tf.sparse.reduce_sum | TensorFlow v2.16.1 (original) (raw)

Computes tf.sparse.add of elements across dimensions of a SparseTensor.

tf.sparse.reduce_sum(
    sp_input, axis=None, keepdims=None, output_is_sparse=False, name=None
)

Used in the notebooks

Used in the tutorials
Client-efficient large-model federated learning via `federated_select` and sparse aggregation

This is the reduction operation for the elementwise tf.sparse.add op.

This Op takes a SparseTensor and is the sparse counterpart totf.reduce_sum(). In particular, this Op also returns a dense Tensorif output_is_sparse is False, or a SparseTensor if output_is_sparseis True.

Reduces sp_input along the dimensions given in axis. Unless keepdims is true, the rank of the tensor is reduced by 1 for each entry in axis. Ifkeepdims is true, the reduced dimensions are retained with length 1.

If axis has no entries, all dimensions are reduced, and a tensor with a single element is returned. Additionally, the axes can be negative, similar to the indexing rules in Python.

For example
'x' represents [[1, ?, 1] where ? is implicitly-zero. x = tf.sparse.SparseTensor([[0, 0], [0, 2], [1, 1]], [1, 1, 1], [2, 3]) tf.sparse.reduce_sum(x) <tf.Tensor: shape=(), dtype=int32, numpy=3> tf.sparse.reduce_sum(x, 0) <tf.Tensor: shape=(3,), dtype=int32, numpy=array([1, 1, 1], dtype=int32)> tf.sparse.reduce_sum(x, 1) # Can also use -1 as the axis <tf.Tensor: shape=(2,), dtype=int32, numpy=array([2, 1], dtype=int32)> tf.sparse.reduce_sum(x, 1, keepdims=True) <tf.Tensor: shape=(2, 1), dtype=int32, numpy= array([[2], [1]], dtype=int32)> tf.sparse.reduce_sum(x, [0, 1]) <tf.Tensor: shape=(), dtype=int32, numpy=3>

For example

'x' represents [[1, ?, 1] where ? is implicitly-zero. x = tf.sparse.SparseTensor([[0, 0], [0, 2], [1, 1]], [1, 1, 1], [2, 3]) tf.sparse.reduce_sum(x) <tf.Tensor: shape=(), dtype=int32, numpy=3> tf.sparse.reduce_sum(x, 0) <tf.Tensor: shape=(3,), dtype=int32, numpy=array([1, 1, 1], dtype=int32)> tf.sparse.reduce_sum(x, 1) # Can also use -1 as the axis <tf.Tensor: shape=(2,), dtype=int32, numpy=array([2, 1], dtype=int32)> tf.sparse.reduce_sum(x, 1, keepdims=True) <tf.Tensor: shape=(2, 1), dtype=int32, numpy= array([[2], [1]], dtype=int32)> tf.sparse.reduce_sum(x, [0, 1]) <tf.Tensor: shape=(), dtype=int32, numpy=3>

Args
sp_input	The SparseTensor to reduce. Should have numeric type.
axis	The dimensions to reduce; list or scalar. If None (the default), reduces all dimensions.
keepdims	If true, retain reduced dimensions with length 1.
output_is_sparse	If true, returns a SparseTensor instead of a denseTensor (the default).
name	A name for the operation (optional).

Returns
The reduced Tensor or the reduced SparseTensor if output_is_sparse is True.