pso 0.2.0

Particle Swarm Optimizer
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
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162

use rand::distributions::Standard;
use rand::rngs::SmallRng;
use rand::Rng;

use super::{StopCondition, TransientBehavior};
use std::f32::consts::PI;

const DEFAULT_PERIOD: usize = 5000;

#[derive(Clone, Debug)]
pub struct TransientModel {
    num_variables: usize,
    num_points: usize,
    period: usize,
    momentums: Vec<f64>,
    motion_coeffs: Vec<Vec<f64>>,
    rand_magnitudes: Vec<f32>,
}

impl TransientModel {
    pub fn new(
        behavior: &TransientBehavior,
        stop_condition: &StopCondition,
        num_variables: usize,
        motion_coefficients: Vec<f64>,
        momentum_magnitude: f64,
    ) -> Self {
        let period = match stop_condition {
            StopCondition::Iterations(iters) => *iters,
            StopCondition::Cost(_) => DEFAULT_PERIOD,
            StopCondition::Both(iters, _) => *iters,
        };

        let num_directions = motion_coefficients.len();
        let num_points = num_directions * num_variables;

        let momentums = Self::gen_momentum_vals(period, behavior.momentum_mode, momentum_magnitude);
        let motion_coeffs = match num_directions {
            2 => Self::motion_coeffs_sol(period, behavior.motion_mode, &motion_coefficients),
            3 => Self::motion_coeffs_syn(period, behavior.motion_mode, &motion_coefficients),
            _ => panic!("Motion Coefficients vector must have a length of 2 or 3!"),
        };
        let rand_magnitudes = Self::gen_rand_mags(period, behavior.stochastic_mode);

        // println!("momentums: {:?}", momentums);

        // println!("motions: {:?}", motion_coeffs);

        Self {
            num_variables,
            num_points,
            period,
            momentums,
            motion_coeffs,
            rand_magnitudes,
        }
    }

    pub fn rand_vecs(&self, iter: usize, rng: &mut SmallRng) -> Vec<Vec<f64>> {
        let index = iter % self.period;
        let rand_mag = self.rand_magnitudes[index];

        let samples: Vec<f32> = rng.sample_iter(Standard).take(self.num_points).collect();
        samples
            .chunks_exact(self.num_variables)
            .map(|dir_chunk| dir_chunk.iter().map(|s| (s * rand_mag) as f64).collect())
            .collect()
    }

    pub fn momentum(&self, iter: usize) -> f64 {
        let index = iter % self.period;
        self.momentums[index]
    }

    pub fn motion_coeffs(&self, iter: usize) -> &[f64] {
        let index = iter % self.period;
        &self.motion_coeffs[index][..]
    }

    fn gen_momentum_vals(
        period: usize,
        fund_momentum_mode: u8,
        momentum_magnitude: f64,
    ) -> Vec<f64> {
        //let progress = (iter as f32) / self.period;
        let period_float = period as f32;
        // PI, 3*PI, 5*PI, 7*PI, ...
        let frequency = PI * (1.0 + ((fund_momentum_mode as f32) * 2.0));

        (0..period)
            .map(|i| {
                let t = (i as f32) / period_float;
                (((frequency * t).cos() * 0.45) + 0.55) as f64 * momentum_magnitude
            })
            .collect()
    }

    fn gen_rand_mags(period: usize, fund_rand_mode: u8) -> Vec<f32> {
        let period_float = period as f32;
        // PI, 3*PI, 5*PI, 7*PI, ...
        let frequency = PI * (1.0 + ((fund_rand_mode as f32) * 2.0));

        (0..period)
            .map(|i| {
                let t = (i as f32) / period_float;
                ((frequency * t).cos() * 0.95) + 1.05
            })
            .collect()
    }

    fn motion_coeffs_syn(
        period: usize,
        fund_motion_mode: u8,
        base_coeffs: &Vec<f64>,
    ) -> Vec<Vec<f64>> {
        let period_float = period as f32;
        let fund_mode_float = fund_motion_mode as f32;

        let loc_freq = PI * (3.0 + (fund_mode_float * 2.0)); // 3*PI, 5*PI, 7*PI, ...
        let tri_freq = PI * (2.0 + (fund_mode_float * 2.0)); // 5*PI, 7*PI, 9*PI, ...
        let glo_freq = PI * (3.0 + (fund_mode_float * 2.0)); // 3*PI, 5*PI, 7*PI, ...

        (0..period)
            .map(|i| {
                let t = (i as f32) / period_float;

                let loc: f32 = ((loc_freq * t).cos() * 0.45) + 0.55;
                let tri: f32 = ((tri_freq * t).sin() * 0.53) + 0.47;
                let glo: f32 = ((glo_freq * (t + 1.0)).cos() * 0.53) + 0.47;

                vec![
                    (loc as f64) * base_coeffs[0],
                    (tri as f64) * base_coeffs[1],
                    (glo as f64) * base_coeffs[2],
                ]
            })
            .collect()
    }

    fn motion_coeffs_sol(
        period: usize,
        fund_motion_mode: u8,
        base_coeffs: &Vec<f64>,
    ) -> Vec<Vec<f64>> {
        let period_float = period as f32;
        let fund_mode_float = fund_motion_mode as f32;

        let loc_freq = PI * (3.0 + (fund_mode_float * 2.0)); // 3*PI, 5*PI, 7*PI, ...
        let tri_freq = PI * (3.0 + (fund_mode_float * 2.0)); // 5*PI, 7*PI, 9*PI, ...

        (0..period)
            .map(|i| {
                let t = (i as f32) / period_float;

                let loc: f32 = ((loc_freq * t).cos() * 0.45) + 0.55;
                let tri: f32 = ((tri_freq * (t + 1.0)).cos() * 0.55) + 0.45;

                vec![(loc as f64) * base_coeffs[0], (tri as f64) * base_coeffs[1]]
            })
            .collect()
    }
}