mirrors/z3
3
0
Fork 0
mirror of https://github.com/Z3Prover/z3 synced 2025-04-08 10:25:18 +00:00
Code Activity

working on convex lemma gneralization

Browse source 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
This commit is contained in:
Nikolaj Bjorner 2013-08-27 21:11:45 -07:00
parent 2d6b3fa284
commit e5541bad17
2 changed files with 80 additions and 30 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

108
src/muz_qe/pdr_generalizers.cpp
View file

@ -154,8 +154,6 @@ namespace pdr {
m_sigma(m),
m_trail(m),
m_is_closure(is_closure) {
m_sigma.push_back(m.mk_fresh_const("sigma", a.mk_real()));
m_sigma.push_back(m.mk_fresh_const("sigma", a.mk_real()));
}
void core_convex_hull_generalizer::operator()(model_node& n, expr_ref_vector const& core, bool uses_level, cores& new_cores) {
@ -182,29 +180,18 @@ namespace pdr {
//
void core_convex_hull_generalizer::method1(model_node& n, expr_ref_vector const& core, bool uses_level, cores& new_cores) {
manager& pm = n.pt().get_pdr_manager();
expr_ref_vector conv1(m), conv2(m), core1(m), core2(m), eqs(m);
expr_ref_vector conv1(m), conv2(m), core1(m), core2(m);
unsigned orig_size = new_cores.size();
if (core.empty()) {
new_cores.push_back(std::make_pair(core, uses_level));
return;
}
add_variables(n, 2, eqs);
add_variables(n, 2, conv1);
if (!mk_convex(core, 0, conv1)) {
new_cores.push_back(std::make_pair(core, uses_level));
IF_VERBOSE(0, verbose_stream() << "Non-convex: " << mk_pp(pm.mk_and(core), m) << "\n";);
return;
}
conv1.append(eqs);
if (m_is_closure) {
conv1.push_back(a.mk_ge(m_sigma[0].get(), a.mk_numeral(rational(0), a.mk_real())));
conv1.push_back(a.mk_ge(m_sigma[1].get(), a.mk_numeral(rational(0), a.mk_real())));
}
else {
// is interior:
conv1.push_back(a.mk_gt(m_sigma[0].get(), a.mk_numeral(rational(0), a.mk_real())));
conv1.push_back(a.mk_gt(m_sigma[1].get(), a.mk_numeral(rational(0), a.mk_real())));
}
conv1.push_back(m.mk_eq(a.mk_numeral(rational(1), a.mk_real()), a.mk_add(m_sigma[0].get(), m_sigma[1].get())));
expr_ref fml = n.pt().get_formulas(n.level(), false);
expr_ref_vector fmls(m);
datalog::flatten_and(fml, fmls);
@ -281,10 +268,6 @@ namespace pdr {
m_trail.push_back(goal);
m_models.insert(goal, new_model);
}
conv1.push_back(a.mk_gt(m_sigma[0].get(), a.mk_numeral(rational(0), a.mk_real())));
conv1.push_back(a.mk_gt(m_sigma[1].get(), a.mk_numeral(rational(0), a.mk_real())));
conv1.push_back(m.mk_eq(a.mk_numeral(rational(1), a.mk_real()), a.mk_add(m_sigma[0].get(), m_sigma[1].get())));
obj_map<expr, expr*>::iterator it = m_models.begin(), end = m_models.end();
for (; it != end; ++it) {
if (it->m_key == goal) {
@ -342,21 +325,76 @@ namespace pdr {
});
// now create the convex closure of the consequences:
expr_ref_vector conv(m);
expr_ref tmp(m), zero(m);
expr_ref_vector conv(m), es(m), enabled(m);
zero = a.mk_numeral(rational(0), a.mk_real());
add_variables(n, consequences.size(), conv);
for (unsigned i = 0; i < consequences.size(); ++i) {
if (m_sigma.size() == i) {
m_sigma.push_back(m.mk_fresh_const("sigma", a.mk_real()));
es.reset();
tmp = m.mk_not(consequences[i].get());
datalog::flatten_and(tmp, es);
if (!mk_convex(es, i, conv)) {
IF_VERBOSE(0, verbose_stream() << "Failed to create convex closure\n";);
return;
}
conv.push_back(a.mk_ge(m_sigma[i].get(), a.mk_numeral(rational(0), a.mk_real())));
;; // mk_convex
es.reset();
//
// enabled[i] = not (sigma_i = 0 and z_i1 = 0 and .. and z_im = 0)
//
es.push_back(m.mk_eq(m_sigma[i].get(), zero));
for (unsigned j = 0; j < n.pt().sig_size(); ++j) {
func_decl* fn0 = n.pt().sig(j);
func_decl* fn1 = pm.o2n(fn0, 0);
expr* var;
VERIFY (m_vars[i].find(fn1, var));
es.push_back(m.mk_eq(var, zero));
}
enabled.push_back(m.mk_not(m.mk_and(es.size(), es.c_ptr())));
}
// the convex closure was created of all consequences.
// now determine a subset of enabled constraints.
smt::kernel ctx(m, m_ctx.get_fparams(), m_ctx.get_params().p);
for (unsigned i = 0; i < conv.size(); ++i) {
ctx.assert_expr(conv[i].get());
}
for (unsigned i = 0; i < core.size(); ++i) {
ctx.assert_expr(core[i]);
}
vector<unsigned_vector> transversal;
while (l_true == ctx.check()) {
model_ref md;
ctx.get_model(md);
expr_ref_vector lits(m);
unsigned_vector pos;
for (unsigned i = 0; i < consequences.size(); ++i) {
if (md->eval(enabled[i].get(), tmp, false)) {
if (m.is_true(tmp)) {
lits.push_back(tmp);
pos.push_back(i);
}
}
}
transversal.push_back(pos);
SASSERT(!lits.empty());
tmp = m.mk_not(m.mk_and(lits.size(), lits.c_ptr()));
TRACE("pdr", tout << "add block: " << mk_pp(tmp, m) << "\n";);
ctx.assert_expr(tmp);
}
//
// we could no longer satisfy core using a partition.
//
}
void core_convex_hull_generalizer::add_variables(model_node& n, unsigned num_vars, expr_ref_vector& eqs) {
void core_convex_hull_generalizer::add_variables(model_node& n, unsigned num_vars, expr_ref_vector& fmls) {
manager& pm = n.pt().get_pdr_manager();
if (m_vars.size() < num_vars) {
m_vars.resize(num_vars);
}
while (m_vars.size() < num_vars) {
m_vars.resize(m_vars.size()+1);
m_sigma.push_back(m.mk_fresh_const("sigma", a.mk_real()));
}
if (!m_vars[0].contains(n.pt().head())) {
expr_ref var(m);
m_vars[0].insert(n.pt().head(), 0);
@ -384,8 +422,20 @@ namespace pdr {
VERIFY (m_vars[j].find(fn1, var));
vars.push_back(var);
}
eqs.push_back(m.mk_eq(m.mk_const(fn1), a.mk_add(num_vars, vars.c_ptr())));
fmls.push_back(m.mk_eq(m.mk_const(fn1), a.mk_add(num_vars, vars.c_ptr())));
}
if (m_is_closure) {
for (unsigned i = 0; i < num_vars; ++i) {
fmls.push_back(a.mk_ge(m_sigma[i].get(), a.mk_numeral(rational(0), a.mk_real())));
}
}
else {
// is interior:
for (unsigned i = 0; i < num_vars; ++i) {
fmls.push_back(a.mk_gt(m_sigma[i].get(), a.mk_numeral(rational(0), a.mk_real())));
}
}
fmls.push_back(m.mk_eq(a.mk_numeral(rational(1), a.mk_real()), a.mk_add(num_vars, m_sigma.c_ptr())));
}
expr_ref core_convex_hull_generalizer::mk_closure(expr* e) {

2
src/muz_qe/pdr_generalizers.h
View file

@ -90,7 +90,7 @@ namespace pdr {
void method1(model_node& n, expr_ref_vector const& core, bool uses_level, cores& new_cores);
void method2(model_node& n, expr_ref_vector& core, bool& uses_level);
void method3(model_node& n, expr_ref_vector const& core, bool uses_level, cores& new_cores);
void add_variables(model_node& n, unsigned num_vars, expr_ref_vector& eqs);
void add_variables(model_node& n, unsigned num_vars, expr_ref_vector& fmls);
public:
core_convex_hull_generalizer(context& ctx, bool is_closure);
virtual ~core_convex_hull_generalizer() {}