Merge branch 'development' into add-contribution-guides

Author: Mec-iS

.github/workflows/ci.yml

@@ -2,23 +2,24 @@ name: CI
 
 on:
   push:
-    branches: [ main, development ]
+    branches: [main, development]
   pull_request:
-    branches: [ development ]
+    branches: [development]
 
 jobs:
   tests:
     runs-on: "${{ matrix.platform.os }}-latest"
     strategy:
       matrix:
-        platform: [
-          { os: "windows", target: "x86_64-pc-windows-msvc" },
-          { os: "windows", target: "i686-pc-windows-msvc" },
-          { os: "ubuntu", target: "x86_64-unknown-linux-gnu" },
-          { os: "ubuntu", target: "i686-unknown-linux-gnu" },
-          { os: "ubuntu", target: "wasm32-unknown-unknown" },
-          { os: "macos", target: "aarch64-apple-darwin" },
-        ]    
+        platform:
+          [
+            { os: "windows", target: "x86_64-pc-windows-msvc" },
+            { os: "windows", target: "i686-pc-windows-msvc" },
+            { os: "ubuntu", target: "x86_64-unknown-linux-gnu" },
+            { os: "ubuntu", target: "i686-unknown-linux-gnu" },
+            { os: "ubuntu", target: "wasm32-unknown-unknown" },
+            { os: "macos", target: "aarch64-apple-darwin" },
+          ]
     env:
       TZ: "/usr/share/zoneinfo/your/location"
     steps:
@@ -40,7 +41,7 @@ jobs:
           default: true
       - name: Install test runner for wasm
         if: matrix.platform.target == 'wasm32-unknown-unknown'
-        run:  curl https://rustwasm.github.io/wasm-pack/installer/init.sh -sSf | sh
+        run: curl https://rustwasm.github.io/wasm-pack/installer/init.sh -sSf | sh
       - name: Stable Build
         uses: actions-rs/cargo@v1
         with:

CHANGELOG.md

@@ -6,6 +6,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## Added
+- Seeds to multiple algorithims that depend on random number generation.
+- Added feature `js` to use WASM in browser
+
+## BREAKING CHANGE
+- Added a new parameter to `train_test_split` to define the seed.
+
+## [0.2.1] - 2022-05-10
+
 ## Added
 - L2 regularization penalty to the Logistic Regression
 - Getters for the naive bayes structs

Cargo.toml

@@ -16,22 +16,29 @@ categories = ["science"]
 default = ["datasets"]
 ndarray-bindings = ["ndarray"]
 nalgebra-bindings = ["nalgebra"]
-datasets = []
+datasets = ["rand_distr", "std"]
+fp_bench = ["itertools"]
+std = ["rand/std", "rand/std_rng"]
+# wasm32 only
+js = ["getrandom/js"]
 
 [dependencies]
 ndarray = { version = "0.15", optional = true }
-nalgebra = { version = "0.23.0", optional = true }
-num-traits = "0.2.12"
-num = "0.4.0"
-rand = "0.8.3"
-rand_distr = "0.4.0"
-serde = { version = "1.0.115", features = ["derive"], optional = true }
+nalgebra = { version = "0.31", optional = true }
+num-traits = "0.2"
+num = "0.4"
+rand = { version = "0.8", default-features = false, features = ["small_rng"] }
+rand_distr = { version = "0.4", optional = true }
+serde = { version = "1", features = ["derive"], optional = true }
+itertools = { version = "0.10.3", optional = true }
+cfg-if = "1.0.0"
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
-getrandom = { version = "0.2", features = ["js"] }
+getrandom = { version = "0.2", optional = true }
 
 [dev-dependencies]
-criterion = "0.3"
+smartcore = { path = ".", features = ["fp_bench"] }
+criterion = { version = "0.4", default-features = false }
 serde_json = "1.0"
 bincode = "1.3.1"
 
@@ -46,3 +53,8 @@ harness = false
 name = "naive_bayes"
 harness = false
 required-features = ["ndarray-bindings", "nalgebra-bindings"]
+
+[[bench]]
+name = "fastpair"
+harness = false
+required-features = ["fp_bench"]

LICENSE

@@ -186,7 +186,7 @@
       same "printed page" as the copyright notice for easier
       identification within third-party archives.
 
-   Copyright [yyyy] [name of copyright owner]
+   Copyright 2019-present at SmartCore developers (smartcorelib.org) 
 
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.

benches/fastpair.rs

@@ -0,0 +1,56 @@
+use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion};
+
+// to run this bench you have to change the declaraion in mod.rs ---> pub mod fastpair;
+use smartcore::algorithm::neighbour::fastpair::FastPair;
+use smartcore::linalg::naive::dense_matrix::*;
+use std::time::Duration;
+
+fn closest_pair_bench(n: usize, m: usize) -> () {
+    let x = DenseMatrix::<f64>::rand(n, m);
+    let fastpair = FastPair::new(&x);
+    let result = fastpair.unwrap();
+
+    result.closest_pair();
+}
+
+fn closest_pair_brute_bench(n: usize, m: usize) -> () {
+    let x = DenseMatrix::<f64>::rand(n, m);
+    let fastpair = FastPair::new(&x);
+    let result = fastpair.unwrap();
+
+    result.closest_pair_brute();
+}
+
+fn bench_fastpair(c: &mut Criterion) {
+    let mut group = c.benchmark_group("FastPair");
+
+    // with full samples size (100) the test will take too long
+    group.significance_level(0.1).sample_size(30);
+    // increase from default 5.0 secs
+    group.measurement_time(Duration::from_secs(60));
+
+    for n_samples in [100_usize, 1000_usize].iter() {
+        for n_features in [10_usize, 100_usize, 1000_usize].iter() {
+            group.bench_with_input(
+                BenchmarkId::from_parameter(format!(
+                    "fastpair --- n_samples: {}, n_features: {}",
+                    n_samples, n_features
+                )),
+                n_samples,
+                |b, _| b.iter(|| closest_pair_bench(*n_samples, *n_features)),
+            );
+            group.bench_with_input(
+                BenchmarkId::from_parameter(format!(
+                    "brute --- n_samples: {}, n_features: {}",
+                    n_samples, n_features
+                )),
+                n_samples,
+                |b, _| b.iter(|| closest_pair_brute_bench(*n_samples, *n_features)),
+            );
+        }
+    }
+    group.finish();
+}
+
+criterion_group!(benches, bench_fastpair);
+criterion_main!(benches);

src/algorithm/neighbour/bbd_tree.rs

@@ -59,7 +59,7 @@ impl<T: RealNumber> BBDTree<T> {
         tree
     }
 
-    pub(in crate) fn clustering(
+    pub(crate) fn clustering(
         &self,
         centroids: &[Vec<T>],
         sums: &mut Vec<Vec<T>>,

src/algorithm/neighbour/distances.rs

@@ -0,0 +1,48 @@
+//!
+//! Dissimilarities for vector-vector distance
+//!
+//! Representing distances as pairwise dissimilarities, so to build a
+//! graph of closest neighbours. This representation can be reused for
+//! different implementations (initially used in this library for FastPair).
+use std::cmp::{Eq, Ordering, PartialOrd};
+
+#[cfg(feature = "serde")]
+use serde::{Deserialize, Serialize};
+
+use crate::math::num::RealNumber;
+
+///
+/// The edge of the subgraph is defined by `PairwiseDistance`.
+/// The calling algorithm can store a list of distsances as
+/// a list of these structures.
+///
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone, Copy)]
+pub struct PairwiseDistance<T: RealNumber> {
+    /// index of the vector in the original `Matrix` or list
+    pub node: usize,
+
+    /// index of the closest neighbor in the original `Matrix` or same list
+    pub neighbour: Option<usize>,
+
+    /// measure of distance, according to the algorithm distance function
+    /// if the distance is None, the edge has value "infinite" or max distance
+    /// each algorithm has to match
+    pub distance: Option<T>,
+}
+
+impl<T: RealNumber> Eq for PairwiseDistance<T> {}
+
+impl<T: RealNumber> PartialEq for PairwiseDistance<T> {
+    fn eq(&self, other: &Self) -> bool {
+        self.node == other.node
+            && self.neighbour == other.neighbour
+            && self.distance == other.distance
+    }
+}
+
+impl<T: RealNumber> PartialOrd for PairwiseDistance<T> {
+    fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
+        self.distance.partial_cmp(&other.distance)
+    }
+}