dypdl_heuristic_search/search_algorithm/
breadth_first_search.rs

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
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
use super::data_structure::{exceed_bound, BfsNode, StateRegistry, SuccessorGenerator};
use super::rollout::get_solution_cost_and_suffix;
use super::search::{Parameters, Search, SearchInput, Solution};
use super::util::{update_bound_if_better, update_solution, TimeKeeper};
use dypdl::{variable_type, Transition, TransitionInterface};
use std::collections::VecDeque;
use std::error::Error;
use std::fmt;
use std::mem;
use std::rc::Rc;

/// Parameters for breadth-first search (BrFS).
#[derive(Debug, PartialEq, Clone, Copy, Default)]
pub struct BrfsParameters<T> {
    /// Keep nodes in all layers for duplicate detection.
    ///
    /// BrFS searches layer by layer, where the i th layer contains states that can be reached with i transitions.
    /// By default, BrFS only keeps states in the current layer to check for duplicates.
    /// If `keep_all_layers` is `true`, BrFS keeps states in all layers to check for duplicates.
    pub keep_all_layers: bool,
    /// Common parameters.
    pub parameters: Parameters<T>,
}

/// Breadth-first search solver.
///
/// This solver uses forward search based on the shortest path problem.
/// It only works with problems where the cost expressions are in the form of `cost + w`, `cost * w`, `max(cost, w)`, or `min(cost, w)`
/// where `cost` is `IntegerExpression::Cost`or `ContinuousExpression::Cost` and `w` is a numeric expression independent of `cost`.
///
/// It searches nodes in the breadth-first order.
///
/// Type parameter `N` is a node type that implements `BfsNode`.
/// Type parameter `E` is a type of a function that evaluates a transition and insert a successor node into a state registry.
/// The last argument of the function is the primal bound of the solution cost.
/// Type parameter `B` is a type of a function that combines the g-value (the cost to a state) and the base cost.
/// It should be the same function as the cost expression, e.g., `cost + base_cost` for `cost + w`.
///
/// # Examples
///
/// ```
/// use dypdl::prelude::*;
/// use dypdl_heuristic_search::{BrfsParameters, Search};
/// use dypdl_heuristic_search::search_algorithm::{
///     BreadthFirstSearch, FNode, SearchInput, SuccessorGenerator,
/// };
/// use std::rc::Rc;
///
/// let mut model = Model::default();
/// let variable = model.add_integer_variable("variable", 0).unwrap();
/// model.add_base_case(
///     vec![Condition::comparison_i(ComparisonOperator::Ge, variable, 1)]
/// ).unwrap();
/// let mut increment = Transition::new("increment");
/// increment.set_cost(IntegerExpression::Cost + 1);
/// increment.add_effect(variable, variable + 1).unwrap();
/// model.add_forward_transition(increment.clone()).unwrap();
/// let model = Rc::new(model);
///
/// let state = model.target.clone();
/// let cost = 0;
/// let h_evaluator = |_: &_| Some(0);
/// let f_evaluator = |g, h, _: &_| g + h;
/// let primal_bound = None;
/// let node = FNode::generate_root_node(
///     state,
///     cost,
///     &model,
///     &h_evaluator,
///     &f_evaluator,
///     primal_bound,
/// );
/// let generator = SuccessorGenerator::<Transition>::from_model(model.clone(), false);
/// let input = SearchInput {
///     node,
///     generator,
///     solution_suffix: &[],
/// };
/// let transition_evaluator =
///     move |node: &FNode<_>, transition, registry: &mut _, primal_bound| {
///         node.insert_successor_node(
///             transition,
///             registry,
///             &h_evaluator,
///             &f_evaluator,
///             primal_bound,
///         )
///     };
/// let base_cost_evaluator = |cost, base_cost| cost + base_cost;
/// let parameters = BrfsParameters::default();
///
/// let mut solver = BreadthFirstSearch::<_, FNode<_>, _, _>::new(
///     input, transition_evaluator, base_cost_evaluator, parameters,
/// );
/// let solution = solver.search().unwrap();
/// assert_eq!(solution.cost, Some(1));
/// assert_eq!(solution.transitions, vec![increment]);
/// assert!(!solution.is_infeasible);
/// ```
pub struct BreadthFirstSearch<'a, T, N, E, B, V = Transition>
where
    T: variable_type::Numeric + fmt::Display + Ord + 'static,
    N: BfsNode<T, V>,
    E: FnMut(&N, Rc<V>, &mut StateRegistry<T, N>, Option<T>) -> Option<(Rc<N>, bool)>,
    B: FnMut(T, T) -> T,
    V: TransitionInterface + Clone + Default,
    Transition: From<V>,
{
    generator: SuccessorGenerator<V>,
    suffix: &'a [V],
    transition_evaluator: E,
    base_cost_evaluator: B,
    keep_all_layers: bool,
    primal_bound: Option<T>,
    get_all_solutions: bool,
    quiet: bool,
    open: VecDeque<Rc<N>>,
    next_open: VecDeque<Rc<N>>,
    registry: StateRegistry<T, N>,
    layer_index: usize,
    layer_dual_bound: Option<T>,
    time_keeper: TimeKeeper,
    solution: Solution<T>,
}

impl<'a, T, N, E, B, V> BreadthFirstSearch<'a, T, N, E, B, V>
where
    T: variable_type::Numeric + fmt::Display + Ord + 'static,
    N: BfsNode<T, V> + Clone,
    E: FnMut(&N, Rc<V>, &mut StateRegistry<T, N>, Option<T>) -> Option<(Rc<N>, bool)>,
    B: FnMut(T, T) -> T,
    V: TransitionInterface + Clone + Default,
    Transition: From<V>,
{
    /// Create a new breadth-first search solver.
    pub fn new(
        input: SearchInput<'a, N, V>,
        transition_evaluator: E,
        base_cost_evaluator: B,
        parameters: BrfsParameters<T>,
    ) -> BreadthFirstSearch<'a, T, N, E, B, V> {
        let mut time_keeper = parameters
            .parameters
            .time_limit
            .map_or_else(TimeKeeper::default, TimeKeeper::with_time_limit);
        let get_all_solutions = parameters.parameters.get_all_solutions;
        let primal_bound = parameters.parameters.primal_bound;
        let quiet = parameters.parameters.quiet;
        let mut open = VecDeque::new();
        let next_open = VecDeque::new();
        let mut registry = StateRegistry::new(input.generator.model.clone());

        if let Some(capacity) = parameters.parameters.initial_registry_capacity {
            registry.reserve(capacity);
        }

        let mut solution = Solution::default();
        if let Some(node) = input.node.clone() {
            let result = registry.insert(node);
            open.push_back(result.information.unwrap());
            solution.generated += 1;
        } else {
            solution.is_infeasible = true;
        }

        time_keeper.stop();

        BreadthFirstSearch {
            generator: input.generator,
            suffix: input.solution_suffix,
            transition_evaluator,
            base_cost_evaluator,
            keep_all_layers: parameters.keep_all_layers,
            primal_bound,
            get_all_solutions,
            quiet,
            open,
            next_open,
            registry,
            layer_index: 0,
            layer_dual_bound: None,
            time_keeper,
            solution,
        }
    }
}

impl<'a, T, N, E, B, V> Search<T> for BreadthFirstSearch<'a, T, N, E, B, V>
where
    T: variable_type::Numeric + fmt::Display + Ord + 'static,
    N: BfsNode<T, V>,
    E: FnMut(&N, Rc<V>, &mut StateRegistry<T, N>, Option<T>) -> Option<(Rc<N>, bool)>,
    B: FnMut(T, T) -> T,
    V: TransitionInterface + Clone + Default,
    Transition: From<V>,
{
    fn search_next(&mut self) -> Result<(Solution<T>, bool), Box<dyn Error>> {
        if self.solution.is_terminated() {
            return Ok((self.solution.clone(), true));
        }

        self.time_keeper.start();
        let model = &self.generator.model;
        let suffix = self.suffix;

        loop {
            if self.open.is_empty() {
                if let Some(bound) = self.layer_dual_bound {
                    self.solution.time = self.time_keeper.elapsed_time();
                    update_bound_if_better(&mut self.solution, bound, model, self.quiet);
                }

                if !self.quiet {
                    println!(
                        "Searched layer: {}, expanded: {}, elapsed time: {}",
                        self.layer_index,
                        self.solution.expanded,
                        self.time_keeper.elapsed_time()
                    );
                }

                if self.next_open.is_empty() {
                    break;
                }

                mem::swap(&mut self.open, &mut self.next_open);

                if !self.keep_all_layers {
                    self.registry.clear();
                }

                self.layer_index += 1;
                self.layer_dual_bound = None;
            }

            while let Some(node) = self.open.pop_front() {
                if node.is_closed() {
                    continue;
                }
                node.close();

                if let Some(dual_bound) = node.bound(model) {
                    if exceed_bound(model, dual_bound, self.primal_bound) {
                        continue;
                    }
                }

                if let Some((cost, suffix)) = get_solution_cost_and_suffix(
                    model,
                    &*node,
                    suffix,
                    &mut self.base_cost_evaluator,
                ) {
                    if !exceed_bound(model, cost, self.primal_bound) {
                        self.primal_bound = Some(cost);
                        let time = self.time_keeper.elapsed_time();
                        update_solution(&mut self.solution, &*node, cost, suffix, time, self.quiet);
                        self.time_keeper.stop();

                        return Ok((self.solution.clone(), self.solution.is_optimal));
                    } else if self.get_all_solutions {
                        let mut solution = self.solution.clone();
                        let time = self.time_keeper.elapsed_time();
                        update_solution(&mut solution, &*node, cost, suffix, time, true);
                        self.time_keeper.stop();

                        return Ok((solution, false));
                    }
                    continue;
                }

                if self.time_keeper.check_time_limit(self.quiet) {
                    self.solution.time_out = true;
                    self.solution.time = self.time_keeper.elapsed_time();
                    self.time_keeper.stop();

                    return Ok((self.solution.clone(), true));
                }

                self.solution.expanded += 1;

                for transition in self.generator.applicable_transitions(node.state()) {
                    if let Some((successor, new_generated)) = (self.transition_evaluator)(
                        &node,
                        transition,
                        &mut self.registry,
                        self.primal_bound,
                    ) {
                        if let Some(bound) = successor.bound(model) {
                            if !exceed_bound(model, bound, self.layer_dual_bound) {
                                self.layer_dual_bound = Some(bound);
                            }
                        }

                        self.next_open.push_back(successor);

                        if new_generated {
                            self.solution.generated += 1;
                        }
                    }
                }
            }
        }

        self.solution.is_infeasible = self.solution.cost.is_none();
        self.solution.is_optimal = self.solution.cost.is_some();
        self.solution.best_bound = self.solution.cost;
        self.solution.time = self.time_keeper.elapsed_time();
        self.time_keeper.stop();
        Ok((self.solution.clone(), true))
    }
}