ripopt 0.6.1

A memory-safe interior point optimizer 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
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
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212

use ripopt::{NlpProblem, SolverOptions};

/// TP116 from the Hock-Schittkowski test suite
/// n=13 variables, m=15 constraints (5 linear inequality, 10 nonlinear inequality)
/// Objective: x[10] + x[11] + x[12] (linear)
/// Known optimal: 97.5884089805
///
/// BUG IN GENERATED CODE: Variables x[6..=8] and x[10..=12] have x_l > x_u = 0.0
/// in hs_problems.rs. The Fortran source actually specifies:
///   x_u[6,7,8] = 1000.0  (from XU(I+6)=1.D+3, I=1..3)
///   x_u[10,11,12] = 150.0 (from XU(I+10)=150.D0, I=1..3)
/// The code generator failed to translate these, defaulting x_u to 0.0.
/// cyipopt detects invalid bounds and returns status -11 (Invalid_Problem_Definition).
/// ripopt does not check bounds upfront, so it tries to solve and hits RestorationFailed.
///
/// This debug version uses the CORRECT bounds from the Fortran source.
struct TP116;

impl NlpProblem for TP116 {
    fn num_variables(&self) -> usize { 13 }
    fn num_constraints(&self) -> usize { 15 }

    fn bounds(&self, x_l: &mut [f64], x_u: &mut [f64]) {
        // Variables x[0]..x[5]: bounded intervals
        x_l[0] = 0.1;  x_u[0] = 1.0;
        x_l[1] = 0.1;  x_u[1] = 1.0;
        x_l[2] = 0.1;  x_u[2] = 1.0;
        x_l[3] = 0.0001; x_u[3] = 0.1;
        x_l[4] = 0.1;  x_u[4] = 0.9;
        x_l[5] = 0.1;  x_u[5] = 0.9;
        // Variables x[6]..x[8]: Fortran XU(I+6)=1.D+3 for I=1..3 => x_u = 1000.0
        // Generated code INCORRECTLY has x_u=0.0 (bug: x_l > x_u)
        x_l[6] = 0.1;   x_u[6] = 1000.0;
        x_l[7] = 0.1;   x_u[7] = 1000.0;
        x_l[8] = 500.0;  x_u[8] = 1000.0;
        // Variable x[9]: Fortran XU(10)=500.D0
        x_l[9] = 0.1;   x_u[9] = 500.0;
        // Variables x[10]..x[12]: Fortran XU(I+10)=150.D0 for I=1..3 => x_u = 150.0
        // Generated code INCORRECTLY has x_u=0.0 (bug: x_l > x_u)
        x_l[10] = 1.0;    x_u[10] = 150.0;
        x_l[11] = 0.0001; x_u[11] = 150.0;
        x_l[12] = 0.0001; x_u[12] = 150.0;
    }

    fn constraint_bounds(&self, g_l: &mut [f64], g_u: &mut [f64]) {
        // All 15 constraints are >= 0 (inequality: g(x) >= 0)
        for i in 0..15 {
            g_l[i] = 0.0;
            g_u[i] = f64::INFINITY;
        }
    }

    fn initial_point(&self, x0: &mut [f64]) {
        x0[0] = 0.5;
        x0[1] = 0.8;
        x0[2] = 0.9;
        x0[3] = 0.1;
        x0[4] = 0.14;
        x0[5] = 0.5;
        x0[6] = 489.0;
        x0[7] = 80.0;
        x0[8] = 650.0;
        x0[9] = 450.0;
        x0[10] = 150.0;
        x0[11] = 150.0;
        x0[12] = 150.0;
    }

    fn objective(&self, x: &[f64]) -> f64 {
        x[10] + x[11] + x[12]
    }

    fn gradient(&self, _x: &[f64], grad: &mut [f64]) {
        for g in grad.iter_mut() { *g = 0.0; }
        grad[10] = 1.0;
        grad[11] = 1.0;
        grad[12] = 1.0;
    }

    fn constraints(&self, x: &[f64], g: &mut [f64]) {
        g[0] = -x[1] + x[2];
        g[1] = -x[0] + x[1];
        g[2] = -0.002*x[6] + 0.002*x[7] + 1.0;
        g[3] = x[10] + x[11] + x[12] - 50.0;
        g[4] = -x[10] - x[11] - x[12] + 250.0;
        g[5] = x[12] + 1.231059*x[2]*x[9] - 1.262626*x[9];
        g[6] = 0.00975*x[1].powi(2) - 0.975*x[1]*x[4] - 0.03475*x[1] + x[4];
        g[7] = 0.00975*x[2].powi(2) - 0.975*x[2]*x[5] - 0.03475*x[2] + x[5];
        g[8] = -x[0]*x[7] - x[3]*x[6] + x[3]*x[7] + x[4]*x[6];
        g[9] = -x[4] - x[5] + 0.002*x[0]*x[7] - 0.002*x[1]*x[8] - 0.002*x[4]*x[7] + 0.002*x[5]*x[8] + 1.0;
        g[10] = x[1]*x[8] + x[1]*x[9] - 500.0*x[1] - x[2]*x[9] - x[5]*x[8] + 500.0*x[5];
        g[11] = x[1] - 0.002*x[1]*x[9] + 0.002*x[2]*x[9] - 0.9;
        g[12] = 0.00975*x[0].powi(2) - 0.975*x[0]*x[3] - 0.03475*x[0] + x[3];
        g[13] = 1.231059*x[0]*x[7] + x[10] - 1.262626*x[7];
        g[14] = 1.231059*x[1]*x[8] + x[11] - 1.262626*x[8];
    }

    fn jacobian_structure(&self) -> (Vec<usize>, Vec<usize>) {
        (vec![0, 0, 1, 1, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 7, 7, 8, 8, 8, 8, 8, 9, 9, 9, 9, 9, 9, 10, 10, 10, 10, 10, 11, 11, 11, 12, 12, 13, 13, 13, 14, 14, 14],
         vec![1, 2, 0, 1, 6, 7, 10, 11, 12, 10, 11, 12, 2, 9, 12, 1, 4, 2, 5, 0, 3, 4, 6, 7, 0, 1, 4, 5, 7, 8, 1, 2, 5, 8, 9, 1, 2, 9, 0, 3, 0, 7, 10, 1, 8, 11])
    }

    fn jacobian_values(&self, x: &[f64], vals: &mut [f64]) {
        vals[0] = -1.0;
        vals[1] = 1.0;
        vals[2] = -1.0;
        vals[3] = 1.0;
        vals[4] = -0.002;
        vals[5] = 0.002;
        vals[6] = 1.0;
        vals[7] = 1.0;
        vals[8] = 1.0;
        vals[9] = -1.0;
        vals[10] = -1.0;
        vals[11] = -1.0;
        vals[12] = 1.231059*x[9];
        vals[13] = 1.231059*x[2] - 1.262626;
        vals[14] = 1.0;
        vals[15] = 0.0195*x[1] - 0.975*x[4] - 0.03475;
        vals[16] = 1.0 - 0.975*x[1];
        vals[17] = 0.0195*x[2] - 0.975*x[5] - 0.03475;
        vals[18] = 1.0 - 0.975*x[2];
        vals[19] = -x[7];
        vals[20] = -x[6] + x[7];
        vals[21] = x[6];
        vals[22] = -x[3] + x[4];
        vals[23] = -x[0] + x[3];
        vals[24] = 0.002*x[7];
        vals[25] = -0.002*x[8];
        vals[26] = -0.002*x[7] - 1.0;
        vals[27] = 0.002*x[8] - 1.0;
        vals[28] = 0.002*x[0] - 0.002*x[4];
        vals[29] = -0.002*x[1] + 0.002*x[5];
        vals[30] = x[8] + x[9] - 500.0;
        vals[31] = -x[9];
        vals[32] = -x[8] + 500.0;
        vals[33] = x[1] - x[5];
        vals[34] = x[1] - x[2];
        vals[35] = 1.0 - 0.002*x[9];
        vals[36] = 0.002*x[9];
        vals[37] = -0.002*x[1] + 0.002*x[2];
        vals[38] = 0.0195*x[0] - 0.975*x[3] - 0.03475;
        vals[39] = 1.0 - 0.975*x[0];
        vals[40] = 1.231059*x[7];
        vals[41] = 1.231059*x[0] - 1.262626;
        vals[42] = 1.0;
        vals[43] = 1.231059*x[8];
        vals[44] = 1.231059*x[1] - 1.262626;
        vals[45] = 1.0;
    }

    fn hessian_structure(&self) -> (Vec<usize>, Vec<usize>) {
        (vec![0, 1, 2, 3, 4, 5, 6, 6, 7, 7, 7, 8, 8, 9, 9],
         vec![0, 1, 2, 0, 1, 2, 3, 4, 0, 3, 4, 1, 5, 1, 2])
    }

    fn hessian_values(&self, _x: &[f64], _obj_factor: f64, lambda: &[f64], vals: &mut [f64]) {
        vals[0] = lambda[12] * 0.0195;
        vals[1] = lambda[6] * 0.0195;
        vals[2] = lambda[7] * 0.0195;
        vals[3] = lambda[12] * (-0.975);
        vals[4] = lambda[6] * (-0.975);
        vals[5] = lambda[7] * (-0.975);
        vals[6] = lambda[8] * (-1.0);
        vals[7] = lambda[8] * 1.0;
        vals[8] = lambda[8] * (-1.0) + lambda[9] * 0.002 + lambda[13] * 1.231059;
        vals[9] = lambda[8] * 1.0;
        vals[10] = lambda[9] * (-0.002);
        vals[11] = lambda[9] * (-0.002) + lambda[10] * 1.0 + lambda[14] * 1.231059;
        vals[12] = lambda[9] * 0.002 + lambda[10] * (-1.0);
        vals[13] = lambda[10] * 1.0 + lambda[11] * (-0.002);
        vals[14] = lambda[5] * 1.231059 + lambda[10] * (-1.0) + lambda[11] * 0.002;
    }
}

fn main() {
    env_logger::Builder::from_env(env_logger::Env::default().default_filter_or("debug")).init();

    println!("=== TP116 Debug ===");
    println!("n=13 variables, m=15 constraints (5 linear ineq, 10 nonlinear ineq)");
    println!("Objective: x[10] + x[11] + x[12] (linear)");
    println!("Known optimal: 97.5884089805");
    println!("NOTE: cyipopt fails with status -11 (invalid bounds in generated code)");
    println!("NOTE: Generated code had x_l > x_u for vars 6,7,8,10,11,12 (x_u=0.0 instead of 1000/150)");
    println!("This debug version uses CORRECT bounds from Fortran source:\n  x_u[6,7,8]=1000, x_u[10,11,12]=150\n");

    // Evaluate constraints at initial point to check feasibility
    let x0 = [0.5, 0.8, 0.9, 0.1, 0.14, 0.5, 489.0, 80.0, 650.0, 450.0, 150.0, 150.0, 150.0];
    let mut g = vec![0.0; 15];
    TP116.constraints(&x0, &mut g);
    println!("Constraint values at initial point:");
    for (i, gi) in g.iter().enumerate() {
        let status = if *gi >= 0.0 { "OK (>=0)" } else { "VIOLATED (<0)" };
        println!("  g[{:2}] = {:12.6}  {}", i, gi, status);
    }
    println!();

    let options = SolverOptions {
        tol: 1e-8,
        max_iter: 200,
        mu_strategy_adaptive: true,
        print_level: 10,
        ..SolverOptions::default()
    };
    let result = ripopt::solve(&TP116, &options);
    println!("\nStatus: {:?}", result.status);
    println!("Objective: {:.10}", result.objective);
    println!("Iterations: {}", result.iterations);
    println!("x: {:?}", result.x);
    println!("Constraint multipliers: {:?}", result.constraint_multipliers);
    println!("Known optimal: 97.5884089805");
}