tensorflow scatter_ Nd function

The function prototype

tf.scatter_nd(
    indices, updates, shape, name=None
)

Function description

Will tensor updates The shape of extends to the new shape shape, Tensor in new shape shape The coordinates in are indices, Other element values are 0.

Parameters indices Index representing tensor , Parameters updates Represents the value scattered to the output tensor , Parameters shape Represents the shape of the output tensor .

Function USES

1、 One dimensional tensor

>>> indices = tf.constant([[1], [3], [5], [7]])
>>> updates = tf.constant([2, 3, 4, 5])
>>> shape = tf.constant([8])
>>> tf.scatter_nd(indices, updates, shape)
<tf.Tensor: shape=(8,), dtype=int32, numpy=array([0, 2, 0, 3, 0, 4, 0, 5])>

2、 Two dimensional tensor

>>> indices = tf.constant([[1], [2]])
>>> updates = tf.constant([[1, 2], [3, 4]])
>>> updates
<tf.Tensor: shape=(2, 2), dtype=int32, numpy=
array([[1, 2],
       [3, 4]])>
>>> shape = tf.constant([4, 2])
>>> tf.scatter_nd(indices, updates, shape)
<tf.Tensor: shape=(4, 2), dtype=int32, numpy=
array([[0, 0],
       [1, 2],
       [3, 4],
       [0, 0]])>

3、 Three dimensional tensor

>>> indices = tf.constant([[0], [2]])
>>> updates = tf.constant([[[5, 5, 5, 5], [6, 6, 6, 6],
                            [7, 7, 7, 7], [8, 8, 8, 8]],
                           [[5, 5, 5, 5], [6, 6, 6, 6],
                            [7, 7, 7, 7], [8, 8, 8, 8]]])
>>> updates
<tf.Tensor: shape=(2, 4, 4), dtype=int32, numpy=
array([[[5, 5, 5, 5],
        [6, 6, 6, 6],
        [7, 7, 7, 7],
        [8, 8, 8, 8]],

       [[5, 5, 5, 5],
        [6, 6, 6, 6],
        [7, 7, 7, 7],
        [8, 8, 8, 8]]])>
>>> shape = tf.constant([4, 4, 4])
>>> tf.scatter_nd(indices, updates, shape)
<tf.Tensor: shape=(4, 4, 4), dtype=int32, numpy=
array([[[5, 5, 5, 5],
        [6, 6, 6, 6],
        [7, 7, 7, 7],
        [8, 8, 8, 8]],

       [[0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0]],

       [[5, 5, 5, 5],
        [6, 6, 6, 6],
        [7, 7, 7, 7],
        [8, 8, 8, 8]],

       [[0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 0]]])>