heuropt/examples/compare.rs

//! Multi-seed algorithm comparison harness.
//!
//! Runs every applicable optimizer on each test problem across N seeds and
//! prints aggregate quality metrics. Adding a new algorithm to the
//! comparison is a single-line edit to the runner table — see the bottom
//! of this file.
//!
//! ```bash
//! cargo run --release --example compare
//! ```

use std::f64::consts::PI;
use std::time::Instant;

use heuropt::metrics::{hypervolume::hypervolume_2d, spacing::spacing};
use heuropt::prelude::*;

const SEEDS: u64 = 10;

const ZDT1_DIM: usize = 30;
const ZDT1_BUDGET: usize = 25_000;
// Standard ZDT1 reference point. Using [11, 11] (rather than the
// near-front [1.1, 1.1]) so under-converged algorithms with large `g`
// values still register a meaningful — if poor — hypervolume.
const ZDT1_REFERENCE: [f64; 2] = [11.0, 11.0];

const RASTRIGIN_DIM: usize = 5;
const RASTRIGIN_BUDGET: usize = 50_000;

// -----------------------------------------------------------------------------
// Test problems
// -----------------------------------------------------------------------------

struct Zdt1 {
    dim: usize,
}

impl Problem for Zdt1 {
    type Decision = Vec<f64>;

    fn objectives(&self) -> ObjectiveSpace {
        ObjectiveSpace::new(vec![Objective::minimize("f1"), Objective::minimize("f2")])
    }

    fn evaluate(&self, x: &Vec<f64>) -> Evaluation {
        let f1 = x[0];
        let tail_sum: f64 = x[1..].iter().sum();
        let g = 1.0 + 9.0 * tail_sum / (self.dim as f64 - 1.0);
        let f2 = g * (1.0 - (f1 / g).sqrt());
        Evaluation::new(vec![f1, f2])
    }
}

struct Rastrigin {
    dim: usize,
}

impl Problem for Rastrigin {
    type Decision = Vec<f64>;

    fn objectives(&self) -> ObjectiveSpace {
        ObjectiveSpace::new(vec![Objective::minimize("f")])
    }

    fn evaluate(&self, x: &Vec<f64>) -> Evaluation {
        let n = self.dim as f64;
        let value = 10.0 * n
            + x.iter().map(|v| v * v - 10.0 * (2.0 * PI * v).cos()).sum::<f64>();
        Evaluation::new(vec![value])
    }
}

// -----------------------------------------------------------------------------
// Run results + metrics aggregation
// -----------------------------------------------------------------------------

#[derive(Clone)]
struct MoRun {
    front: Vec<Candidate<Vec<f64>>>,
    wall_ms: u128,
}

#[derive(Clone)]
struct SoRun {
    best_value: f64,
    wall_ms: u128,
}

fn mean_l2_to_zdt1_front(front: &[Candidate<Vec<f64>>]) -> f64 {
    if front.is_empty() {
        return f64::INFINITY;
    }
    let samples: Vec<(f64, f64)> = (0..=1000)
        .map(|i| {
            let f1 = i as f64 / 1000.0;
            (f1, 1.0 - f1.sqrt())
        })
        .collect();
    let mut total = 0.0;
    for c in front {
        let f1 = c.evaluation.objectives[0];
        let f2 = c.evaluation.objectives[1];
        let mut best = f64::INFINITY;
        for &(rf1, rf2) in &samples {
            let d = ((rf1 - f1).powi(2) + (rf2 - f2).powi(2)).sqrt();
            if d < best {
                best = d;
            }
        }
        total += best;
    }
    total / front.len() as f64
}

fn mean_std(values: &[f64]) -> (f64, f64) {
    let n = values.len() as f64;
    let mean = values.iter().sum::<f64>() / n;
    let var = values.iter().map(|v| (v - mean).powi(2)).sum::<f64>() / n;
    (mean, var.sqrt())
}

// -----------------------------------------------------------------------------
// ZDT1 algorithm runners
// -----------------------------------------------------------------------------

fn zdt1_random(seed: u64) -> MoRun {
    let problem = Zdt1 { dim: ZDT1_DIM };
    let initializer = RealBounds::new(vec![(0.0, 1.0); ZDT1_DIM]);
    let config = RandomSearchConfig {
        iterations: ZDT1_BUDGET,
        batch_size: 1,
        seed,
    };
    let mut opt = RandomSearch::new(config, initializer);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    MoRun { front: result.pareto_front, wall_ms: t0.elapsed().as_millis() }
}

fn zdt1_paes(seed: u64) -> MoRun {
    let problem = Zdt1 { dim: ZDT1_DIM };
    let initializer = RealBounds::new(vec![(0.0, 1.0); ZDT1_DIM]);
    let variation = BoundedGaussianMutation::new(0.05, vec![(0.0, 1.0); ZDT1_DIM]);
    let config = PaesConfig {
        iterations: ZDT1_BUDGET,
        archive_size: 100,
        seed,
    };
    let mut opt = Paes::new(config, initializer, variation);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    MoRun { front: result.pareto_front, wall_ms: t0.elapsed().as_millis() }
}

fn zdt1_spea2(seed: u64) -> MoRun {
    let problem = Zdt1 { dim: ZDT1_DIM };
    let bounds = vec![(0.0, 1.0); ZDT1_DIM];
    let initializer = RealBounds::new(bounds.clone());
    let variation = CompositeVariation {
        crossover: SimulatedBinaryCrossover::new(bounds.clone(), 15.0, 0.5),
        mutation: PolynomialMutation::new(bounds, 20.0, 1.0 / ZDT1_DIM as f64),
    };
    let pop = 100;
    let arc = 100;
    // SPEA2 evaluates `pop_size` per generation after the initial population.
    let gens = (ZDT1_BUDGET - pop) / pop;
    let config = Spea2Config {
        population_size: pop,
        archive_size: arc,
        generations: gens,
        seed,
    };
    let mut opt = Spea2::new(config, initializer, variation);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    MoRun { front: result.pareto_front, wall_ms: t0.elapsed().as_millis() }
}

fn zdt1_nsga2(seed: u64) -> MoRun {
    let problem = Zdt1 { dim: ZDT1_DIM };
    let bounds = vec![(0.0, 1.0); ZDT1_DIM];
    let initializer = RealBounds::new(bounds.clone());
    let variation = CompositeVariation {
        crossover: SimulatedBinaryCrossover::new(bounds.clone(), 15.0, 0.5),
        mutation: PolynomialMutation::new(bounds, 20.0, 1.0 / ZDT1_DIM as f64),
    };
    let pop = 100;
    let gens = ZDT1_BUDGET / pop;
    let config = Nsga2Config { population_size: pop, generations: gens, seed };
    let mut opt = Nsga2::new(config, initializer, variation);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    MoRun { front: result.pareto_front, wall_ms: t0.elapsed().as_millis() }
}

// -----------------------------------------------------------------------------
// Rastrigin algorithm runners
// -----------------------------------------------------------------------------

fn rastrigin_random(seed: u64) -> SoRun {
    let problem = Rastrigin { dim: RASTRIGIN_DIM };
    let initializer = RealBounds::new(vec![(-5.12, 5.12); RASTRIGIN_DIM]);
    let config = RandomSearchConfig {
        iterations: RASTRIGIN_BUDGET,
        batch_size: 1,
        seed,
    };
    let mut opt = RandomSearch::new(config, initializer);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    SoRun {
        best_value: result.best.unwrap().evaluation.objectives[0],
        wall_ms: t0.elapsed().as_millis(),
    }
}

fn rastrigin_paes(seed: u64) -> SoRun {
    let problem = Rastrigin { dim: RASTRIGIN_DIM };
    let initializer = RealBounds::new(vec![(-5.12, 5.12); RASTRIGIN_DIM]);
    let variation = BoundedGaussianMutation::new(0.3, vec![(-5.12, 5.12); RASTRIGIN_DIM]);
    let config = PaesConfig {
        iterations: RASTRIGIN_BUDGET,
        archive_size: 32,
        seed,
    };
    let mut opt = Paes::new(config, initializer, variation);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    SoRun {
        best_value: result.best.unwrap().evaluation.objectives[0],
        wall_ms: t0.elapsed().as_millis(),
    }
}

fn rastrigin_nsga2(seed: u64) -> SoRun {
    let problem = Rastrigin { dim: RASTRIGIN_DIM };
    let bounds = vec![(-5.12, 5.12); RASTRIGIN_DIM];
    let initializer = RealBounds::new(bounds.clone());
    let variation = CompositeVariation {
        crossover: SimulatedBinaryCrossover::new(bounds.clone(), 15.0, 0.5),
        mutation: PolynomialMutation::new(bounds, 20.0, 1.0 / RASTRIGIN_DIM as f64),
    };
    let pop = 50;
    let gens = RASTRIGIN_BUDGET / pop;
    let config = Nsga2Config { population_size: pop, generations: gens, seed };
    let mut opt = Nsga2::new(config, initializer, variation);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    SoRun {
        best_value: result.best.unwrap().evaluation.objectives[0],
        wall_ms: t0.elapsed().as_millis(),
    }
}

fn rastrigin_de(seed: u64) -> SoRun {
    let problem = Rastrigin { dim: RASTRIGIN_DIM };
    let bounds = RealBounds::new(vec![(-5.12, 5.12); RASTRIGIN_DIM]);
    let pop = 50;
    let gens = (RASTRIGIN_BUDGET - pop) / pop; // initial pop also evaluates
    let config = DifferentialEvolutionConfig {
        population_size: pop,
        generations: gens,
        differential_weight: 0.5,
        crossover_probability: 0.9,
        seed,
    };
    let mut opt = DifferentialEvolution::new(config, bounds);
    let t0 = Instant::now();
    let result = opt.run(&problem);
    SoRun {
        best_value: result.best.unwrap().evaluation.objectives[0],
        wall_ms: t0.elapsed().as_millis(),
    }
}

// -----------------------------------------------------------------------------
// Main
// -----------------------------------------------------------------------------

fn run_zdt1_comparison() {
    println!(
        "== ZDT1 (dim={ZDT1_DIM}, {ZDT1_BUDGET} evals/run × {SEEDS} seeds) =="
    );
    println!("metric arrows: hypervolume↑ (higher better), others↓ (lower better)");
    println!();
    println!(
        "{:<14} {:>16} {:>14} {:>14} {:>10} {:>10}",
        "algorithm", "hypervolume", "spacing", "mean L2", "front", "ms",
    );
    println!("{}", "-".repeat(82));

    let zdt1 = Zdt1 { dim: ZDT1_DIM };
    let zdt1_objs = zdt1.objectives();

    type Runner = fn(u64) -> MoRun;
    let runners: &[(&str, Runner)] = &[
        ("RandomSearch", zdt1_random),
        ("PAES", zdt1_paes),
        ("SPEA2", zdt1_spea2),
        ("NSGA-II", zdt1_nsga2),
    ];

    for (name, runner) in runners {
        let runs: Vec<MoRun> = (0..SEEDS).map(runner).collect();
        let hv: Vec<f64> = runs
            .iter()
            .map(|r| hypervolume_2d(&r.front, &zdt1_objs, ZDT1_REFERENCE))
            .collect();
        let sp: Vec<f64> =
            runs.iter().map(|r| spacing(&r.front, &zdt1_objs)).collect();
        let l2: Vec<f64> =
            runs.iter().map(|r| mean_l2_to_zdt1_front(&r.front)).collect();
        let fs: Vec<f64> = runs.iter().map(|r| r.front.len() as f64).collect();
        let ms: Vec<f64> = runs.iter().map(|r| r.wall_ms as f64).collect();

        let (hv_m, hv_s) = mean_std(&hv);
        let (sp_m, sp_s) = mean_std(&sp);
        let (l2_m, l2_s) = mean_std(&l2);
        let (fs_m, _) = mean_std(&fs);
        let (ms_m, _) = mean_std(&ms);

        println!(
            "{:<14} {:>16} {:>14} {:>14} {:>10} {:>10}",
            name,
            format!("{hv_m:.4}±{hv_s:.4}"),
            format!("{sp_m:.4}±{sp_s:.4}"),
            format!("{l2_m:.4}±{l2_s:.4}"),
            format!("{fs_m:.0}"),
            format!("{ms_m:.0}"),
        );
    }
}

fn run_rastrigin_comparison() {
    println!();
    println!(
        "== Rastrigin (dim={RASTRIGIN_DIM}, {RASTRIGIN_BUDGET} evals/run × {SEEDS} seeds) =="
    );
    println!("global minimum: f = 0  (lower is better)");
    println!();
    println!("{:<14} {:>20} {:>10}", "algorithm", "best f", "ms");
    println!("{}", "-".repeat(48));

    type Runner = fn(u64) -> SoRun;
    let runners: &[(&str, Runner)] = &[
        ("RandomSearch", rastrigin_random),
        ("PAES", rastrigin_paes),
        ("NSGA-II", rastrigin_nsga2),
        ("DE", rastrigin_de),
    ];

    for (name, runner) in runners {
        let runs: Vec<SoRun> = (0..SEEDS).map(runner).collect();
        let best: Vec<f64> = runs.iter().map(|r| r.best_value).collect();
        let ms: Vec<f64> = runs.iter().map(|r| r.wall_ms as f64).collect();

        let (b_m, b_s) = mean_std(&best);
        let (ms_m, _) = mean_std(&ms);

        println!(
            "{:<14} {:>20} {:>10}",
            name,
            format!("{b_m:.4e} ± {b_s:.2e}"),
            format!("{ms_m:.0}"),
        );
    }
}

fn main() {
    run_zdt1_comparison();
    run_rastrigin_comparison();
}