tensorplay.functional.bitwise_right_shift

tensorplay.functional.bitwise_right_shift(input, other, *, out=None) → Tensor[source]

Computes the right arithmetic shift of input by other bits. The input tensor must be of integral type. This operator supports broadcasting to a common shape and type promotion. In any case, if the value of the right operand is negative or is greater or equal to the number of bits in the promoted left operand, the behavior is undefined.

The operation applied is:

$$\text{out}_i = \text{input}_i >> \text{other}_i$$

Parameters:

• input (Tensor or Scalar) – the first input tensor

• other (Tensor or Scalar) – the second input tensor

Keyword Arguments:

out (Tensor, optional) – the output tensor.

Example:

>>> tensorplay.bitwise_right_shift(tensorplay.tensor([-2, -7, 31], dtype=tensorplay.int8), tensorplay.tensor([1, 0, 3], dtype=tensorplay.int8))
tensor([-1, -7,  3], dtype=tensorplay.int8)