [-1, num_components]))
337 num_components = 2
344 num_components=num_components,
369 np.reshape(cat_probs_values, [-1, num_components]),
370 np.reshape(stacked_mean_res, [-1, num_components]),
371 np.reshape(stacked_dev_res, [-1, num_components]))
401 dist = make_univariate_mixture(batch_shape=[], num_components=2,
428 batch_shape=[], num_components=2, event_shape=[3],
456 dist = make_univariate_mixture(batch_shape=[2, 3], num_components=2,
484 batch_shape=[2, 3], num_components=2, event_shape=[4],
510 num_components = 3
513 batch_shape=batch_shape, num_components=num_components,
525 for c in range(num_components):
571 num_components = 3
573 batch_shape=[], num_components=num_components, event_shape=[2],
584 for c in range(num_components):
596 num_components = 3
598 batch_shape=[2, 3], num_components=num_components,
609 for c in range(num_components):
624 num_components = 3
634 num_components=num_components, event_shape=[4],
655 for c in range(num_components):
678 batch_shape=batch_shape, num_components=2, event_shape=(4,),
693 # for i in num_components, batchwise.
823 num_components, batch_size, num_features,
832 num_components=num_components,
842 (name, use_gpu, num_components, batch_size, num_features,
845 use_gpu, num_components, batch_size, num_features, sample_size,
852 def create_distribution(batch_size, num_components, num_features):
854 logits=np.random.randn(batch_size, num_components))
857 for _ in range(num_components)
861 for _ in range(num_components)
871 for num_components in 1, 8, 16:
879 num_components=num_components,
892 def create_distribution(batch_size, num_components, num_features):
894 logits=np.random.randn(batch_size, num_components))
897 for _ in range(num_components)
902 for _ in range(num_components)
913 for num_components in 1, 8, 16:
921 num_components=num_components,
