SpatialVector

Published configurations

SpatialVector represents a small vector with runtime active dimension and bounded storage. The suite separates arithmetic, storage lifetime, scalar products, norms and cross products.

Scalar products

The right-conjugated product is \(\langle u,v\rangle=\sum_i u_i\overline{v_i}\). Bilinear contractions \(\sum_i u_iv_i\) are separate cases. Complex cross products use the ordinary bilinear convention.

Representations and result lifetime

Spatial values are compared with bounded Eigen values, dynamic Eigen vectors and selected fixed vectors. Owned-result cases include materialization and assignment. Destination cases evaluate into prepared, disjoint storage and are compared separately. Compound assignments include copying the left operand.

Measurement contract

Inputs, numerical references and array allocation are excluded. Scalar-loop oracles check results before and after timing. Each sweep processes one or thirty-two inputs. Calling-thread CPU time is reported per sweep; divide by the operations counter for time per value. Empty shapes are retained where defined; normalization requires nonempty vectors.

Benchmarks

Timing history

RodinMathSpatialVectorBenchmarks · cpu_time · slow · ranks: 1