oximo 0.7.0

A Rust algebraic modeling library for solving optimization problems
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
213
214
215
216
217
218
219
220

//! Common adapter contracts.

#![cfg(any(
    feature = "highs",
    feature = "clarabel",
    feature = "pounce",
    feature = "gurobi",
    feature = "mosek",
    feature = "gams",
    feature = "baron"
))]

use oximo::prelude::*;

fn close(actual: Option<f64>, expected: f64) {
    let actual = actual.expect("missing result value");
    assert!((actual - expected).abs() < 1e-5, "got {actual}, expected {expected}");
}

fn range_and_offset<S: Solver>(mut solver: S, opts: &S::Options, require_dual: bool) {
    for slope in [2.0, -2.0] {
        let model = Model::new("range_offset_conformance");
        variable!(model, -10.0 <= x <= 10.0);
        constraint!(model, band, 1.0 <= 2.0 * x + 3.0 <= 9.0);
        objective!(model, Max, slope * x + 7.0);
        let result = solver.solve(&model, opts).unwrap();
        let expected_x = if slope > 0.0 { 3.0 } else { -1.0 };
        assert_eq!(result.termination, TerminationStatus::Optimal);
        assert_eq!(result.primal_status, PrimalStatus::OptimalPoint);
        close(result.value_of(x).unwrap(), expected_x);
        close(result.objective(), slope * expected_x + 7.0);
        close(result.best_bound, slope * expected_x + 7.0);
        assert!(result.gap.is_none());
        if require_dual {
            assert_eq!(result.dual_status, DualStatus::FeasiblePoint);
            // Both the native interval and split-row representations must
            // return sensitivity in the original row and objective units.
            close(result.dual_of(model.constraint_handle("band").unwrap()).unwrap(), slope / 2.0);
        }
    }
}

fn parameter_refresh<S: Solver>(mut solver: S, opts: &S::Options) {
    let model = Model::new("parameter_refresh_conformance");
    variable!(model, 0.0 <= x <= 20.0);
    param!(model, a = 2.0);
    param!(model, shift = 3.0);
    param!(model, price = 2.0);
    param!(model, offset = 7.0);
    constraint!(model, cap, a * x + shift <= 9.0);
    objective!(model, Max, price * x + offset);
    // Objective-only updates followed by row coefficient/constant updates.
    for (av, sv, pv, ov) in [(2.0, 3.0, 2.0, 7.0), (2.0, 3.0, 3.0, 11.0), (4.0, 1.0, 3.0, 11.0)] {
        model.set_param(a, av).unwrap();
        model.set_param(shift, sv).unwrap();
        model.set_param(price, pv).unwrap();
        model.set_param(offset, ov).unwrap();
        let result = solver.solve(&model, opts).unwrap();
        let expected_x = (9.0 - sv) / av;
        assert_eq!(result.termination, TerminationStatus::Optimal);
        close(result.value_of(x).unwrap(), expected_x);
        close(result.objective(), pv * expected_x + ov);
        close(result.best_bound, pv * expected_x + ov);
        assert!(result.gap.is_none());
    }
}

macro_rules! adapter {
    ($feature:literal, $name:ident, $solver:expr, $opts:expr, $dual:expr) => {
        #[cfg(feature = $feature)]
        mod $name {
            use super::*;
            #[test]
            fn maximization_offset_and_both_range_sides() {
                range_and_offset($solver, &$opts, $dual);
            }
            #[test]
            fn parameters_between_cold_solves() {
                parameter_refresh($solver, &$opts);
            }
            #[test]
            fn quadratic_cross_terms_and_objective_signs() {
                quadratic_cross_terms($solver, &$opts);
            }
            #[test]
            fn explicit_and_detected_cones() {
                cone_forms($solver, &$opts);
            }
        }
    };
}

fn quadratic_cross_terms<S: Solver>(mut solver: S, opts: &S::Options) {
    for maximize in [false, true] {
        let model = Model::new("quadratic_cross_terms");
        variable!(model, x >= 0.0);
        variable!(model, y >= 0.0);
        constraint!(model, balance, x + y == 1.0);
        let polynomial = 2.0 * x.powi(2) + x * y + y.powi(2) + x + y;
        if maximize {
            objective!(model, Max, 7.0 - polynomial);
        } else {
            objective!(model, Min, polynomial + 7.0);
        }
        let result = solver.solve(&model, opts).unwrap();
        assert!(result.has_solution());
        close(result.value_of(x).unwrap(), 0.25);
        close(result.value_of(y).unwrap(), 0.75);
        close(result.objective(), if maximize { 5.125 } else { 8.875 });
        assert!(matches!(
            result.termination,
            TerminationStatus::Optimal | TerminationStatus::LocallyOptimal
        ));
        if result.termination == TerminationStatus::Optimal {
            close(result.best_bound, result.objective().unwrap());
        }
    }
}

fn cone_forms<S: Solver>(mut solver: S, opts: &S::Options) {
    for explicit in [false, true] {
        let model = Model::new("cone_forms");
        variable!(model, x >= 0.0);
        variable!(model, y >= 0.0);
        variable!(model, t >= 0.0);
        constraint!(model, fix_x, x == 3.0);
        constraint!(model, fix_y, y == 4.0);
        if explicit {
            soc_constraint!(model, cone, [2.0 * x + 1.0, y + 20.0] <= 2.0 * t + 5.0);
        } else {
            constraint!(model, cone, x.powi(2) + y.powi(2) <= t.powi(2));
        }
        objective!(model, Min, t + 7.0);
        assert_eq!(model.kind(), ModelKind::SOCP);
        let result = solver.solve(&model, opts);
        if !solver.supports(ModelKind::SOCP) {
            assert!(matches!(result, Err(SolverError::UnsupportedKind(ModelKind::SOCP))));
            continue;
        }
        let result = result.unwrap();
        assert!(result.has_solution());
        close(result.value_of(x).unwrap(), 3.0);
        close(result.value_of(y).unwrap(), 4.0);
        // sqrt(7^2 + 24^2) = 25 for the explicit cone.
        close(result.value_of(t).unwrap(), if explicit { 10.0 } else { 5.0 });
        close(result.objective(), if explicit { 17.0 } else { 12.0 });
    }
}

adapter!("highs", highs, oximo::solvers::Highs, oximo::HighsOptions::default(), true);
adapter!("clarabel", clarabel, oximo::solvers::Clarabel, oximo::ClarabelOptions::default(), true);
adapter!("pounce", pounce, oximo::solvers::Pounce, oximo::pounce::PounceOptions::default(), true);
adapter!("gurobi", gurobi, oximo::solvers::Gurobi, oximo::GurobiOptions::default(), true);
adapter!("mosek", mosek, oximo::solvers::Mosek, oximo::MosekOptions::default(), true);
adapter!("gams", gams, oximo::solvers::Gams::new(), oximo::GamsOptions::default(), true);
adapter!("baron", baron, oximo::solvers::Baron::new(), oximo::BaronOptions::default(), false);

macro_rules! resident {
    ($feature:literal, $name:ident, $solver:expr, $opts:expr) => {
        #[cfg(feature = $feature)]
        #[test]
        fn $name() {
            parameter_refresh($solver.persistent(), &$opts);
        }
    };
}

resident!("highs", highs_resident, oximo::solvers::Highs, oximo::HighsOptions::default());
resident!(
    "clarabel",
    clarabel_resident,
    oximo::solvers::Clarabel,
    oximo::ClarabelOptions::default()
);
resident!(
    "pounce",
    pounce_resident,
    oximo::solvers::Pounce,
    oximo::pounce::PounceOptions::default()
);
resident!("gurobi", gurobi_resident, oximo::solvers::Gurobi, oximo::GurobiOptions::default());
resident!("mosek", mosek_resident, oximo::solvers::Mosek, oximo::MosekOptions::default());

#[cfg(feature = "clarabel")]
#[test]
fn clarabel_iteration_limit_without_a_certified_point() {
    let model = Model::new("limit_without_point");
    variable!(model, x >= 0.0);
    constraint!(model, cap, x <= 3.0);
    objective!(model, Max, x + 7.0);
    let result = oximo::solvers::Clarabel
        .solve(&model, &oximo::ClarabelOptions::default().max_iter(0))
        .unwrap();
    assert_eq!(result.termination, TerminationStatus::IterationLimit);
    assert_eq!(result.primal_status, PrimalStatus::NoSolution);
    assert!(result.solutions.is_empty());
    assert!(result.best_bound.is_none());
    assert!(result.gap.is_none());
    assert!(result.dual.is_empty());
}

#[cfg(feature = "pounce")]
#[test]
fn pounce_nlp_limit_requires_feasibility_evidence() {
    for initial in [0.0, 1.5] {
        let model = Model::new("limit_point_evidence");
        variable!(model, 0.0 <= x <= 3.0, initial = initial);
        constraint!(model, floor, x >= 1.0);
        objective!(model, Min, (x - 2.0).powi(2) + 7.0);
        let opts = oximo::pounce::PounceOptions::default()
            .solver_selection(oximo::pounce::PounceSolverSelection::Nlp)
            .max_iter(0);
        let result = oximo::solvers::Pounce.solve(&model, &opts).unwrap();
        assert_eq!(result.termination, TerminationStatus::IterationLimit);
        assert_eq!(result.has_solution(), initial >= 1.0);
        assert!(result.best_bound.is_none());
        assert!(result.gap.is_none());
        assert!(result.dual.is_empty());
    }
}