cupynumeric.vecdot#

cupynumeric.vecdot(
x1: ndarray,
x2: ndarray,
/,
*,
axis: int = -1,
) ndarray#

Compute the vector dot product along an axis.

If x1 is complex, its complex conjugate is used for the product. The dimensions outside the contracted axis are broadcast together.

Parameters:
  • x1 (array_like) – First input array. If x1 is complex, the complex conjugate is taken before calculating the dot product.

  • x2 (array_like) – Second input array.

  • axis (int, optional) – Axis over which to compute the vector dot product. Only negative axes are currently supported.

Returns:

output – The vector dot product of x1 and x2.

Return type:

ndarray

Notes

The contracted axis length must match exactly before broadcasting the other dimensions.

See also

numpy.vecdot

Availability:

Multiple GPUs, Multiple CPUs