solvr 0.2.0

Advanced computing library for real-world problem solving - optimization, differential equations, interpolation, statistics, and more
Documentation 
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

//! CPU implementations of global optimization algorithms.

mod basinhopping;
mod differential_evolution;
mod dual_annealing;
mod shgo;
mod simulated_annealing;

#[cfg(test)]
mod tests {
    use crate::optimize::global::{
        BasinHoppingAlgorithms, DifferentialEvolutionAlgorithms, DualAnnealingAlgorithms,
        GlobalOptions, SimulatedAnnealingAlgorithms,
    };
    use numr::error::Result;
    use numr::runtime::cpu::{CpuClient, CpuDevice, CpuRuntime};
    use numr::tensor::Tensor;

    fn sphere_tensor(x: &Tensor<CpuRuntime>) -> Result<f64> {
        let data: Vec<f64> = x.to_vec();
        Ok(data.iter().map(|&xi| xi * xi).sum())
    }

    #[test]
    fn test_compare_methods() {
        let device = CpuDevice::new();
        let client = CpuClient::new(device.clone());

        let lower = Tensor::<CpuRuntime>::from_slice(&[-5.0, -5.0], &[2], &device);
        let upper = Tensor::<CpuRuntime>::from_slice(&[5.0, 5.0], &[2], &device);

        // DE and DA converge faster
        let opts = GlobalOptions {
            max_iter: 500,
            seed: Some(42),
            ..Default::default()
        };

        let de_result = client
            .differential_evolution(sphere_tensor, &lower, &upper, &opts)
            .expect("DE failed");
        let da_result = client
            .dual_annealing(sphere_tensor, &lower, &upper, &opts)
            .expect("DA failed");

        assert!(de_result.fun < 1e-4);
        assert!(da_result.fun < 1e-4);

        // SA needs more iterations for similar convergence
        let sa_opts = GlobalOptions {
            max_iter: 5000,
            seed: Some(42),
            ..Default::default()
        };

        let sa_result = client
            .simulated_annealing(sphere_tensor, &lower, &upper, &sa_opts)
            .expect("SA failed");
        assert!(sa_result.fun < 1.0); // SA is stochastic, use relaxed tolerance
    }

    #[test]
    fn test_basinhopping() {
        let device = CpuDevice::new();
        let client = CpuClient::new(device.clone());

        let x0 = Tensor::<CpuRuntime>::from_slice(&[2.0, 2.0, 2.0], &[3], &device);
        let lower = Tensor::<CpuRuntime>::from_slice(&[-10.0, -10.0, -10.0], &[3], &device);
        let upper = Tensor::<CpuRuntime>::from_slice(&[10.0, 10.0, 10.0], &[3], &device);

        let opts = GlobalOptions {
            max_iter: 50,
            seed: Some(42),
            ..Default::default()
        };

        let result = client
            .basinhopping(sphere_tensor, &x0, &lower, &upper, &opts)
            .expect("basinhopping failed");
        assert!(result.fun < 1e-4);
    }
}