Repository navigation
Expand file tree
/
Copy pathhugo.toml
More file actions
112 lines (103 loc) · 7.08 KB
/
Copy pathhugo.toml
File metadata and controls
112 lines (103 loc) · 7.08 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
baseURL = 'https://go-ndarray.github.io/'
locale = 'en-us'
title = 'go-ndarray'
# No theme -- the layout is the go-* family template go-fft.github.io uses
# (itself go-authn.github.io's): one self-contained layouts/index.html,
# partials/styles.html (inline :root tokens, a verbatim copy of go-fft's) and
# partials/theme-toggle.html (system / light / dark).
disableKinds = ['taxonomy', 'term', 'rss', 'sitemap', '404']
[markup.goldmark.renderer]
unsafe = true
# The design -- layouts/partials/styles.html, theme-toggle.html,
# fork-ribbon.html and icons/{github,book}.html -- comes from the shared Hugo
# module github.com/go-fleettools/landingkit, not from copies here: the copies
# drifted. Its version is pinned in go.mod (Renovate bumps it), and kit.go
# keeps `go mod tidy` from deleting that requirement. A file of the same name
# under layouts/partials/ would silently override the module's.
[module]
[[module.imports]]
path = "github.com/go-fleettools/landingkit"
[params]
description = "go-ndarray is a pure-Go (CGO=0) NumPy-style n-dimensional array library in 13 dtypes — slicing/views, broadcasting, ufuncs, reductions, numpy.linalg, numpy.fft and numpy.random (NumPy's numbers for the same seed), multicore + SIMD on amd64, arm64, ppc64le, loong64, riscv64 and s390x, with a Workspace arena for loops."
# The hero's buttons: rendered by the kit's cta.html, in the kit's order.
[[params.cta]]
kind = "docs"
href = "https://go-ndarray.github.io/docs/"
label = "Read the documentation →"
[[params.cta]]
kind = "try"
href = "/playground/"
label = "Try it in your browser"
[[params.repos]]
name = "ndarray"
kind = "library · v0.46.0"
href = "https://github.com/go-ndarray/ndarray"
blurb = "The library, in <strong>13 dtypes</strong> with NumPy 2 promotion. Creation (<code>Zeros</code>/<code>Arange</code>/<code>Linspace</code>/…), strided <strong>views</strong>, broadcasting elementwise + ufuncs, reductions (<code>Sum</code> in <strong>NumPy's pairwise order, within a stated error bound checked against <code>math/big</code></strong>, from <a href=\"https://github.com/go-simd/floats\">go-simd/floats</a>; <code>Max</code>/<code>ArgMax</code>/<code>CumSum</code>/<code>Clip</code>/<code>Where</code>), manipulation (<code>Reshape</code>/<code>Concatenate</code>/<code>Stack</code>), linalg (<code>MatMul</code>/<code>Dot</code>) on <a href=\"https://github.com/go-blas/blas\">go-blas/blas</a>, a panel-packed cache-blocked GEMM with a go-asmgen SIMD-FMA micro-kernel, the same kernels as before v0.42.0 (NEON 4×8 on arm64, AVX2/FMA 6×8 on amd64, VSX 8×8 on ppc64le, LASX 8×8 on loong64, RVV 6×8 on riscv64, 4×8 on s390x), and a <code>Workspace</code> arena for loops (<code>Use</code>/<code>Reset</code>/<code>Detach</code>). Packages <code>linalg</code>, <code>fft</code> and <code>random</code> are <code>numpy.linalg</code>, <code>numpy.fft</code> and <code>numpy.random</code>."
[[params.repos]]
name = "docs"
kind = "site"
href = "https://go-ndarray.github.io/docs/"
blurb = "Versioned documentation site (Hugo + Hextra): API reference, the roadmap, and honest benchmark pages versus NumPy 2.5.3 / OpenBLAS 0.3.34. <a href=\"https://github.com/go-ndarray/docs\">Source →</a>"
[[params.repos]]
name = "brand"
kind = "assets"
href = "https://github.com/go-ndarray/brand"
blurb = "Logos and icons for the organization, in SVG / PNG / JPG / ICO / ICNS across colour, white and black variants."
[params.brand]
# Generated by logos/landing-kit/kit.py from static/img/logo.svg: the colours are
# READ from the logo's gradient, never chosen. Do not add other keys:
# neutrals (backgrounds, surfaces, greys) belong to the shared sheet.
# Contrast: base on white 5.48:1, bright on the dark background 10.05:1.
base = "#047857"
deep = "#024935"
bright = "#34d399"
shadow = "rgba(4,120,87,.7)"
glow = "rgba(4,120,87,.6)"
wash = "rgba(4,120,87,.10)"
washDeep = "rgba(2,73,53,.06)"
chipBg = "#e7faf3"
chipBorder = "#c2f2e0"
chipFg = "#024935"
chipBgDark = "rgba(52,211,153,.14)"
chipBorderDark = "rgba(52,211,153,.32)"
chipFgDark = "#8fe7c7"
statsEnd = "#cef4e7"
bandText = "#dff8ef"
# The API cards (#api) and the count in the hero come from this list, never
# from a number typed beside it.
[[params.families]]
name = "Dtypes"
what = "bool, int8…int64, uint8…uint64, float32/float64, complex64/complex128, chosen at run time with NumPy 2 promotion: `AsType`, `Data[T]`, `Item[T]`, weak Go scalars with `Scalar`."
[[params.families]]
name = "Creation"
what = "`Zeros`, `Ones`, `Full`, `Arange`, `Linspace`, `FromData`, `Eye` and the rest of NumPy's constructors, in any dtype (`ZerosOf`, `FromSlice`)."
[[params.families]]
name = "Views and indexing"
what = "`Reshape`, `Transpose` and NumPy basic indexing with `Slice(All(), A(0), …)`: strided views that share data, no copy. `Ravel` is a view too when the elements are one row-major run, as `numpy.ravel`; `Flatten` and `Copy` give an independent array."
[[params.families]]
name = "Broadcasting and ufuncs"
what = "Elementwise arithmetic with NumPy broadcasting, and the unary functions, including `Exp` and `Log` ported from Arm's optimized-routines (≤ 0.51 ULP); `Round` is half to even, as `np.round` (`RoundHalfAway` keeps half away from zero), and float64 `Abs`, `Neg`, `Square`, `Floor`, `Ceil` and `Round` run on go-simd/floats vector kernels."
[[params.families]]
name = "Reductions"
what = "`Sum`, `Mean`, `Max`, `ArgMax`, `CumSum`, `Clip`, `Where`, whole-array or along an axis, multicore on large arrays."
[[params.families]]
name = "Manipulation"
what = "`Concatenate`, `Stack` and the shape operations NumPy users reach for."
[[params.families]]
name = "Linear algebra"
what = "`MatMul` and `Dot` run on [go-blas/blas](https://github.com/go-blas/blas), the panel-packed, cache-blocked GEMM with a go-asmgen SIMD micro-kernel per architecture (the same kernels ndarray carried before v0.42.0); package `linalg` is `numpy.linalg` (`Solve`, `Inv`, `Det`, `QR`, `SVD`, `Eig`, `Lstsq`, …), checked against numpy 2.5.3 on 1,040 cases."
[[params.families]]
name = "Fourier transforms"
what = "Package `fft` is `numpy.fft` on arrays: `FFT`, `RFFT`, `FFT2`, `FFTN` and the rest, with `n`/`s`, `axes` and `norm`, over go-fft."
[[params.families]]
name = "Random numbers"
what = "Package `random` is `numpy.random` with NumPy's numbers for the same seed: PCG64, MT19937, Philox, SFC64, 40 samplers, the legacy `RandomState`, and allocation-free scalar draws (`Float64`, `FillUniform`, `IntN`, `FillIntN`, `ShuffleInt64`) for solvers, NumPy's stream consumed call for call."
[[params.families]]
name = "Notebooks and storage"
what = "`Repr` prints like NumPy, `Display` gives a notebook text and HTML, and `MarshalBinary` writes the same bytes on every architecture."
[[params.families]]
name = "Workspace"
what = "An arena for loops: `Use`, `Reset`, `Detach` — results stop costing the garbage collector from one pass to the next."
# The architectures the tests RUN on, read from the library's CI.
[params.mantra]
goarch = ["amd64", "arm64", "riscv64", "loong64", "ppc64le", "s390x", "arm"]