[tannevaled]

This PR contains the following updates:

Package	Change	Age	Confidence
github.com/go-ndarray/ndarray	v0.34.0 → v0.37.5
github.com/go-simd/floats	v0.7.0 → v0.12.0
github.com/go-sparse/sparse	v0.3.0 → v0.3.2

---

Release Notes

go-ndarray/ndarray (github.com/go-ndarray/ndarray)

v0.37.5

Compare Source

What's Changed

• perf: refbench against NumPy on one core; replace the multicore page by @​tannevaled in #​105

Full Changelog: go-ndarray/ndarray@v0.37.4...v0.37.5

v0.37.4

Compare Source

What's Changed

• docs: float64 Exp on 16 cores is slowed by allocation, not the kernel by @​tannevaled in #​102

Full Changelog: go-ndarray/ndarray@v0.37.3...v0.37.4

v0.37.3

Compare Source

What's Changed

• floats v0.11.0: SIMD Max/Min restored on amd64 and arm64 by @​tannevaled in #​101

Full Changelog: go-ndarray/ndarray@v0.37.2...v0.37.3

v0.37.2

Compare Source

What's Changed

• floats v0.10.0: drop the max/min zero fallback and the moved sum/elementwise assembly by @​tannevaled in #​100

Full Changelog: go-ndarray/ndarray@v0.37.1...v0.37.2

v0.37.1

Compare Source

SumP runs on several cores again, with the bits of floats.SumPairwise (NumPy order) whatever the core count: it cuts the NumPy pairwise tree where NumPy cuts it and adds the subtrees back up the same tree (#​99). Results are unchanged from v0.37.0.

v0.37.0

Compare Source

K1 of the go-scikit performance campaign (go-scikit/roadmap#27): ndarray imports go-simd/floats v0.9.0.

Changed result: Sum (and mean, var, std built on it) is now NumPy's pairwise sum (floats.SumPairwise), bit-identical to np.sum of a contiguous 1-D float64 array on every architecture. It replaces lane-parallel sums whose last bits differed from NumPy's and between architectures; it is serial at every size.

Exp, Log and the correctly rounded float32 Exp/Log, and on amd64/arm64 the float64 add/sub/mul/div/sqrt/max/min/clip kernels, are go-simd/floats' (the AVX2 and NEON exp/log kernels written here moved down: #​94, #​95, #​96).

v0.36.0

Compare Source

What's Changed

• AVX2 kernel for float64 Log: 1.55x faster than NumPy by @​tannevaled in #​96

Full Changelog: go-ndarray/ndarray@v0.35.0...v0.36.0

v0.35.0

Compare Source

What's Changed

• fix(deps): update dependencies by @​tannevaled in #​89
• docs: design of the dense linear-algebra foundation (scipy.linalg side) by @​tannevaled in #​92
• Float32 results from the Workspace; cheaper range check in the AVX2 log by @​tannevaled in #​94

Full Changelog: go-ndarray/ndarray@v0.34.0...v0.35.0

go-simd/floats (github.com/go-simd/floats)

v0.12.0

Compare Source

What's Changed

• Polyval, Chebval: Horner and Clenshaw over an array of points (K2) by @​tannevaled in #​39

Full Changelog: go-simd/floats@v0.11.0...v0.12.0

v0.11.0

Compare Source

What's Changed

• Max, Min: SSE2 and NEON kernels (SIMD regression since ndarray v0.37.0) by @​tannevaled in #​38

Full Changelog: go-simd/floats@v0.10.0...v0.11.0

v0.10.0

Compare Source

What's Changed

• ci: on-demand workflow for the 2^32 correct-rounding proofs by @​tannevaled in #​33
• Max, Min: order -0 < +0 like NumPy by @​tannevaled in #​34
• amd64: SSE2 elementwise kernels for CPUs without AVX2+FMA by @​tannevaled in #​35
• ppc64le, loong64, riscv64, s390x: pairwise-sum block and elementwise kernels by @​tannevaled in #​36
• README: changes in v0.10.0 by @​tannevaled in #​37

Full Changelog: go-simd/floats@v0.9.0...v0.10.0

v0.9.0

Compare Source

What's Changed

• Exp, Log, Float32Exp, Float32Log: the exp/log kernels of go-ndarray (K1) by @​tannevaled in #​29

Full Changelog: go-simd/floats@v0.8.0...v0.9.0

v0.8.0

Compare Source

What's Changed

• bench: replace numpycmp with a refbench module (bench/perf) by @​tannevaled in #​27
• Add, Sub, Div, Sqrt, ClampScalar: the elementwise kernels of go-ndarray (K1) by @​tannevaled in #​28

Full Changelog: go-simd/floats@v0.7.0...v0.8.0

go-sparse/sparse (github.com/go-sparse/sparse)

v0.3.2

Compare Source

What's Changed

• docs: record the first refbench run (invalid: load above the gate) by @​tannevaled in #​17
• perf: NormFro, COO conversions, Laplacian assembly, Prune (causes from scipy's code) by @​tannevaled in #​18

Full Changelog: v0.3.1...v0.3.2

v0.3.1

Compare Source

What's Changed

• docs: coverage.md against scipy.sparse, MATLAB and SuiteSparse by @​tannevaled in #​14
• docs: sparse direct-solver design (design-direct.md) by @​tannevaled in #​15
• fix(deps): update dependencies by @​tannevaled in #​5
• perf: refbench benchmarks (QDLDL, scipy) and allocations that grew with the input by @​tannevaled in #​16

Full Changelog: v0.3.0...v0.3.1

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • At any time (no schedule defined)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Enabled.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

👻 Immortal: This PR will be recreated if closed unmerged. Get config help if that's undesired.

---

• If you want to rebase/retry this PR, check this box

---

This PR has been generated by Mend Renovate CLI.