tide-maxflow 0.1.0

Tide max flow algorithm — a push-pull-relabel variant with O(1) array-based data structures
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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185

//! Cross-validation binary: generate the same deterministic graphs as
//! the Haskell Validate.hs and output (graph_id, name, n, m, max_flow, steps).

use tide_maxflow::max_flow;

fn layered_graph(layers: usize, width: usize) -> (usize, usize, usize, Vec<(usize, usize, i64)>) {
    let s = 0;
    let t = layers * width + 1;
    let n = t + 1;
    let vid = |l: usize, p: usize| l * width + p + 1;
    let mut edges = Vec::new();
    for p in 0..width {
        edges.push((s, vid(0, p), (50 + p) as i64));
    }
    for l in 0..layers - 1 {
        for p in 0..width {
            for d in 0..=2 {
                edges.push((vid(l, p), vid(l + 1, (p + d) % width), (10 + p + d) as i64));
            }
        }
    }
    for p in 0..width {
        edges.push((vid(layers - 1, p), t, (50 + p) as i64));
    }
    (n, s, t, edges)
}

fn grid_graph(rows: usize, cols: usize) -> (usize, usize, usize, Vec<(usize, usize, i64)>) {
    let s = 0;
    let t = rows * cols + 1;
    let n = t + 1;
    let vid = |r: usize, c: usize| r * cols + c + 1;
    let mut edges = Vec::new();
    for c in 0..cols {
        edges.push((s, vid(0, c), (50 + c) as i64));
    }
    for r in 0..rows {
        for c in 0..cols - 1 {
            edges.push((vid(r, c), vid(r, c + 1), (5 + r + c) as i64));
        }
    }
    for r in 0..rows - 1 {
        for c in 0..cols {
            edges.push((vid(r, c), vid(r + 1, c), (5 + r + c) as i64));
        }
    }
    for c in 0..cols {
        edges.push((vid(rows - 1, c), t, (50 + c) as i64));
    }
    (n, s, t, edges)
}

fn bipartite_graph(left: usize, right: usize) -> (usize, usize, usize, Vec<(usize, usize, i64)>) {
    let s = 0;
    let t = left + right + 1;
    let n = t + 1;
    let mut edges = Vec::new();
    for i in 0..left {
        edges.push((s, 1 + i, (100 + (i * 7) % 50) as i64));
    }
    for i in 0..left {
        for j in 0..right {
            edges.push((1 + i, 1 + left + j, (10 + (i * 13 + j * 7) % 40) as i64));
        }
    }
    for j in 0..right {
        edges.push((1 + left + j, t, (100 + (j * 11) % 50) as i64));
    }
    (n, s, t, edges)
}

fn random_graph(
    n_inner: usize,
    edge_prob_pct: usize,
    max_cap: i64,
) -> (usize, usize, usize, Vec<(usize, usize, i64)>) {
    let s = 0;
    let t = n_inner + 1;
    let n = t + 1;
    let mut edges = Vec::new();
    let mut lcg: u64 = 12345;

    // Source edges
    for v in 1..=n_inner {
        lcg = lcg.wrapping_mul(6364136223846793005).wrapping_add(1);
        let cap = (lcg % max_cap as u64) as i64 + 1;
        edges.push((s, v, cap));
    }
    // Inner edges
    for u in 1..=n_inner {
        for v in 1..=n_inner {
            if u != v {
                lcg = lcg.wrapping_mul(6364136223846793005).wrapping_add(1);
                if (lcg % 100) < edge_prob_pct as u64 {
                    lcg = lcg.wrapping_mul(6364136223846793005).wrapping_add(1);
                    let cap = (lcg % max_cap as u64) as i64 + 1;
                    edges.push((u, v, cap));
                }
            }
        }
    }
    // Sink edges
    for v in 1..=n_inner {
        lcg = lcg.wrapping_mul(6364136223846793005).wrapping_add(1);
        let cap = (lcg % max_cap as u64) as i64 + 1;
        edges.push((v, t, cap));
    }
    (n, s, t, edges)
}

fn parallel_chains(k: usize, len: usize) -> (usize, usize, usize, Vec<(usize, usize, i64)>) {
    let s = 0;
    let t = k * len + 1;
    let n = t + 1;
    let mut edges = Vec::new();
    for chain in 0..k {
        let first = 1 + chain * len;
        let last = first + len - 1;
        edges.push((s, first, (50 + chain * 3) as i64));
        for i in 0..len - 1 {
            edges.push((first + i, first + i + 1, (20 + chain + i) as i64));
        }
        edges.push((last, t, (50 + chain * 3) as i64));
    }
    (n, s, t, edges)
}

fn main() {
    println!("graph_id,name,n,m,max_flow,steps");

    let graphs: Vec<(&str, (usize, usize, usize, Vec<(usize, usize, i64)>))> = vec![
        ("layered_5x5", layered_graph(5, 5)),
        ("layered_10x10", layered_graph(10, 10)),
        ("layered_20x10", layered_graph(20, 10)),
        ("layered_50x50", layered_graph(50, 50)),
        ("grid_5x5", grid_graph(5, 5)),
        ("grid_10x10", grid_graph(10, 10)),
        ("grid_20x20", grid_graph(20, 20)),
        ("grid_50x50", grid_graph(50, 50)),
        ("bipartite_5x5", bipartite_graph(5, 5)),
        ("bipartite_10x10", bipartite_graph(10, 10)),
        ("bipartite_20x20", bipartite_graph(20, 20)),
        ("bipartite_50x50", bipartite_graph(50, 50)),
        ("random_20_10pct", random_graph(20, 10, 100)),
        ("random_20_30pct", random_graph(20, 30, 100)),
        ("random_50_10pct", random_graph(50, 10, 100)),
        ("random_50_30pct", random_graph(50, 30, 100)),
        ("chains_5x10", parallel_chains(5, 10)),
        ("chains_10x20", parallel_chains(10, 20)),
        ("chains_20x10", parallel_chains(20, 10)),
        (
            "clrs",
            (
                6,
                0,
                5,
                vec![
                    (0, 1, 16),
                    (0, 2, 13),
                    (1, 2, 10),
                    (1, 3, 12),
                    (2, 1, 4),
                    (2, 4, 14),
                    (3, 2, 9),
                    (3, 5, 20),
                    (4, 3, 7),
                    (4, 5, 4),
                ],
            ),
        ),
    ];

    for (gid, (name, (n, s, t, edges))) in graphs.iter().enumerate() {
        let result = max_flow(*n, *s, *t, edges);
        println!(
            "{},{},{},{},{},{}",
            gid,
            name,
            n,
            edges.len(),
            result.flow,
            result.steps
        );
    }
}