3
0
Fork 0
mirror of https://github.com/Z3Prover/z3 synced 2025-08-17 00:32:16 +00:00
z3/src/sat/sat_local_search.h
Nikolaj Bjorner af96e42724 fixing local search
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
2018-03-15 21:11:55 -07:00

299 lines
9.9 KiB
C++

/*++
Copyright (c) 2017 Microsoft Corporation
Module Name:
sat_local_search.h
Abstract:
Local search module for cardinality clauses.
Author:
Sixue Liu 2017-2-21
Notes:
--*/
#ifndef _SAT_LOCAL_SEARCH_H_
#define _SAT_LOCAL_SEARCH_H_
#include "util/vector.h"
#include "sat/sat_types.h"
#include "sat/sat_config.h"
#include "util/rlimit.h"
namespace sat {
class parallel;
class local_search_config {
unsigned m_random_seed;
int m_best_known_value;
local_search_mode m_mode;
bool m_phase_sticky;
public:
local_search_config() {
m_random_seed = 0;
m_best_known_value = INT_MAX;
m_mode = local_search_mode::wsat;
m_phase_sticky = false;
}
unsigned random_seed() const { return m_random_seed; }
unsigned best_known_value() const { return m_best_known_value; }
local_search_mode mode() const { return m_mode; }
bool phase_sticky() const { return m_phase_sticky; }
void set_random_seed(unsigned s) { m_random_seed = s; }
void set_best_known_value(unsigned v) { m_best_known_value = v; }
void set_config(config const& cfg) {
m_mode = cfg.m_local_search_mode;
m_random_seed = cfg.m_random_seed;
m_phase_sticky = cfg.m_phase_sticky;
}
};
class local_search {
struct pbcoeff {
unsigned m_constraint_id;
unsigned m_coeff;
pbcoeff(unsigned id, unsigned coeff):
m_constraint_id(id), m_coeff(coeff) {}
};
typedef svector<bool> bool_vector;
typedef svector<pbcoeff> coeff_vector;
// data structure for a term in objective function
struct ob_term {
bool_var var_id; // variable id, begin with 1
int coefficient; // non-zero integer
ob_term(bool_var v, int c): var_id(v), coefficient(c) {}
};
struct var_info {
bool m_value; // current solution
unsigned m_bias; // bias for current solution in percentage.
// if bias is 0, then value is always false, if 100, then always true
bool m_unit; // is this a unit literal
bool m_conf_change; // whether its configure changes since its last flip
bool m_in_goodvar_stack;
int m_score;
int m_slack_score;
int m_time_stamp; // the flip time stamp
int m_cscc; // how many times its constraint state configure changes since its last flip
bool_var_vector m_neighbors; // neighborhood variables
coeff_vector m_watch[2];
literal_vector m_bin[2];
var_info():
m_value(true),
m_bias(50),
m_unit(false),
m_conf_change(true),
m_in_goodvar_stack(false),
m_score(0),
m_slack_score(0),
m_cscc(0)
{}
};
struct constraint {
unsigned m_id;
unsigned m_k;
int m_slack;
unsigned m_size;
literal_vector m_literals;
constraint(unsigned k, unsigned id) : m_id(id), m_k(k), m_slack(0), m_size(0) {}
void push(literal l) { m_literals.push_back(l); ++m_size; }
unsigned size() const { return m_size; }
literal const& operator[](unsigned idx) const { return m_literals[idx]; }
literal const* begin() const { return m_literals.begin(); }
literal const* end() const { return m_literals.end(); }
};
local_search_config m_config;
// objective function: maximize
svector<ob_term> ob_constraint; // the objective function *constraint*, sorted in decending order
// information about the variable
int_vector coefficient_in_ob_constraint; // var! initialized to be 0
vector<var_info> m_vars;
svector<bool_var> m_units;
inline int score(bool_var v) const { return m_vars[v].m_score; }
inline void inc_score(bool_var v) { m_vars[v].m_score++; }
inline void dec_score(bool_var v) { m_vars[v].m_score--; }
inline int slack_score(bool_var v) const { return m_vars[v].m_slack_score; }
inline void inc_slack_score(bool_var v) { m_vars[v].m_slack_score++; }
inline void dec_slack_score(bool_var v) { m_vars[v].m_slack_score--; }
inline bool already_in_goodvar_stack(bool_var v) const { return m_vars[v].m_in_goodvar_stack; }
inline bool conf_change(bool_var v) const { return m_vars[v].m_conf_change; }
inline int time_stamp(bool_var v) const { return m_vars[v].m_time_stamp; }
inline int cscc(bool_var v) const { return m_vars[v].m_cscc; }
inline void inc_cscc(bool_var v) { m_vars[v].m_cscc++; }
inline bool cur_solution(bool_var v) const { return m_vars[v].m_value; }
inline void set_best_unsat();
/* TBD: other scores */
vector<constraint> m_constraints;
literal_vector m_assumptions;
literal_vector m_prop_queue;
unsigned m_num_non_binary_clauses;
bool m_is_pb;
inline bool is_pos(literal t) const { return !t.sign(); }
inline bool is_true(bool_var v) const { return cur_solution(v); }
inline bool is_true(literal l) const { return cur_solution(l.var()) != l.sign(); }
inline bool is_false(literal l) const { return cur_solution(l.var()) == l.sign(); }
inline bool is_unit(bool_var v) const { return m_vars[v].m_unit; }
inline bool is_unit(literal l) const { return m_vars[l.var()].m_unit; }
unsigned num_constraints() const { return m_constraints.size(); } // constraint index from 1 to num_constraint
unsigned constraint_slack(unsigned ci) const { return m_constraints[ci].m_slack; }
// unsat constraint stack
bool m_is_unsat;
unsigned_vector m_unsat_stack; // store all the unsat constraits
unsigned_vector m_index_in_unsat_stack; // which position is a contraint in the unsat_stack
// configuration changed decreasing variables (score>0 and conf_change==true)
bool_var_vector m_goodvar_stack;
// information about solution
unsigned m_best_unsat;
double m_best_unsat_rate;
double m_last_best_unsat_rate;
int m_objective_value; // the objective function value corresponds to the current solution
bool_vector m_best_solution; // !var: the best solution so far
int m_best_objective_value = -1; // the objective value corresponds to the best solution so far
// for non-known instance, set as maximal
int m_best_known_value = INT_MAX; // best known value for this instance
unsigned m_max_steps = (1 << 30);
// dynamic noise
double m_noise = 9800; // normalized by 10000
double m_noise_delta = 0.05;
reslimit m_limit;
random_gen m_rand;
parallel* m_par;
model m_model;
void init();
void reinit();
void reinit_orig();
void init_cur_solution();
void init_slack();
void init_scores();
void init_goodvars();
bool_var pick_var_gsat();
void flip_gsat(bool_var v);
void pick_flip_walksat();
void flip_walksat(bool_var v);
bool propagate(literal lit);
void add_propagation(literal lit);
void walksat();
void gsat();
void unsat(unsigned c);
void sat(unsigned c);
bool tie_breaker_sat(bool_var v1, bool_var v2);
bool tie_breaker_ccd(bool_var v1, bool_var v2);
void set_parameters();
void calculate_and_update_ob();
bool all_objectives_are_met() const;
void verify_solution() const;
void verify_unsat_stack() const;
void verify_constraint(constraint const& c) const;
unsigned constraint_value(constraint const& c) const;
unsigned constraint_coeff(constraint const& c, literal l) const;
void print_info(std::ostream& out);
void extract_model();
bool check_goodvar();
void add_clause(unsigned sz, literal const* c);
void add_unit(literal lit);
std::ostream& display(std::ostream& out) const;
std::ostream& display(std::ostream& out, constraint const& c) const;
std::ostream& display(std::ostream& out, unsigned v, var_info const& vi) const;
public:
local_search();
reslimit& rlimit() { return m_limit; }
~local_search();
void add_soft(bool_var v, int weight);
void add_cardinality(unsigned sz, literal const* c, unsigned k);
void add_pb(unsigned sz, literal const* c, unsigned const* coeffs, unsigned k);
lbool check();
lbool check(unsigned sz, literal const* assumptions, parallel* p = 0);
local_search_config& config() { return m_config; }
unsigned num_vars() const { return m_vars.size() - 1; } // var index from 1 to num_vars
unsigned num_non_binary_clauses() const { return m_num_non_binary_clauses; }
void import(solver& s, bool init);
void set_phase(bool_var v, lbool f);
bool get_phase(bool_var v) const { return is_true(v); }
model& get_model() { return m_model; }
};
}
#endif