gspx 0.1.2

Sparse graph signal processing and spectral graph wavelets in Rust
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

use ndarray::{Array2, ArrayView1, ArrayView2, Zip};
use num_complex::Complex64;

use crate::error::GspError;

pub(super) fn arrays_close(a: ArrayView1<'_, f64>, b: ArrayView1<'_, f64>, atol: f64) -> bool {
    if a.len() != b.len() {
        return false;
    }
    a.iter().zip(b.iter()).all(|(x, y)| (*x - *y).abs() <= atol)
}

pub(super) fn scalar_close(a: f64, b: f64, rtol: f64) -> bool {
    (a - b).abs() <= rtol * a.abs().max(b.abs()).max(1.0)
}

pub(super) fn real_times_complex(
    a: ArrayView2<f64>,
    b: ArrayView2<Complex64>,
) -> Array2<Complex64> {
    let mut out = Array2::<Complex64>::zeros((a.nrows(), b.ncols()));
    for i in 0..a.nrows() {
        for j in 0..b.ncols() {
            let mut sum = Complex64::new(0.0, 0.0);
            for k in 0..a.ncols() {
                sum += Complex64::new(a[[i, k]], 0.0) * b[[k, j]];
            }
            out[[i, j]] = sum;
        }
    }
    out
}

pub(super) fn peak_local_max(
    image: ArrayView2<'_, f64>,
    min_distance: usize,
    num_peaks: usize,
) -> Vec<(usize, usize)> {
    let mut coords = Vec::<(usize, usize, f64)>::new();
    for r in 0..image.nrows() {
        for c in 0..image.ncols() {
            let center = image[[r, c]];
            if center <= 0.0 {
                continue;
            }
            let r_lo = r.saturating_sub(min_distance);
            let r_hi = (r + min_distance + 1).min(image.nrows());
            let c_lo = c.saturating_sub(min_distance);
            let c_hi = (c + min_distance + 1).min(image.ncols());
            let mut is_max = true;
            'outer: for rr in r_lo..r_hi {
                for cc in c_lo..c_hi {
                    if image[[rr, cc]] > center {
                        is_max = false;
                        break 'outer;
                    }
                }
            }
            if is_max {
                coords.push((r, c, center));
            }
        }
    }
    coords.sort_by(|a, b| b.2.partial_cmp(&a.2).unwrap_or(std::cmp::Ordering::Equal));
    let mut suppressed = vec![false; image.nrows() * image.ncols()];
    let mut out = Vec::new();
    for (r, c, _) in coords {
        let idx = r * image.ncols() + c;
        if suppressed[idx] {
            continue;
        }
        out.push((r, c));
        if out.len() >= num_peaks {
            break;
        }
        let r_lo = r.saturating_sub(min_distance);
        let r_hi = (r + min_distance + 1).min(image.nrows());
        let c_lo = c.saturating_sub(min_distance);
        let c_hi = (c + min_distance + 1).min(image.ncols());
        for rr in r_lo..r_hi {
            for cc in c_lo..c_hi {
                suppressed[rr * image.ncols() + cc] = true;
            }
        }
    }
    out
}

pub(super) fn gradient_real_axis(
    a: ArrayView2<'_, f64>,
    grid: ArrayView1<'_, f64>,
    axis: usize,
) -> Result<Array2<f64>, GspError> {
    let mut g = Array2::<f64>::zeros(a.raw_dim());
    if axis == 0 {
        if a.nrows() < 2 || grid.len() < 2 || a.nrows() != grid.len() {
            return Err(GspError::Dimensions(format!(
                "axis-0 gradient requires at least 2 rows and matching grid length; got matrix {}x{} and grid {}",
                a.nrows(),
                a.ncols(),
                grid.len()
            )));
        }
        for j in 0..a.ncols() {
            g[[0, j]] = (a[[1, j]] - a[[0, j]]) / (grid[1] - grid[0]);
            for i in 1..(a.nrows() - 1) {
                g[[i, j]] = (a[[i + 1, j]] - a[[i - 1, j]]) / (grid[i + 1] - grid[i - 1]);
            }
            g[[a.nrows() - 1, j]] = (a[[a.nrows() - 1, j]] - a[[a.nrows() - 2, j]])
                / (grid[grid.len() - 1] - grid[grid.len() - 2]);
        }
    } else {
        if a.ncols() < 2 || grid.len() < 2 || a.ncols() != grid.len() {
            return Err(GspError::Dimensions(format!(
                "axis-1 gradient requires at least 2 cols and matching grid length; got matrix {}x{} and grid {}",
                a.nrows(),
                a.ncols(),
                grid.len()
            )));
        }
        for i in 0..a.nrows() {
            g[[i, 0]] = (a[[i, 1]] - a[[i, 0]]) / (grid[1] - grid[0]);
            for j in 1..(a.ncols() - 1) {
                g[[i, j]] = (a[[i, j + 1]] - a[[i, j - 1]]) / (grid[j + 1] - grid[j - 1]);
            }
            g[[i, a.ncols() - 1]] = (a[[i, a.ncols() - 1]] - a[[i, a.ncols() - 2]])
                / (grid[grid.len() - 1] - grid[grid.len() - 2]);
        }
    }
    Ok(g)
}

pub(super) fn gradient_complex_axis(
    a: ArrayView2<'_, Complex64>,
    grid: ArrayView1<'_, f64>,
    axis: usize,
) -> Result<Array2<Complex64>, GspError> {
    let real = gradient_real_axis(a.mapv(|z| z.re).view(), grid, axis)?;
    let imag = gradient_real_axis(a.mapv(|z| z.im).view(), grid, axis)?;
    let mut out = Array2::<Complex64>::zeros(real.raw_dim());
    Zip::from(&mut out)
        .and(&real)
        .and(&imag)
        .for_each(|o, &r, &i| *o = Complex64::new(r, i));
    Ok(out)
}

pub(super) fn stddev(x: ArrayView1<'_, f64>) -> f64 {
    if x.is_empty() {
        return 0.0;
    }
    let mean = x.sum() / x.len() as f64;
    let var = x.iter().map(|v| (v - mean).powi(2)).sum::<f64>() / x.len() as f64;
    var.sqrt()
}