Merge pull request #115 from AdaWorldAPI/claude/continue-lance-graph-ndarray-Ld786

feat(hpc): publish wht_f32 + i2/i8 quantization + kmeans/squared_l2 for bgz-tensor

Author: AdaWorldAPI

COMPARISON.md

@@ -186,7 +186,7 @@
 | Crate | Upstream | **Fork** | Detail |
 |-------|----------|----------|--------|
 | `crates/p64` | Not present | **P64** | Palette64 data structure — convergence highway between ndarray and lance-graph |
-| `crates/phyllotactic-manifold` | Not present | **Phyllotactic Manifold** | Golden-angle spiral geometry for uniform point distribution |
+| `crates/fractal` | Not present | **Phyllotactic Manifold** | Golden-angle spiral geometry for uniform point distribution |
 
 ## Burn Backend (20 ops files)
 

Cargo.lock

@@ -50,7 +50,7 @@ libc = { version = "0.2.82", optional = true }
 matrixmultiply = { version = "0.3.2", default-features = false, features=["cgemm"] }
 blake3 = "1"
 p64 = { path = "crates/p64" }
-phyllotactic-manifold = { path = "crates/phyllotactic-manifold" }
+fractal = { path = "crates/fractal", default-features = false }
 
 serde = { version = "1.0", optional = true, default-features = false, features = ["alloc"] }
 rawpointer = { version = "0.2" }
@@ -81,7 +81,7 @@ blas = ["dep:cblas-sys", "dep:libc"]
 
 serde = ["dep:serde"]
 
-std = ["num-traits/std", "matrixmultiply/std"]
+std = ["num-traits/std", "matrixmultiply/std", "fractal/std"]
 rayon = ["dep:rayon", "std"]
 
 matrixmultiply-threading = ["matrixmultiply/threading"]

Cargo.toml

@@ -1,12 +1,17 @@
 [package]
-name = "phyllotactic-manifold"
+name = "fractal"
 version = "0.1.0"
 edition = "2021"
 rust-version = "1.94"
 description = "7+1 phyllotactic SIMD manifold with Euler-gamma singularity correction"
 license = "MIT OR Apache-2.0"
 
+[features]
+default = ["std"]
+std = []
+
 [dependencies]
+libm = { version = "0.2", default-features = false }
 
 [dev-dependencies]
 criterion = { version = "0.5", features = ["html_reports"] }

crates/phyllotactic-manifold/Cargo.toml crates/fractal/Cargo.tomlcrates/phyllotactic-manifold/Cargo.toml renamed to crates/fractal/Cargo.toml

@@ -1,14 +1,14 @@
 use criterion::{black_box, criterion_group, criterion_main, BenchmarkId, Criterion};
-use phyllotactic_manifold::*;
+use fractal::*;
 
 fn make_seeds(n: usize) -> Vec<[i8; 34]> {
     let mut seeds = Vec::with_capacity(n);
     for i in 0..n {
         let mut seed = [0i8; 34];
         seed[0] = (i % 128) as i8; // HEEL
         seed[33] = ((i * 7) % 128) as i8; // GAMMA
-        for j in 1..33 {
-            seed[j] = ((i * 13 + j * 37) % 256) as i8;
+        for (j, slot) in seed.iter_mut().enumerate().take(33).skip(1) {
+            *slot = ((i * 13 + j * 37) % 256) as i8;
         }
         seeds.push(seed);
     }
@@ -59,10 +59,10 @@ fn bench_resonance(c: &mut Criterion) {
     let threshold = 1000.0;
 
     // Pre-encode
-    let flat8_enc: Vec<_> = seeds.iter().map(|s| flat8::encode(s)).collect();
-    let spiral8_enc: Vec<_> = seeds.iter().map(|s| spiral8::encode(s)).collect();
-    let spiral8g_enc: Vec<_> = seeds.iter().map(|s| spiral8_gamma::encode(s)).collect();
-    let s7p1_enc: Vec<_> = seeds.iter().map(|s| seven_plus_one::encode(s)).collect();
+    let flat8_enc: Vec<_> = seeds.iter().map(flat8::encode).collect();
+    let spiral8_enc: Vec<_> = seeds.iter().map(spiral8::encode).collect();
+    let spiral8g_enc: Vec<_> = seeds.iter().map(spiral8_gamma::encode).collect();
+    let s7p1_enc: Vec<_> = seeds.iter().map(seven_plus_one::encode).collect();
 
     let mut group = c.benchmark_group("resonance");
 
@@ -85,11 +85,7 @@ fn bench_resonance(c: &mut Criterion) {
     group.bench_function("spiral8_gamma", |b| {
         b.iter(|| {
             for (x, y) in &spiral8g_enc {
-                black_box(spiral8_gamma::resonance(
-                    black_box(x),
-                    black_box(y),
-                    threshold,
-                ));
+                black_box(spiral8_gamma::resonance(black_box(x), black_box(y), threshold));
             }
         })
     });
@@ -107,7 +103,7 @@ fn bench_resonance(c: &mut Criterion) {
 
 fn bench_clam48(c: &mut Criterion) {
     let seeds = make_seeds(1024);
-    let manifolds: Vec<_> = seeds.iter().map(|s| seven_plus_one::encode(s)).collect();
+    let manifolds: Vec<_> = seeds.iter().map(seven_plus_one::encode).collect();
 
     c.bench_function("clam48_extraction", |b| {
         b.iter(|| {
@@ -123,8 +119,8 @@ fn bench_dead_zone(c: &mut Criterion) {
         let mut s = [0i8; 34];
         s[0] = 42;
         s[33] = 7;
-        for i in 1..33 {
-            s[i] = (i as i8).wrapping_mul(13).wrapping_add(37);
+        for (i, slot) in s.iter_mut().enumerate().take(33).skip(1) {
+            *slot = (i as i8).wrapping_mul(13).wrapping_add(37);
         }
         s
     };
@@ -151,12 +147,5 @@ fn bench_encode_scaling(c: &mut Criterion) {
     group.finish();
 }
 
-criterion_group!(
-    benches,
-    bench_encode,
-    bench_resonance,
-    bench_clam48,
-    bench_dead_zone,
-    bench_encode_scaling,
-);
+criterion_group!(benches, bench_encode, bench_resonance, bench_clam48, bench_dead_zone, bench_encode_scaling,);
 criterion_main!(benches);