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
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
use ripopt::{NlpProblem, SolverOptions};
use std::f64::consts::PI;
/// TP374 from the Hock-Schittkowski test suite.
///
/// Minimize x[9]
/// subject to 35 nonlinear inequality constraints g_i(x) >= 0
///
/// The constraints involve trigonometric sums:
/// A(z, x) = sum_{k=1}^{9} x[k-1] * cos(k*z)
/// B(z, x) = sum_{k=1}^{9} x[k-1] * sin(k*z)
/// G(z, x) = A(z, x)^2 + B(z, x)^2
///
/// Constraints 0..9 (i=0..9): G(z_i, x) - (1 - x[9])^2 >= 0, z_i = pi/4 * (i * 0.1)
/// Constraints 10..19 (i=10..19): (1 + x[9])^2 - G(z_i, x) >= 0, z_i = pi/4 * ((i-10) * 0.1)
/// Constraints 20..34 (i=20..34): x[9]^2 - G(z_i, x) >= 0, z_i = pi/4 * (1.2 + (i-20) * 0.2)
///
/// n=10, m=35, all 35 constraints are nonlinear inequalities (ninl=35).
/// Known optimal: f* = 0.233264
///
/// NOTE: The generated code in hs_problems.rs has EMPTY constraint/jacobian bodies
/// because the code generator failed to translate the Fortran trigonometric
/// constraint functions. This file provides a correct hand-translated implementation
/// from the original Fortran source (PROB.FOR subroutine TP374).
struct TP374;
// Helper functions matching the Fortran TP374A, TP374B, TP374G
fn tp374_a(z: f64, x: &[f64]) -> f64 {
let mut val = 0.0;
for k in 1..=9 {
val += x[k - 1] * (k as f64 * z).cos();
}
val
}
fn tp374_b(z: f64, x: &[f64]) -> f64 {
let mut val = 0.0;
for k in 1..=9 {
val += x[k - 1] * (k as f64 * z).sin();
}
val
}
fn tp374_g(z: f64, x: &[f64]) -> f64 {
let a = tp374_a(z, x);
let b = tp374_b(z, x);
a * a + b * b
}
impl NlpProblem for TP374 {
fn num_variables(&self) -> usize { 10 }
fn num_constraints(&self) -> usize { 35 }
fn bounds(&self, x_l: &mut [f64], x_u: &mut [f64]) {
for i in 0..10 {
x_l[i] = f64::NEG_INFINITY;
x_u[i] = f64::INFINITY;
}
}
fn constraint_bounds(&self, g_l: &mut [f64], g_u: &mut [f64]) {
for i in 0..35 {
g_l[i] = 0.0;
g_u[i] = f64::INFINITY;
}
}
fn initial_point(&self, x0: &mut [f64]) {
for i in 0..10 {
x0[i] = 0.1;
}
}
fn objective(&self, x: &[f64]) -> f64 {
x[9]
}
fn gradient(&self, _x: &[f64], grad: &mut [f64]) {
for i in 0..9 { grad[i] = 0.0; }
grad[9] = 1.0;
}
fn constraints(&self, x: &[f64], g: &mut [f64]) {
// Constraints 0..9: G(z_i, x) - (1 - x[9])^2 >= 0
for i in 0..10 {
let z = PI / 4.0 * (i as f64 * 0.1);
g[i] = tp374_g(z, x) - (1.0 - x[9]).powi(2);
}
// Constraints 10..19: (1 + x[9])^2 - G(z_i, x) >= 0
for i in 10..20 {
let z = PI / 4.0 * ((i - 10) as f64 * 0.1);
g[i] = (1.0 + x[9]).powi(2) - tp374_g(z, x);
}
// Constraints 20..34: x[9]^2 - G(z_i, x) >= 0
for i in 20..35 {
let z = PI / 4.0 * (1.2 + (i - 20) as f64 * 0.2);
g[i] = x[9].powi(2) - tp374_g(z, x);
}
}
fn jacobian_structure(&self) -> (Vec<usize>, Vec<usize>) {
// Each of the 35 constraints depends on all 10 variables
let mut rows = Vec::with_capacity(35 * 10);
let mut cols = Vec::with_capacity(35 * 10);
for i in 0..35 {
for j in 0..10 {
rows.push(i);
cols.push(j);
}
}
(rows, cols)
}
fn jacobian_values(&self, x: &[f64], vals: &mut [f64]) {
let mut idx = 0;
// Constraints 0..9: g_i = G(z, x) - (1 - x[9])^2
// dg/dx[k-1] = 2*(A*cos(k*z) + B*sin(k*z)) for k=1..9
// dg/dx[9] = 2*(1 - x[9])
for i in 0..10 {
let z = PI / 4.0 * (i as f64 * 0.1);
let a = tp374_a(z, x);
let b = tp374_b(z, x);
for k in 1..=9 {
vals[idx] = 2.0 * (a * (k as f64 * z).cos() + b * (k as f64 * z).sin());
idx += 1;
}
vals[idx] = 2.0 * (1.0 - x[9]);
idx += 1;
}
// Constraints 10..19: g_i = (1 + x[9])^2 - G(z, x)
// dg/dx[k-1] = -2*(A*cos(k*z) + B*sin(k*z)) for k=1..9
// dg/dx[9] = 2*(1 + x[9])
for i in 10..20 {
let z = PI / 4.0 * ((i - 10) as f64 * 0.1);
let a = tp374_a(z, x);
let b = tp374_b(z, x);
for k in 1..=9 {
vals[idx] = -2.0 * (a * (k as f64 * z).cos() + b * (k as f64 * z).sin());
idx += 1;
}
vals[idx] = 2.0 * (1.0 + x[9]);
idx += 1;
}
// Constraints 20..34: g_i = x[9]^2 - G(z, x)
// dg/dx[k-1] = -2*(A*cos(k*z) + B*sin(k*z)) for k=1..9
// dg/dx[9] = 2*x[9]
for i in 20..35 {
let z = PI / 4.0 * (1.2 + (i - 20) as f64 * 0.2);
let a = tp374_a(z, x);
let b = tp374_b(z, x);
for k in 1..=9 {
vals[idx] = -2.0 * (a * (k as f64 * z).cos() + b * (k as f64 * z).sin());
idx += 1;
}
vals[idx] = 2.0 * x[9];
idx += 1;
}
}
fn hessian_structure(&self) -> (Vec<usize>, Vec<usize>) {
// 10x10 lower triangle: all pairs (i, j) with i >= j
let mut rows = Vec::new();
let mut cols = Vec::new();
for i in 0..10 {
for j in 0..=i {
rows.push(i);
cols.push(j);
}
}
(rows, cols)
}
fn hessian_values(&self, _x: &[f64], _obj_factor: f64, lambda: &[f64], vals: &mut [f64]) {
// Objective Hessian is zero (f = x[9] is linear).
// Constraint Hessians:
// For constraints 0..9 (group 1): g = A^2 + B^2 - (1-x9)^2
// d2g/dx[j]dx[k] = 2*(cos(j*z)*cos(k*z) + sin(j*z)*sin(k*z))
// = 2*cos((j-k)*z) for j,k in 1..9
// d2g/dx9^2 = 2
// cross terms dx[k]*dx9 = 0 for k<9
//
// For constraints 10..19 (group 2): g = (1+x9)^2 - A^2 - B^2
// d2g/dx[j]dx[k] = -2*cos((j-k)*z) for j,k in 1..9
// d2g/dx9^2 = 2
// cross terms dx[k]*dx9 = 0 for k<9
//
// For constraints 20..34 (group 3): g = x9^2 - A^2 - B^2
// d2g/dx[j]dx[k] = -2*cos((j-k)*z) for j,k in 1..9
// d2g/dx9^2 = 2
// cross terms dx[k]*dx9 = 0 for k<9
// Initialize to zero
for v in vals.iter_mut() { *v = 0.0; }
// Helper: map lower-triangle (i, j) with i >= j to flat index
// Index of (i,j) = i*(i+1)/2 + j
let lt = |i: usize, j: usize| -> usize { i * (i + 1) / 2 + j };
// Group 1: constraints 0..9
for ci in 0..10 {
let lam = lambda[ci];
if lam == 0.0 { continue; }
let z = PI / 4.0 * (ci as f64 * 0.1);
// x[j-1], x[k-1] block (j,k = 1..9, so indices 0..8)
for j in 1..=9usize {
for k in 1..=j {
let h = 2.0 * ((j as f64 - k as f64) * z).cos();
vals[lt(j - 1, k - 1)] += lam * h;
}
}
// x[9]^2 term: d2g/dx9^2 = 2
vals[lt(9, 9)] += lam * 2.0;
}
// Group 2: constraints 10..19
for ci in 10..20 {
let lam = lambda[ci];
if lam == 0.0 { continue; }
let z = PI / 4.0 * ((ci - 10) as f64 * 0.1);
for j in 1..=9usize {
for k in 1..=j {
let h = -2.0 * ((j as f64 - k as f64) * z).cos();
vals[lt(j - 1, k - 1)] += lam * h;
}
}
vals[lt(9, 9)] += lam * 2.0;
}
// Group 3: constraints 20..34
for ci in 20..35 {
let lam = lambda[ci];
if lam == 0.0 { continue; }
let z = PI / 4.0 * (1.2 + (ci - 20) as f64 * 0.2);
for j in 1..=9usize {
for k in 1..=j {
let h = -2.0 * ((j as f64 - k as f64) * z).cos();
vals[lt(j - 1, k - 1)] += lam * h;
}
}
vals[lt(9, 9)] += lam * 2.0;
}
}
}
fn main() {
env_logger::Builder::from_env(env_logger::Env::default().default_filter_or("debug"))
.init();
let problem = TP374;
// Verify constraint evaluation at initial point
let mut x0 = vec![0.0; 10];
problem.initial_point(&mut x0);
let mut g0 = vec![0.0; 35];
problem.constraints(&x0, &mut g0);
println!("Initial point: {:?}", x0);
println!("f(x0) = {}", problem.objective(&x0));
println!("Constraints at x0 (first 10): {:?}", &g0[..10]);
println!("Constraints at x0 (10..20): {:?}", &g0[10..20]);
println!("Constraints at x0 (20..35): {:?}", &g0[20..35]);
println!();
let options = SolverOptions {
print_level: 10,
max_iter: 3000,
mu_strategy_adaptive: true,
tol: 1e-8,
..SolverOptions::default()
};
let result = ripopt::solve(&problem, &options);
println!("\n=== RESULT ===");
println!("Status: {:?}", result.status);
println!("Objective: {:.10}", result.objective);
println!("x: {:?}", result.x);
println!("Iterations: {}", result.iterations);
println!("Known optimal: 0.233264");
if !result.x.is_empty() {
let mut g_final = vec![0.0; 35];
problem.constraints(&result.x, &mut g_final);
println!("\nConstraint values at solution:");
for (i, gv) in g_final.iter().enumerate() {
let status = if *gv < -1e-6 { " ** VIOLATED **" } else { "" };
println!(" g[{:2}] = {:+.8e}{}", i, gv, status);
}
}
}