sum_axis_f64

void sum_axis_f64(const double * in, double * out, int outer, int reduce_dim, int inner)

Double-precision counterpart to sum_axis_f32.

Uses vDSP_sveD on the contiguous (inner == 1) path and the generic axis_reduce template otherwise.

Math

$\text{out}[o, i] = \sum_{r=0}^{\text{reduce\_dim}-1} \text{in}[o, r, i].$

inconst double*

Source buffer of shape (outer, reduce_dim, inner).

outdouble*

Destination buffer of shape (outer, inner).

outer, reduce_dim, innerstd::size_t

Layout extents; see sum_axis_f32.

Loading…

void sum_axis_f64(const double * in, double * out, int outer, int reduce_dim, int inner)