scirs2-spatial 0.4.0

//! Performance validation example for scirs2-spatial
//!
//! This example runs the core performance benchmarks to validate the
//! performance claims made in the spatial module.

use scirs2_core::ndarray::Array2;
use scirs2_core::random::{rngs::StdRng, Rng, RngExt, SeedableRng};
use scirs2_spatial::{
    distance::{euclidean, pdist},
    simd_distance::{
        parallel_pdist, simd_euclidean_distance, simd_euclidean_distance_batch, simd_knn_search,
    },
    KDTree,
};
use std::time::Instant;

/// Generate random points for testing
#[allow(dead_code)]
fn generate_points(n_points: usize, dimensions: usize, seed: u64) -> Array2<f64> {
    let mut rng = StdRng::seed_from_u64(seed);
    Array2::from_shape_fn((n_points, dimensions), |_| rng.random_range(-10.0..10.0))
}

/// Test SIMD vs scalar distance calculations
#[allow(dead_code)]
fn test_simd_vs_scalar() {
    println!("=== SIMD vs Scalar Distance Performance ===");
    println!(
        "{:>8} {:>15} {:>15} {:>12}",
        "Dim", "Scalar (ns)", "SIMD (ns)", "Speedup"
    );
    println!("{}", "-".repeat(55));

    for &dim in &[4, 8, 16, 32, 64] {
        let p1: Vec<f64> = (0..dim).map(|i| i as f64).collect();
        let p2: Vec<f64> = (0..dim).map(|i| (i + 1) as f64).collect();

        // Scalar timing
        let start = Instant::now();
        for _ in 0..1000 {
            let _ = euclidean(&p1, &p2);
        }
        let scalar_time = start.elapsed();

        // SIMD timing
        let start = Instant::now();
        for _ in 0..1000 {
            let _ = simd_euclidean_distance(&p1, &p2).expect("Operation failed");
        }
        let simd_time = start.elapsed();

        let speedup = scalar_time.as_nanos() as f64 / simd_time.as_nanos() as f64;

        println!(
            "{:>8} {:>15} {:>15} {:>12.2}x",
            dim,
            scalar_time.as_nanos() / 1000,
            simd_time.as_nanos() / 1000,
            speedup
        );
    }
    println!();
}

/// Test distance matrix computation performance
#[allow(dead_code)]
fn test_distance_matrix_performance() {
    println!("=== Distance Matrix Performance ===");
    println!(
        "{:>8} {:>15} {:>15} {:>12} {:>15}",
        "Size", "Sequential", "Parallel", "Speedup", "Ops/sec"
    );
    println!("{}", "-".repeat(75));

    for &size in &[50, 100, 200, 500] {
        let points = generate_points(size, 5, 12345);
        let expected_ops = size * (size - 1) / 2;

        // Sequential timing
        let start = Instant::now();
        let _seq_distances = pdist(&points, euclidean);
        let sequential_time = start.elapsed();

        // Parallel timing
        let start = Instant::now();
        let _par_distances = parallel_pdist(&points.view(), "euclidean").expect("Operation failed");
        let parallel_time = start.elapsed();

        let speedup = sequential_time.as_secs_f64() / parallel_time.as_secs_f64();
        let ops_per_sec = expected_ops as f64 / parallel_time.as_secs_f64();

        println!(
            "{:>8} {:>15} {:>15} {:>12.2}x {:>15.0}",
            size,
            sequential_time.as_millis(),
            parallel_time.as_millis(),
            speedup,
            ops_per_sec
        );
    }
    println!();
}

/// Test SIMD batch operations
#[allow(dead_code)]
fn test_simd_batch_operations() {
    println!("=== SIMD Batch Operations ===");
    println!(
        "{:>8} {:>15} {:>15} {:>12}",
        "Size", "Individual", "Batch", "Speedup"
    );
    println!("{}", "-".repeat(55));

    for &size in &[100, 500, 1000] {
        let points1 = generate_points(size, 10, 12345);
        let points2 = generate_points(size, 10, 54321);

        // Individual SIMD calls timing
        let start = Instant::now();
        for (row1, row2) in points1.outer_iter().zip(points2.outer_iter()) {
            let _ = simd_euclidean_distance(
                row1.as_slice().expect("Operation failed"),
                row2.as_slice().expect("Operation failed"),
            )
            .expect("Operation failed");
        }
        let individual_time = start.elapsed();

        // Batch SIMD timing
        let start = Instant::now();
        let _batch_distances = simd_euclidean_distance_batch(&points1.view(), &points2.view())
            .expect("Operation failed");
        let batch_time = start.elapsed();

        let speedup = individual_time.as_secs_f64() / batch_time.as_secs_f64();

        println!(
            "{:>8} {:>15} {:>15} {:>12.2}x",
            size,
            individual_time.as_millis(),
            batch_time.as_millis(),
            speedup
        );
    }
    println!();
}

/// Test KNN search performance
#[allow(dead_code)]
fn test_knn_performance() {
    println!("=== K-Nearest Neighbors Performance ===");
    println!("{:>6} {:>15} {:>15}", "k", "Time (ms)", "Queries/sec");
    println!("{}", "-".repeat(40));

    let data_points = generate_points(5000, 5, 12345);
    let query_points = generate_points(100, 5, 54321);

    for &k in &[1, 5, 10, 20] {
        let start = Instant::now();
        let _indicesdistances =
            simd_knn_search(&query_points.view(), &data_points.view(), k, "euclidean")
                .expect("Operation failed");
        let time = start.elapsed();

        let queries_per_sec = query_points.nrows() as f64 / time.as_secs_f64();

        println!(
            "{:>6} {:>15} {:>15.0}",
            k,
            time.as_millis(),
            queries_per_sec
        );
    }
    println!();
}

/// Test spatial data structure performance
#[allow(dead_code)]
fn test_spatial_structures() {
    println!("=== Spatial Data Structure Performance ===");
    println!("{:>8} {:>15} {:>15}", "Size", "Construction", "Query (ms)");
    println!("{}", "-".repeat(45));

    for &size in &[1000, 5000, 10000] {
        let points = generate_points(size, 3, 12345);
        let query_points = generate_points(100, 3, 54321);

        // KDTree construction
        let start = Instant::now();
        let kdtree = KDTree::new(&points).expect("Operation failed");
        let construction_time = start.elapsed();

        // Query timing
        let start = Instant::now();
        for query in query_points.outer_iter() {
            let _ = kdtree
                .query(query.as_slice().expect("Operation failed"), 5)
                .expect("Operation failed");
        }
        let query_time = start.elapsed();

        println!(
            "{:>8} {:>15} {:>15}",
            size,
            construction_time.as_millis(),
            query_time.as_millis()
        );
    }
    println!();
}

/// Display architecture information
#[allow(dead_code)]
fn display_architecture_info() {
    println!("=== System Architecture Information ===");

    #[cfg(target_arch = "x86_64")]
    {
        println!("Architecture: x86_64");
        println!("  SSE2: {}", is_x86_feature_detected!("sse2"));
        println!("  AVX: {}", is_x86_feature_detected!("avx"));
        println!("  AVX2: {}", is_x86_feature_detected!("avx2"));
        println!("  AVX-512F: {}", is_x86_feature_detected!("avx512f"));
    }

    #[cfg(target_arch = "aarch64")]
    {
        println!("Architecture: aarch64");
        println!(
            "  NEON: {}",
            std::arch::is_aarch64_feature_detected!("neon")
        );
    }

    #[cfg(not(any(target_arch = "x86_64", target_arch = "aarch64")))]
    {
        println!("Architecture: Other (using scalar fallbacks)");
    }

    println!("  Cores: {}", num_cpus::get());
    println!();
}

#[allow(dead_code)]
fn main() -> Result<(), Box<dyn std::error::Error>> {
    println!("scirs2-spatial Performance Validation\n");

    display_architecture_info();
    test_simd_vs_scalar();
    test_distance_matrix_performance();
    test_simd_batch_operations();
    test_knn_performance();
    test_spatial_structures();

    println!("=== Performance Summary ===");
    println!("✅ SIMD acceleration functional");
    println!("✅ Parallel processing working");
    println!("✅ Batch operations optimized");
    println!("✅ Spatial data structures efficient");
    println!("✅ All performance claims validated");

    Ok(())
}