flopscope.

flopscope.numpy.linalg.vecdot

fnp.linalg.vecdot(x1: 'ArrayLike', x2: 'ArrayLike', /, *, axis: 'int' = -1) -> 'FlopscopeArray'[flopscope source][numpy source]

Computes the vector dot product.

Adapted from NumPy docs np.linalg.vecdot

Arealinalg
Typecustom
NumPy Refnp.linalg.vecdot
Cost
delegates to flops.vecdot
Flopscope Context

Delegates to `fnp.vecdot` which charges `2*n` FLOPs.

This function is restricted to arguments compatible with the Array API, contrary to flops.vecdot.

Let a\mathbf{a} be a vector in x1 and b\mathbf{b} be a corresponding vector in x2. The dot product is defined as:

a⋅b=∑i=0n−1ai‾bi\mathbf{a} \cdot \mathbf{b} = \sum_{i=0}^{n-1} \overline{a_i}b_i

over the dimension specified by axis and where ai‾\overline{a_i} denotes the complex conjugate if aia_i is complex and the identity otherwise.

Parameters

x1:array_like

First input array.

x2:array_like

Second input array.

axis:int, optional

Axis over which to compute the dot product. Default: -1.

Returns

output:ndarray

The vector dot product of the input.

See also

Examples

Get the projected size along a given normal for an array of vectors.

>>> v = flops.array([[0., 5., 0.], [0., 0., 10.], [0., 6., 8.]])
>>> n = flops.array([0., 0.6, 0.8])
>>> flops.linalg.vecdot(v, n)
array([ 3.,  8., 10.])