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
|
#define IL_STD // required by CPLEX when using STL classes.
#include <assert.h>
#include <ilcplex/cplex.h>
#include "cpu_time.h"
#include "linprog/cplex.h"
class CPXSolution : public Solution
{
private:
CPXENVptr env;
CPXLPptr lp;
const LinearProgram &linprog;
const unsigned int num_cols;
const unsigned int num_rows;
unsigned int num_coeffs;
double *values;
bool solved;
void solve_model(double var_lb, double var_ub);
bool setup_objective(double lb, double ub);
bool add_rows();
bool load_coeffs();
public:
CPXSolution(const LinearProgram &lp, unsigned int max_num_vars,
double var_lb = 0.0, double var_ub = 1.0);
~CPXSolution();
double get_value(unsigned int var) const
{
return values[var];
}
bool is_solved() const
{
return solved;
}
};
CPXSolution::CPXSolution(const LinearProgram& lp, unsigned int max_num_vars,
double var_lb, double var_ub)
: env(0),
lp(0),
linprog(lp),
num_cols(max_num_vars),
num_rows(lp.get_equalities().size() +
lp.get_inequalities().size()),
num_coeffs(0),
values(0),
solved(false)
{
if (num_cols > 0)
{
values = new double[num_cols];
solve_model(var_lb, var_ub);
} else
// Trivial case: no variables.
solved = true;
}
void CPXSolution::solve_model(double var_lb, double var_ub)
{
int err;
#if DEBUG_LP_OVERHEADS >= 3
static DEFINE_CPU_CLOCK(model_costs);
static DEFINE_CPU_CLOCK(solver_costs);
static DEFINE_CPU_CLOCK(extract_costs);
model_costs.start();
#endif
env = CPXopenCPLEX(&err);
if (!env)
return;
lp = CPXcreateprob(env, &err, "blocking");
if (!lp)
return;
if (!setup_objective(var_lb, var_ub) ||
!add_rows() ||
!load_coeffs())
return;
#if DEBUG_LP_OVERHEADS >= 3
model_costs.stop();
solver_costs.start();
#endif
err = CPXlpopt(env, lp);
if (err != 0)
return;
#if DEBUG_LP_OVERHEADS >= 3
solver_costs.stop();
extract_costs.start();
#endif
err = CPXsolution(env, lp, NULL, NULL, values, NULL, NULL, NULL);
solved = err == 0;
#if DEBUG_LP_OVERHEADS >= 3
extract_costs.stop();
std::cout << model_costs << std::endl
<< solver_costs << std::endl
<< extract_costs << std::endl;
#endif
}
CPXSolution::~CPXSolution()
{
int status;
if (lp) {
status = CPXfreeprob(env, &lp);
assert(status == 0);
}
if (env) {
status = CPXcloseCPLEX(&env);
assert(status == 0);
}
delete [] values;
}
bool CPXSolution::setup_objective(double lb, double ub)
{
const LinearExpression *obj = linprog.get_objective();
int err;
double *all = new double[num_cols * 3];
double *vals = all;
double *lbs = all + num_cols;
double *ubs = all + 2 * num_cols;
assert(obj->get_terms().size() <= num_cols);
for (unsigned int i = 0; i < num_cols; i++)
{
vals[i] = 0;
lbs[i] = lb;
ubs[i] = ub;
}
foreach(obj->get_terms(), term)
vals[term->second] = term->first;
CPXchgobjsen(env, lp, CPX_MAX);
err = CPXnewcols(env, lp, num_cols, vals, lbs, ubs, NULL, NULL);
delete [] all;
return err == 0;
}
bool CPXSolution::add_rows()
{
double *bounds = new double[num_rows];
char *senses = new char[num_rows];
int err;
unsigned int r = 0;
foreach(linprog.get_equalities(), equ)
{
bounds[r] = equ->second;
senses[r] = 'E'; // equality constraint
num_coeffs += equ->first->get_terms().size();
r++;
}
foreach(linprog.get_inequalities(), inequ)
{
bounds[r] = inequ->second;
senses[r] = 'L'; // less-than-or-equal constraint
num_coeffs += inequ->first->get_terms().size();
r++;
}
err = CPXnewrows(env, lp, num_rows, bounds, senses, NULL, NULL);
delete [] bounds;
delete [] senses;
return err == 0;
}
bool CPXSolution::load_coeffs()
{
unsigned int r = 0;
int err;
foreach(linprog.get_equalities(), equ)
{
foreach(equ->first->get_terms(), term)
{
err = CPXchgcoef(env, lp, r, term->second, term->first);
if (err != 0)
return false;
}
r++;
}
foreach(linprog.get_inequalities(), inequ)
{
foreach(inequ->first->get_terms(), term)
{
err = CPXchgcoef(env, lp, r, term->second, term->first);
if (err != 0)
return false;
}
r++;
}
return true;
}
Solution *cpx_solve(const LinearProgram& lp, unsigned int max_num_vars)
{
CPXSolution *sol = new CPXSolution(lp, max_num_vars);
if (sol->is_solved())
return sol;
else
{
delete sol;
return NULL;
}
}
|
