rlu 0.7.0

Sparse LU Decomposition (Gilbert-Peierls)
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

use std::iter::zip;

fn main() {
    let n = 5;
    let mut a_mat: rlu::Matrix<usize, f64> = vec![vec![]; n];
    for i in 0..n {
        if i == n - 1 {
            a_mat[i].push((0, 1.0));
        }
        if i >= 2 {
            a_mat[i].push((i - 2, 3.0));
        }
        if i >= 1 {
            a_mat[i].push((i - 1, 4.0));
        }
        a_mat[i].push((i, 5.0));
        if i + 1 < n {
            a_mat[i].push((i + 1, 2.0));
        }
        if i == 0 {
            a_mat[i].push((n - 1, 1.0));
        }
    }
    // M = |5 4 3     1|
    //     |2 5 4 3    |
    //     |  2 5 4 3  |
    //     |    2 5 4 3|
    //     |      2 5 4|
    //     |1       2 5|
    #[cfg(feature = "debug")]
    print!("A =\n{}", rlu::matrix_table(&a_mat));

    let (n, rowidx, colptr, values) = matrix_to_csc(&a_mat);

    // let p = None;
    let mut q: Vec<usize> = (0..n).map(|i| i).collect();
    // q.swap(0, 0);
    q.swap(0, 1);
    q.swap(2, 3);
    // q.swap(0, 4);
    // q.swap(5, 6);
    println!("q = {:?}", q);

    let b0: Vec<f64> = (1..=n).map(|i| i as f64).collect();
    // x = [1,2,...,n]'
    // let b0: Vec<f64> = (0..n).map(|i| 1.0 + i as f64 / n as f64).collect();
    // x = [1,...,2]'
    println!("b = {:?}", b0);

    // let pivot = false;
    // let pivot = true;

    {
        let mut x = b0.clone();
        //         let (l_mat, u_mat, p) = lu_decomposition(&a_mat, Some(&q), pivot);
        //
        //         println!("p = {:?}", p);
        //
        //         let mut b = vec![0.0; n];
        //         for i in 0..n {
        //             b[p[i].unwrap()] = b0[i];
        //             // b[i] = b0[i];
        //         }
        //
        //         lsolve(&l_mat, &mut b);
        //         usolve(&u_mat, &mut b);
        //
        //         let mut x = vec![0.0; n]; // inverse permutation
        //         for i in 0..n {
        //             x[q[i]] = b[i];
        //             // x[i] = b[i];
        //         }

        rlu::solve(n, &rowidx, &colptr, &values, Some(&q), &mut x, false).unwrap();

        // Matrix-vector multiply b2 = A*x and print residual.
        let mut b2 = vec![0.0; n];
        for j in 0..n {
            for (i, aij) in &a_mat[j] {
                b2[*i] += *aij * x[j];
            }
        }
        b2.iter().for_each(|bi| println!("{}", bi));
        println!(
            "resid: {}",
            zip(b2, &b0)
                .map(|(b2, b0)| f64::abs(b2 - b0))
                .max_by(|a, b| a.partial_cmp(b).unwrap())
                .unwrap()
        );
    }
    #[cfg(feature = "ignored")]
    {
        let mut b = b0.clone();
        let (l_mat, u_mat): (Matrix, Matrix) = lu_decomposition(&a_mat, None);
        rlu::utsolve(&u_mat, &mut b);
        rlu::ltsolve(&l_mat, &mut b);
        let x = b;

        let mut b2 = vec![0.0; n];
        for i in 0..n {
            for (j, aij) in &a_mat[i] {
                b2[i] += *aij * x[*j];
            }
        }
        println!("{:?}", b2);
    }
}

pub fn matrix_to_csc<I: rlu::Int, S: rlu::Scalar>(
    m: &rlu::Matrix<I, S>,
) -> (usize, Vec<I>, Vec<I>, Vec<S>) {
    let n = m.len();
    let nnz = m.iter().map(|c| c.len()).fold(0, |acc, e| acc + e);
    let mut rowidx: Vec<I> = Vec::with_capacity(nnz);
    let mut colptr: Vec<I> = Vec::with_capacity(n + 1);
    let mut values: Vec<S> = Vec::with_capacity(nnz);

    let mut idxptr: usize = 0;
    for col in m {
        colptr.push(I::from_usize(idxptr));
        for (j, x) in col {
            values.push(*x);
            rowidx.push(*j);
            idxptr += 1
        }
    }
    colptr.push(I::from_usize(idxptr));

    (n, rowidx, colptr, values)
}