Scippy

SCIP

Solving Constraint Integer Programs

heur_feaspump.c
Go to the documentation of this file.
1 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2 /* */
3 /* This file is part of the program and library */
4 /* SCIP --- Solving Constraint Integer Programs */
5 /* */
6 /* Copyright (c) 2002-2023 Zuse Institute Berlin (ZIB) */
7 /* */
8 /* Licensed under the Apache License, Version 2.0 (the "License"); */
9 /* you may not use this file except in compliance with the License. */
10 /* You may obtain a copy of the License at */
11 /* */
12 /* http://www.apache.org/licenses/LICENSE-2.0 */
13 /* */
14 /* Unless required by applicable law or agreed to in writing, software */
15 /* distributed under the License is distributed on an "AS IS" BASIS, */
16 /* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. */
17 /* See the License for the specific language governing permissions and */
18 /* limitations under the License. */
19 /* */
20 /* You should have received a copy of the Apache-2.0 license */
21 /* along with SCIP; see the file LICENSE. If not visit scipopt.org. */
22 /* */
23 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
24 
25 /**@file heur_feaspump.c
26  * @ingroup DEFPLUGINS_HEUR
27  * @brief Objective Feasibility Pump 2.0
28  * @author Timo Berthold
29  * @author Domenico Salvagnin
30  */
31 
32 /*---+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8----+----9----+----0----+----1----+----2*/
33 
34 #include "blockmemshell/memory.h"
35 #include "scip/cons_linear.h"
36 #include "scip/heur_feaspump.h"
37 #include "scip/pub_heur.h"
38 #include "scip/pub_message.h"
39 #include "scip/pub_misc.h"
40 #include "scip/pub_misc_sort.h"
41 #include "scip/pub_var.h"
42 #include "scip/scip_branch.h"
43 #include "scip/scip_cons.h"
44 #include "scip/scip_copy.h"
45 #include "scip/scip_general.h"
46 #include "scip/scip_heur.h"
47 #include "scip/scip_lp.h"
48 #include "scip/scip_mem.h"
49 #include "scip/scip_message.h"
50 #include "scip/scip_nodesel.h"
51 #include "scip/scip_numerics.h"
52 #include "scip/scip_param.h"
53 #include "scip/scip_pricer.h"
54 #include "scip/scip_prob.h"
55 #include "scip/scip_probing.h"
56 #include "scip/scip_randnumgen.h"
57 #include "scip/scip_sol.h"
58 #include "scip/scip_solve.h"
59 #include "scip/scip_solvingstats.h"
60 #include "scip/scip_tree.h"
61 #include "scip/scip_var.h"
62 #include <string.h>
63 
64 #define HEUR_NAME "feaspump"
65 #define HEUR_DESC "objective feasibility pump 2.0"
66 #define HEUR_DISPCHAR SCIP_HEURDISPCHAR_OBJDIVING
67 #define HEUR_PRIORITY -1000000
68 #define HEUR_FREQ 20
69 #define HEUR_FREQOFS 0
70 #define HEUR_MAXDEPTH -1
71 #define HEUR_TIMING SCIP_HEURTIMING_AFTERLPPLUNGE
72 #define HEUR_USESSUBSCIP FALSE /**< does the heuristic use a secondary SCIP instance? */
73 
74 #define DEFAULT_MAXLPITERQUOT 0.01 /**< maximal fraction of diving LP iterations compared to node LP iterations */
75 #define DEFAULT_MAXLPITEROFS 1000 /**< additional number of allowed LP iterations */
76 #define DEFAULT_MAXSOLS 10 /**< total number of feasible solutions found up to which heuristic is called
77  * (-1: no limit) */
78 #define DEFAULT_MAXLOOPS 10000 /**< maximal number of pumping rounds (-1: no limit) */
79 #define DEFAULT_MAXSTALLLOOPS 10 /**< maximal number of pumping rounds without fractionality improvement (-1: no limit) */
80 #define DEFAULT_MINFLIPS 10 /**< minimum number of random variables to flip, if a 1-cycle is encountered */
81 #define DEFAULT_CYCLELENGTH 3 /**< maximum length of cycles to be checked explicitly in each round */
82 #define DEFAULT_PERTURBFREQ 100 /**< number of iterations until a random perturbation is forced */
83 #define DEFAULT_OBJFACTOR 0.1 /**< factor by which the regard of the objective is decreased in each round,
84  * 1.0 for dynamic, depending on solutions already found */
85 #define DEFAULT_ALPHA 1.0 /**< initial weight of the objective function in the convex combination */
86 #define DEFAULT_ALPHADIFF 1.0 /**< threshold difference for the convex parameter to perform perturbation */
87 #define DEFAULT_BEFORECUTS TRUE /**< should the feasibility pump be called at root node before cut separation? */
88 #define DEFAULT_USEFP20 FALSE /**< should an iterative round-and-propagate scheme be used to find the integral points? */
89 #define DEFAULT_PERTSOLFOUND TRUE /**< should a random perturbation be performed if a feasible solution was found? */
90 #define DEFAULT_STAGE3 FALSE /**< should we solve a local branching sub-MIP if no solution could be found? */
91 #define DEFAULT_NEIGHBORHOODSIZE 18 /**< radius of the neighborhood to be searched in stage 3 */
92 #define DEFAULT_COPYCUTS TRUE /**< should all active cuts from the cutpool of the original SCIP be copied to
93  * constraints of the subscip
94  */
95 
96 #define MINLPITER 5000 /**< minimal number of LP iterations allowed in each LP solving call */
97 
98 #define DEFAULT_RANDSEED 13 /**< initial random seed */
99 
100 /** primal heuristic data */
101 struct SCIP_HeurData
102 {
103  SCIP_SOL* sol; /**< working solution */
104  SCIP_SOL* roundedsol; /**< rounded solution */
105  SCIP_Longint nlpiterations; /**< number of LP iterations used in this heuristic */
106  SCIP_Real maxlpiterquot; /**< maximal fraction of diving LP iterations compared to node LP iterations */
107  SCIP_Real objfactor; /**< factor by which the regard of the objective is decreased in each round,
108  * 1.0 for dynamic, depending on solutions already found */
109  SCIP_Real alpha; /**< initial weight of the objective function in the convex combination */
110  SCIP_Real alphadiff; /**< threshold difference for the convex parameter to perform perturbation */
111 
112  int maxlpiterofs; /**< additional number of allowed LP iterations */
113  int maxsols; /**< total number of feasible solutions found up to which heuristic is called
114  * (-1: no limit) */
115  int maxloops; /**< maximum number of loops (-1: no limit) */
116  int maxstallloops; /**< maximal number of pumping rounds without fractionality improvement (-1: no limit) */
117  int minflips; /**< minimum number of random variables to flip, if a 1-cycle is encountered */
118  int cyclelength; /**< maximum length of cycles to be checked explicitly in each round */
119  int perturbfreq; /**< number of iterations until a random perturbation is forced */
120  int nsuccess; /**< number of runs that produced at least one feasible solution */
121  int neighborhoodsize; /**< radius of the neighborhood to be searched in stage 3 */
122 
123  SCIP_RANDNUMGEN* randnumgen; /**< random number generator */
124  SCIP_Bool beforecuts; /**< should the feasibility pump be called at root node before cut separation? */
125  SCIP_Bool usefp20; /**< should an iterative round-and-propagate scheme be used to find the integral points? */
126  SCIP_Bool pertsolfound; /**< should a random perturbation be performed if a feasible solution was found? */
127  SCIP_Bool stage3; /**< should we solve a local branching sub-MIP if no solution could be found? */
128  SCIP_Bool copycuts; /**< should all active cuts from cutpool be copied to constraints in
129  * subproblem?
130  */
131 };
132 
133 /* copies SCIP to probing SCIP and creates variable hashmap */
134 static
136  SCIP* scip, /**< SCIP data structure */
137  SCIP** probingscip, /**< sub-SCIP data structure */
138  SCIP_HASHMAP** varmapfw, /**< mapping of SCIP variables to sub-SCIP variables */
139  SCIP_Bool copycuts, /**< should all active cuts from cutpool of scip copied to constraints in subscip */
140  SCIP_Bool* success /**< was copying successful? */
141  )
142 {
143  /* check if we are already at the maximal tree depth */
144  if( SCIP_MAXTREEDEPTH <= SCIPgetDepth(scip) )
145  {
146  *success = FALSE;
147  return SCIP_OKAY;
148  }
149 
150  /* initializing the subproblem */
151  SCIP_CALL( SCIPcreate(probingscip) );
152 
153  /* create the variable mapping hash map */
154  SCIP_CALL( SCIPhashmapCreate(varmapfw, SCIPblkmem(*probingscip), SCIPgetNVars(scip)) );
155  *success = FALSE;
156 
157  /* copy SCIP instance */
158  SCIP_CALL( SCIPcopyConsCompression(scip, *probingscip, *varmapfw, NULL, "feaspump", NULL, NULL, 0, FALSE, FALSE,
159  FALSE, TRUE, success) );
160 
161  if( copycuts )
162  {
163  /* copies all active cuts from cutpool of sourcescip to linear constraints in targetscip */
164  SCIP_CALL( SCIPcopyCuts(scip, *probingscip, *varmapfw, NULL, FALSE, NULL) );
165  }
166 
167  return SCIP_OKAY;
168 }
169 
170 /** set appropriate parameters for probing SCIP in FP2 */
171 static
173  SCIP* scip, /**< SCIP data structure */
174  SCIP* probingscip /**< sub-SCIP data structure */
175  )
176 {
177  if( SCIPisParamFixed(probingscip, "heuristics/" HEUR_NAME "/freq") )
178  {
179  SCIPwarningMessage(scip, "unfixing parameter heuristics/" HEUR_NAME "/freq in probingscip of " HEUR_NAME " heuristic to avoid recursive calls\n");
180  SCIP_CALL( SCIPunfixParam(probingscip, "heuristics/" HEUR_NAME "/freq") );
181  }
182  SCIP_CALL( SCIPsetIntParam(probingscip, "heuristics/" HEUR_NAME "/freq", -1) );
183 
184  /* do not abort subproblem on CTRL-C */
185  SCIP_CALL( SCIPsetBoolParam(probingscip, "misc/catchctrlc", FALSE) );
186 
187 #ifndef SCIP_DEBUG
188  /* disable output to console */
189  SCIP_CALL( SCIPsetIntParam(probingscip, "display/verblevel", 0) );
190 #endif
191 
192  /* do not multiaggregate variables, because otherwise they have to be skipped in the fix-and-propagate loop */
193  SCIP_CALL( SCIPsetBoolParam(probingscip, "presolving/donotmultaggr", TRUE) );
194 
195  /* limit to root node solving */
196  SCIP_CALL( SCIPsetLongintParam(probingscip, "limits/nodes", 1LL) );
197 
198  /* disable LP solving and expensive techniques */
199  if( SCIPisParamFixed(probingscip, "lp/solvefreq") )
200  {
201  SCIPwarningMessage(scip, "unfixing parameter lp/solvefreq in probingscip of " HEUR_NAME " heuristic to speed up propagation\n");
202  SCIP_CALL( SCIPunfixParam(probingscip, "lp/solvefreq") );
203  }
204  SCIP_CALL( SCIPsetIntParam(probingscip, "lp/solvefreq", -1) );
205  SCIP_CALL( SCIPsetBoolParam(probingscip, "conflict/enable", FALSE) );
206  SCIP_CALL( SCIPsetBoolParam(probingscip, "constraints/disableenfops", TRUE) );
207  SCIP_CALL( SCIPsetBoolParam(probingscip, "constraints/knapsack/negatedclique", FALSE) );
210 
211  return SCIP_OKAY;
212 }
213 
214 /** set appropriate parameters for probing SCIP in Stage 3 */
215 static
217  SCIP* scip, /**< SCIP data structure */
218  SCIP* probingscip /**< sub-SCIP data structure */
219  )
220 {
221  SCIP_CALL( SCIPcopyParamSettings(scip, probingscip) );
222  /* do not abort subproblem on CTRL-C */
223  SCIP_CALL( SCIPsetBoolParam(probingscip, "misc/catchctrlc", FALSE) );
224 
225 #ifndef SCIP_DEBUG
226  /* disable output to console */
227  SCIP_CALL( SCIPsetIntParam(probingscip, "display/verblevel", 0) );
228 #endif
229  /* set limits for the subproblem */
230  SCIP_CALL( SCIPcopyLimits(scip, probingscip) );
231  SCIP_CALL( SCIPsetLongintParam(probingscip, "limits/nodes", 1000LL) );
232  SCIP_CALL( SCIPsetLongintParam(probingscip, "limits/stallnodes", 100LL) );
233 
234  /* forbid recursive call of heuristics and separators solving sub-SCIPs */
235  SCIP_CALL( SCIPsetSubscipsOff(probingscip, TRUE) );
236  if( SCIPisParamFixed(probingscip, "heuristics/" HEUR_NAME "/freq") )
237  {
238  SCIPwarningMessage(scip,"unfixing parameter heuristics/" HEUR_NAME "/freq in probingscip of " HEUR_NAME " heuristic to avoid recursive calls\n");
239  SCIP_CALL( SCIPunfixParam(probingscip, "heuristics/" HEUR_NAME "/freq") );
240  }
241  SCIP_CALL( SCIPsetIntParam(probingscip, "heuristics/feaspump/freq", -1) );
242 
243  /* disable heuristics which aim to feasibility instead of optimality */
244  if( !SCIPisParamFixed(probingscip, "heuristics/octane/freq") )
245  {
246  SCIP_CALL( SCIPsetIntParam(probingscip, "heuristics/octane/freq", -1) );
247  }
248  if( !SCIPisParamFixed(probingscip, "heuristics/objpscostdiving/freq") )
249  {
250  SCIP_CALL( SCIPsetIntParam(probingscip, "heuristics/objpscostdiving/freq", -1) );
251  }
252  if( !SCIPisParamFixed(probingscip, "heuristics/rootsoldiving/freq") )
253  {
254  SCIP_CALL( SCIPsetIntParam(probingscip, "heuristics/rootsoldiving/freq", -1) );
255  }
256 
257  /* disable cutting plane separation */
259 
260  /* disable expensive presolving */
262 
263  /* use best estimate node selection */
264  if( SCIPfindNodesel(probingscip, "estimate") != NULL && !SCIPisParamFixed(probingscip, "nodeselection/estimate/stdpriority") )
265  {
266  SCIP_CALL( SCIPsetIntParam(probingscip, "nodeselection/estimate/stdpriority", INT_MAX/4) );
267  }
268 
269  /* use inference branching */
270  if( SCIPfindBranchrule(probingscip, "inference") != NULL && !SCIPisParamFixed(probingscip, "branching/inference/priority") )
271  {
272  SCIP_CALL( SCIPsetIntParam(probingscip, "branching/inference/priority", INT_MAX/4) );
273  }
274 
275  /* disable conflict analysis */
276  if( !SCIPisParamFixed(probingscip, "conflict/enable") )
277  {
278  SCIP_CALL( SCIPsetBoolParam(probingscip, "conflict/enable", FALSE) );
279  }
280 
281  return SCIP_OKAY;
282 }
283 
284 /** checks whether a variable is one of the currently most fractional ones */
285 static
286 void insertFlipCand(
287  SCIP_VAR** mostfracvars, /**< sorted array of the currently most fractional variables */
288  SCIP_Real* mostfracvals, /**< array of their fractionality, decreasingly sorted */
289  int* nflipcands, /**< number of fractional variables already labeled to be flipped*/
290  int maxnflipcands, /**< typically randomized number of maximum amount of variables to flip */
291  SCIP_VAR* var, /**< variable to be checked */
292  SCIP_Real frac /**< fractional value of the variable */
293  )
294 {
295  int i;
296 
297  assert(mostfracvars != NULL);
298  assert(mostfracvals != NULL);
299  assert(nflipcands != NULL);
300 
301  /* instead of the fractional value use the fractionality */
302  if( frac > 0.5 )
303  frac = 1 - frac;
304 
305  /* if there are already enough candidates and the variable is less fractional, return, else reserve the last entry */
306  if( *nflipcands >= maxnflipcands )
307  {
308  if( frac <= mostfracvals[*nflipcands-1] )
309  return;
310  else
311  (*nflipcands)--;
312  }
313 
314  /* shifting var and frac through the (sorted) arrays */
315  for( i = *nflipcands; i > 0 && mostfracvals[i-1] < frac; i-- )
316  {
317  mostfracvars[i] = mostfracvars[i-1];
318  mostfracvals[i] = mostfracvals[i-1];
319  }
320  assert(0 <= i && i <= *nflipcands && *nflipcands < maxnflipcands);
321 
322  /* insert the variable and its fractionality */
323  mostfracvars[i] = var;
324  mostfracvals[i] = frac;
325 
326  /* we've found another candidate */
327  (*nflipcands)++;
328 }
329 
330 /** set solution value in rounded solution and update objective coefficient accordingly */
331 static
333  SCIP* scip, /**< SCIP data structure */
334  SCIP_HEURDATA* heurdata, /**< heuristic data structure */
335  SCIP_VAR* var, /**< variable */
336  SCIP_Real solval, /**< solution value for this variable */
337  SCIP_Real alpha, /**< weight of original objective function */
338  SCIP_Real scalingfactor /**< factor to scale the original objective function with */
339  )
340 {
341  SCIP_Real lb;
342  SCIP_Real ub;
343  SCIP_Real newobjcoeff;
344  SCIP_Real orgobjcoeff;
345 
346  assert(heurdata != NULL);
347  assert(var != NULL);
348 
349  lb = SCIPvarGetLbLocal(var);
350  ub = SCIPvarGetUbLocal(var);
351 
352  /* update rounded solution */
353  assert(SCIPisFeasLE(scip, lb, solval) && SCIPisFeasLE(scip, solval, ub));
354  assert(SCIPisIntegral(scip,solval));
355  SCIP_CALL( SCIPsetSolVal(scip, heurdata->roundedsol, var, solval) );
356 
357  /* modify objective towards rounded solution value if it is at one of the variable bounds */
358  orgobjcoeff = SCIPvarGetObj(var);
359  if( SCIPisEQ(scip, solval, lb) )
360  newobjcoeff = (1.0 - alpha)/scalingfactor + alpha * orgobjcoeff;
361  else if( SCIPisEQ(scip, solval, ub) )
362  newobjcoeff = - (1.0 - alpha)/scalingfactor + alpha * orgobjcoeff;
363  else
364  newobjcoeff = alpha * orgobjcoeff;
365 
366  SCIP_CALL( SCIPchgVarObjDive(scip, var, newobjcoeff) );
367 
368  return SCIP_OKAY;
369 }
370 
371 
372 /** flips the roundings of the most fractional variables, if a 1-cycle was found */
373 static
375  SCIP* scip, /**< SCIP data structure */
376  SCIP_HEURDATA* heurdata, /**< data of this special heuristic */
377  SCIP_VAR** mostfracvars, /**< sorted array of the currently most fractional variables */
378  int nflipcands, /**< number of variables to flip */
379  SCIP_Real alpha, /**< factor how much the original objective is regarded */
380  SCIP_Real scalingfactor /**< factor to scale the original objective function with */
381  )
382 {
383  int i;
384 
385  /* flip rounding to the opposite side */
386  for( i = 0; i < nflipcands; i++ )
387  {
388  SCIP_VAR* var;
389  SCIP_Real solval;
390  SCIP_Real roundedsolval;
391 
392  var = mostfracvars[i];
393  solval = SCIPvarGetLPSol(var);
394  roundedsolval = SCIPgetSolVal(scip, heurdata->roundedsol, var);
395 
396  assert(! SCIPisFeasIntegral(scip, solval));
397  assert(SCIPisFeasIntegral(scip, roundedsolval));
398 
399  /* flip to the opposite rounded solution value */
400  if( roundedsolval > solval )
401  solval = SCIPfeasFloor(scip, solval);
402  else
403  {
404  solval = SCIPfeasCeil(scip, solval);
405  }
406 
407  SCIPdebugMsg(scip, "1-cycle flip: variable <%s> [%g,%g] LP sol %.15g sol %.15g -> %.15g\n",
409  SCIPvarGetLPSol(var), SCIPgetSolVal(scip, heurdata->roundedsol, var), solval);
410 
411  SCIP_CALL( updateVariableRounding(scip, heurdata, var, solval, alpha, scalingfactor) );
412  }
413  return SCIP_OKAY;
414 }
415 
416 /** flips the roundings of randomly chosen fractional variables, preferring highly fractional ones,
417  * if a longer cycle was found
418  */
419 static
421  SCIP* scip, /**< SCIP data structure */
422  SCIP_HEURDATA* heurdata, /**< data of this special heuristic */
423  SCIP_VAR** vars, /**< array of all variables */
424  int nbinandintvars, /**< number of general integer and 0-1 variables */
425  SCIP_Real alpha, /**< factor how much the original objective is regarded */
426  SCIP_Real scalingfactor /**< factor to scale the original objective function with */
427  )
428 {
429  int i;
430 
431  /* flip variables randomized biased on their fractionality */
432  for( i = 0; i < nbinandintvars; i++ )
433  {
434  SCIP_VAR* var;
435  SCIP_Real solval;
436  SCIP_Real frac;
437  SCIP_Real flipprob;
438  SCIP_Real roundedsolval;
439 
440  var = vars[i];
441  solval = SCIPvarGetLPSol(var);
442 
443  /* skip variables with integral solution values */
444  if( SCIPisFeasIntegral(scip, solval) )
445  continue;
446 
447  frac = SCIPfeasFrac(scip, solval);
448  flipprob = SCIPrandomGetReal(heurdata->randnumgen, -0.3, 0.7);
449 
450  /* flip, iff the sum of the randomized number and the fractionality is big enough */
451  if( MIN(frac, 1.0 - frac) + MAX(flipprob, 0.0) > 0.5 )
452  {
453  roundedsolval = SCIPgetSolVal(scip, heurdata->roundedsol, var);
454  assert(SCIPisFeasIntegral(scip, roundedsolval));
455 
456  /* flip the solution to the opposite side */
457  if( roundedsolval > solval )
458  solval = SCIPfloor(scip, solval);
459  else
460  solval = SCIPceil(scip, solval);
461 
462  /* update rounded solution value and objective coefficient */
463  SCIP_CALL( updateVariableRounding(scip, heurdata, var, solval, alpha, scalingfactor) );
464  }
465  }
466 
467  return SCIP_OKAY;
468 }
469 
470 /** create the extra constraint of local branching and add it to subscip */
471 static
473  SCIP* scip, /**< SCIP data structure of the original problem */
474  SCIP* probingscip, /**< SCIP data structure of the subproblem */
475  SCIP_HASHMAP* varmapfw, /**< mapping of SCIP variables to sub-SCIP variables */
476  SCIP_SOL* bestsol, /**< SCIP solution */
477  SCIP_Real neighborhoodsize /**< rhs for LB constraint */
478  )
479 {
480  SCIP_CONS* cons; /* local branching constraint to create */
481  SCIP_VAR** consvars;
482  SCIP_VAR** vars;
483 
484  int nbinvars;
485  int nconsvars;
486  int i;
487  SCIP_Real lhs;
488  SCIP_Real rhs;
489  SCIP_Real* consvals;
490  char consname[SCIP_MAXSTRLEN];
491 
492  (void) SCIPsnprintf(consname, SCIP_MAXSTRLEN, "%s_localbranchcons", SCIPgetProbName(scip));
493 
494  /* get vars data */
495  SCIP_CALL( SCIPgetVarsData(scip, &vars, NULL, &nbinvars, NULL, NULL, NULL) );
496  /* memory allocation */
497  SCIP_CALL( SCIPallocBufferArray(scip, &consvars, nbinvars) );
498  SCIP_CALL( SCIPallocBufferArray(scip, &consvals, nbinvars) );
499  nconsvars = 0;
500 
501  /* set initial left and right hand sides of local branching constraint */
502  lhs = 0.0;
503  rhs = neighborhoodsize;
504 
505  /* create the distance (to incumbent) function of the binary variables */
506  for( i = 0; i < nbinvars; i++ )
507  {
508  SCIP_Real solval;
509 
510  solval = SCIPgetSolVal(scip, bestsol, vars[i]);
511  assert( SCIPisFeasIntegral(scip, solval) );
512 
513  /* is variable i part of the binary support of closest sol? */
514  if( SCIPisFeasEQ(scip,solval,1.0) )
515  {
516  consvals[nconsvars] = -1.0;
517  rhs -= 1.0;
518  lhs -= 1.0;
519  }
520  else
521  consvals[nconsvars] = 1.0;
522  consvars[nconsvars] = (SCIP_VAR*) SCIPhashmapGetImage(varmapfw, vars[i]);
523  if( consvars[nconsvars] == NULL )
524  continue;
525  SCIP_CALL( SCIPchgVarObj(probingscip, consvars[nconsvars], consvals[nconsvars]) );
526  assert( SCIPvarGetType(consvars[nconsvars]) == SCIP_VARTYPE_BINARY );
527  ++nconsvars;
528  }
529 
530  /* creates localbranching constraint and adds it to subscip */
531  SCIP_CALL( SCIPcreateConsLinear(probingscip, &cons, consname, nconsvars, consvars, consvals,
532  lhs, rhs, FALSE, FALSE, TRUE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE, FALSE) );
533  SCIP_CALL( SCIPaddCons(probingscip, cons) );
534  SCIP_CALL( SCIPreleaseCons(probingscip, &cons) );
535 
536  /* free local memory */
537  SCIPfreeBufferArray(scip, &consvals);
538  SCIPfreeBufferArray(scip, &consvars);
539 
540  return SCIP_OKAY;
541 }
542 
543 /** creates new solutions for the original problem by copying the solutions of the subproblem */
544 static
546  SCIP* scip, /**< original SCIP data structure */
547  SCIP* subscip, /**< SCIP structure of the subproblem */
548  SCIP_HASHMAP* varmapfw, /**< mapping of SCIP variables to sub-SCIP variables */
549  SCIP_HEUR* heur, /**< heuristic structure */
550  SCIP_Bool* success /**< used to store whether new solution was found or not */
551  )
552 {
553  SCIP_VAR** vars; /* the original problem's variables */
554  int nvars;
555  SCIP_VAR** subvars;
556  int i;
557 
558  assert(scip != NULL);
559  assert(subscip != NULL);
560 
561  /* get variables' data */
562  SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, NULL, NULL, NULL, NULL) );
563 
564  /* for copying a solution we need an explicit mapping */
565  SCIP_CALL( SCIPallocBufferArray(scip, &subvars, nvars) );
566  for( i = 0; i < nvars; i++ )
567  subvars[i] = (SCIP_VAR*) SCIPhashmapGetImage(varmapfw, vars[i]);
568 
569  SCIP_CALL( SCIPtranslateSubSols(scip, subscip, heur, subvars, success, NULL) );
570 
571  SCIPfreeBufferArray(scip, &subvars);
572 
573  return SCIP_OKAY;
574 }
575 
576 /** copy method for primal heuristic plugins (called when SCIP copies plugins) */
577 static
578 SCIP_DECL_HEURCOPY(heurCopyFeaspump)
579 {
580  assert(scip != NULL);
581  assert(heur != NULL);
582  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
583 
584  /* call inclusion method of primal heuristic */
586 
587  return SCIP_OKAY;
588 }
589 
590 /** destructor of primal heuristic to free user data (called when SCIP is exiting) */
591 static
592 SCIP_DECL_HEURFREE(heurFreeFeaspump)
593 {
594  SCIP_HEURDATA* heurdata;
595 
596  assert(heur != NULL);
597  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
598  assert(scip != NULL);
599 
600  /* free heuristic data */
601  heurdata = SCIPheurGetData(heur);
602  assert(heurdata != NULL);
603  SCIPfreeBlockMemory(scip, &heurdata);
604  SCIPheurSetData(heur, NULL);
605 
606  return SCIP_OKAY;
607 }
608 
609 
610 /** initialization method of primal heuristic (called after problem was transformed) */
611 static
612 SCIP_DECL_HEURINIT(heurInitFeaspump)
613 {
614  SCIP_HEURDATA* heurdata;
615 
616  assert(heur != NULL);
617  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
618 
619  /* get heuristic data */
620  heurdata = SCIPheurGetData(heur);
621  assert(heurdata != NULL);
622 
623  /* create working solution */
624  SCIP_CALL( SCIPcreateSol(scip, &heurdata->sol, heur) );
625  SCIP_CALL( SCIPcreateSol(scip, &heurdata->roundedsol, heur) );
626 
627  /* initialize data */
628  heurdata->nlpiterations = 0;
629  heurdata->nsuccess = 0;
630 
631  /* create random number generator */
632  SCIP_CALL( SCIPcreateRandom(scip, &heurdata->randnumgen,
634 
635  return SCIP_OKAY;
636 }
637 
638 /** deinitialization method of primal heuristic (called before transformed problem is freed) */
639 static
640 SCIP_DECL_HEUREXIT(heurExitFeaspump)
641 {
642  SCIP_HEURDATA* heurdata;
643 
644  assert(heur != NULL);
645  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
646 
647  /* get heuristic data */
648  heurdata = SCIPheurGetData(heur);
649  assert(heurdata != NULL);
650 
651  /* free working solution */
652  SCIP_CALL( SCIPfreeSol(scip, &heurdata->sol) );
653  SCIP_CALL( SCIPfreeSol(scip, &heurdata->roundedsol) );
654 
655  /* free random number generator */
656  SCIPfreeRandom(scip, &heurdata->randnumgen);
657 
658  return SCIP_OKAY;
659 }
660 
661 
662 /** solving process initialization method of primal heuristic (called when branch and bound process is about to begin) */
663 static
664 SCIP_DECL_HEURINITSOL(heurInitsolFeaspump)
665 {
666  SCIP_HEURDATA* heurdata;
667 
668  heurdata = SCIPheurGetData(heur);
669  assert(heurdata != NULL);
670 
671  /* if the heuristic is called at the root node, we may want to be called directly after the initial root LP solve */
672  if( heurdata->beforecuts && SCIPheurGetFreqofs(heur) == 0 )
674 
675  return SCIP_OKAY;
676 }
677 
678 
679 /** solving process deinitialization method of primal heuristic (called before branch and bound process data is freed) */
680 static
681 SCIP_DECL_HEUREXITSOL(heurExitsolFeaspump)
682 {
683  /* reset the timing mask to its default value */
686  return SCIP_OKAY;
687 }
688 
689 /** calculates an adjusted maximal number of LP iterations */
690 static
692  SCIP_Longint maxnlpiterations, /**< regular maximal number of LP iterations */
693  SCIP_Longint nsolsfound, /**< total number of solutions found so far by SCIP */
694  int nstallloops /**< current number of stalling rounds */
695  )
696 {
697  if( nstallloops <= 1 )
698  {
699  if( nsolsfound == 0 )
700  return 4*maxnlpiterations;
701  else
702  return 2*maxnlpiterations;
703  }
704  else
705  return maxnlpiterations;
706 }
707 
708 /** execution method of primal heuristic */
709 static
710 SCIP_DECL_HEUREXEC(heurExecFeaspump)
711 {
712  SCIP_HEURDATA* heurdata;
713  SCIP_SOL* tmpsol; /* only used for swapping */
714  SCIP_SOL** lastroundedsols;/* solutions of the last pumping rounds (depending on heurdata->cyclelength) */
715  SCIP_SOL* closestsol; /* rounded solution closest to the LP relaxation: used for stage3 */
716  SCIP_Real* lastalphas; /* alpha values associated to solutions in lastroundedsols */
717 
718  SCIP* probingscip; /* copied SCIP structure, used for round-and-propagate loop of feasibility pump 2.0 */
719  SCIP_HASHMAP* varmapfw; /* mapping of SCIP variables to sub-SCIP variables */
720 
721  SCIP_VAR** vars;
722  SCIP_VAR** pseudocands;
723  SCIP_VAR** tmppseudocands;
724  SCIP_VAR** mostfracvars; /* the 30 most fractional variables, needed to avoid 1-cycles */
725  SCIP_VAR* var;
726 
727  SCIP_Real* mostfracvals; /* the values of the variables above */
728  SCIP_Real oldsolval; /* one value of the last solution */
729  SCIP_Real solval; /* one value of the actual solution */
730  SCIP_Real frac; /* the fractional part of the value above */
731  SCIP_Real objfactor; /* factor by which the regard of the objective is decreased in each round, in [0,0.99] */
732  SCIP_Real alpha; /* factor how the original objective is regarded, used for convex combination of two functions */
733  SCIP_Real objnorm; /* Euclidean norm of the objective function, used for scaling */
734  SCIP_Real scalingfactor; /* factor to scale the original objective function with */
735  SCIP_Real mindistance; /* distance of the closest rounded solution from the LP relaxation: used for stage3 */
736 
737  SCIP_Longint nlpiterations; /* number of LP iterations done during one pumping round */
738  SCIP_Longint maxnlpiterations; /* maximum number of LP iterations for this heuristic */
739  SCIP_Longint nsolsfound; /* number of solutions found by this heuristic */
740  SCIP_Longint ncalls; /* number of calls of this heuristic */
741  SCIP_Longint nbestsolsfound; /* current total number of best solution updates in SCIP */
742 
743  SCIP_LPSOLSTAT lpsolstat; /* status of the LP solution */
744 
745  int nvars; /* number of variables */
746  int nbinvars; /* number of 0-1-variables */
747  int nintvars; /* number of integer variables */
748  int nfracs; /* number of fractional variables updated after each pumping round*/
749  int nflipcands; /* how many flipcands (most frac. var.) have been found */
750  int npseudocands;
751  int maxnflipcands; /* maximal number of candidates to flip in the current pumping round */
752  int nloops; /* how many pumping rounds have been made */
753  int maxflips; /* maximum number of flips, if a 1-cycle is found (depending on heurdata->minflips) */
754  int maxloops; /* maximum number of pumping rounds */
755  int nstallloops; /* number of loops without reducing the current best number of factional variables */
756  int maxstallloops; /* maximal number of allowed stalling loops */
757  int bestnfracs; /* best number of fractional variables */
758  int i;
759  int j;
760 
761  SCIP_Bool success;
762  SCIP_Bool lperror;
763  SCIP_Bool* cycles; /* are there short cycles */
764 
765  SCIP_RETCODE retcode;
766 
767  assert(heur != NULL);
768  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
769  assert(scip != NULL);
770  assert(result != NULL);
771  assert(SCIPhasCurrentNodeLP(scip));
772 
773  *result = SCIP_DELAYED;
774 
775  /* do not call heuristic of node was already detected to be infeasible */
776  if( nodeinfeasible )
777  return SCIP_OKAY;
778 
779  /* only call heuristic, if an optimal LP solution is at hand */
781  return SCIP_OKAY;
782 
783  /* only call heuristic, if the LP objective value is smaller than the cutoff bound */
784  if( SCIPisGE(scip, SCIPgetLPObjval(scip), SCIPgetCutoffbound(scip)) )
785  return SCIP_OKAY;
786 
787  /* only call heuristic, if the LP solution is basic (which allows fast resolve in diving) */
788  if( !SCIPisLPSolBasic(scip) )
789  return SCIP_OKAY;
790 
791  *result = SCIP_DIDNOTRUN;
792 
793  /* don't dive two times at the same node */
794  if( SCIPgetLastDivenode(scip) == SCIPgetNNodes(scip) && SCIPgetDepth(scip) > 0 )
795  return SCIP_OKAY;
796 
797  /* only call feaspump once at the root */
798  if( SCIPgetDepth(scip) == 0 && SCIPheurGetNCalls(heur) > 0 )
799  return SCIP_OKAY;
800 
801  /* reset the timing mask to its default value (at the root node it could be different) */
803 
804  /* only call the heuristic before cutting planes if we do not have an incumbent and no pricer exists */
805  if( heurtiming == SCIP_HEURTIMING_DURINGLPLOOP && SCIPgetNSolsFound(scip) > 0 && SCIPgetNPricers(scip) == 0 )
806  return SCIP_OKAY;
807 
808  /* get heuristic's data */
809  heurdata = SCIPheurGetData(heur);
810  assert(heurdata != NULL);
811 
812  /* only apply heuristic, if only a few solutions have been found and no pricer exists */
813  if( heurdata->maxsols >= 0 && SCIPgetNSolsFound(scip) > heurdata->maxsols && SCIPgetNPricers(scip) == 0 )
814  return SCIP_OKAY;
815 
816  /* get all variables of LP and number of fractional variables in LP solution that should be integral */
817  SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, &nbinvars, &nintvars, NULL, NULL) );
818  nfracs = SCIPgetNLPBranchCands(scip);
819  assert(0 <= nfracs && nfracs <= nbinvars + nintvars);
820  if( nfracs == 0 )
821  return SCIP_OKAY;
822 
823  /* calculate the maximal number of LP iterations until heuristic is aborted */
824  nlpiterations = SCIPgetNLPIterations(scip);
825  ncalls = SCIPheurGetNCalls(heur);
826  nsolsfound = 10*SCIPheurGetNBestSolsFound(heur) + heurdata->nsuccess;
827  maxnlpiterations = (SCIP_Longint)((1.0 + 10.0*(nsolsfound+1.0)/(ncalls+1.0)) * heurdata->maxlpiterquot * nlpiterations);
828  maxnlpiterations += heurdata->maxlpiterofs;
829 
830  /* don't try to dive, if we took too many LP iterations during diving */
831  if( heurdata->nlpiterations >= maxnlpiterations )
832  return SCIP_OKAY;
833 
834  /* at the first root call, allow more iterations if there is no feasible solution yet */
835  if( SCIPheurGetNCalls(heur) == 0 && SCIPgetNSolsFound(scip) == 0 && SCIPgetDepth(scip) == 0 )
836  maxnlpiterations += nlpiterations;
837 
838  /* allow at least a certain number of LP iterations in this dive */
839  maxnlpiterations = MAX(maxnlpiterations, heurdata->nlpiterations + MINLPITER);
840 
841  /* calculate maximal number of flips and loops */
842  maxflips = 3*heurdata->minflips;
843  maxloops = (heurdata->maxloops == -1 ? INT_MAX : heurdata->maxloops);
844  maxstallloops = (heurdata->maxstallloops == -1 ? INT_MAX : heurdata->maxstallloops);
845 
846  SCIPdebugMsg(scip, "executing feasibility pump heuristic, nlpiters=%" SCIP_LONGINT_FORMAT ", maxnlpit:%" SCIP_LONGINT_FORMAT ", maxflips:%d \n",
847  nlpiterations, maxnlpiterations, maxflips);
848 
849  *result = SCIP_DIDNOTFIND;
850 
851  probingscip = NULL;
852  varmapfw = NULL;
853 
854  if( heurdata->usefp20 )
855  {
856  SCIP_Bool valid;
857 
858  /* ignore value of valid */
859  SCIP_CALL( setupProbingSCIP(scip, &probingscip, &varmapfw, heurdata->copycuts, &valid) );
860 
861  if( probingscip != NULL )
862  {
863  SCIP_CALL( setupSCIPparamsFP2(scip, probingscip) );
864 
865  retcode = SCIPsolve(probingscip);
866 
867  /* errors in solving the subproblem should not kill the overall solving process;
868  * hence, the return code is caught and a warning is printed, only in debug mode, SCIP will stop. */
869  if( retcode != SCIP_OKAY )
870  {
871 #ifndef NDEBUG
872  SCIP_CALL( retcode );
873 #endif
874  SCIPwarningMessage(scip, "Error while solving subproblem in feaspump heuristic; sub-SCIP terminated with code <%d>\n", retcode);
875 
876  /* free hash map and copied SCIP */
877  SCIPhashmapFree(&varmapfw);
878  SCIP_CALL( SCIPfree(&probingscip) );
879  return SCIP_OKAY;
880  }
881 
882  if( SCIPgetStage(probingscip) != SCIP_STAGE_SOLVING)
883  {
884  SCIP_STATUS probingstatus = SCIPgetStatus(probingscip);
885 
886  if( probingstatus == SCIP_STATUS_OPTIMAL )
887  {
888  assert( SCIPgetNSols(probingscip) > 0 );
889  SCIP_CALL( createNewSols(scip, probingscip, varmapfw, heur, &success) );
890  if( success )
891  *result = SCIP_FOUNDSOL;
892  }
893 
894  /* free hash map and copied SCIP */
895  SCIPhashmapFree(&varmapfw);
896  SCIP_CALL( SCIPfree(&probingscip) );
897  return SCIP_OKAY;
898  }
899  SCIP_CALL( SCIPsetLongintParam(probingscip, "limits/nodes", 2LL) );
900 
901  /* set SCIP into probing mode and create root node of the probing tree */
902  SCIP_CALL( SCIPstartProbing(probingscip) );
903 
904  /* this should always be fulfilled */
905  assert(SCIP_MAXTREEDEPTH > SCIPgetDepth(probingscip));
906 
907  SCIP_CALL( SCIPnewProbingNode(probingscip) );
908 
909  SCIPdebugMsg(scip, "successfully copied SCIP instance -> feasibility pump 2.0 can be used.\n");
910  }
911  else
912  {
913  assert(varmapfw == NULL);
914 
915  SCIPdebugMsg(scip, "SCIP reached the depth limit -> skip heuristic\n");
916  return SCIP_OKAY;
917  }
918  } /*lint !e438*/
919 
920  /* memory allocation */
921  SCIP_CALL( SCIPallocBufferArray(scip, &mostfracvars, maxflips) );
922  SCIP_CALL( SCIPallocBufferArray(scip, &mostfracvals, maxflips) );
923  SCIP_CALL( SCIPallocBufferArray(scip, &lastroundedsols, heurdata->cyclelength) );
924  SCIP_CALL( SCIPallocBufferArray(scip, &lastalphas, heurdata->cyclelength) );
925  SCIP_CALL( SCIPallocBufferArray(scip, &cycles, heurdata->cyclelength) );
926 
927  for( j = 0; j < heurdata->cyclelength; j++ )
928  {
929  SCIP_CALL( SCIPcreateSol(scip, &lastroundedsols[j], heur) );
930  }
931 
932  closestsol = NULL;
933  if( heurdata->stage3 )
934  {
935  SCIP_CALL( SCIPcreateSol(scip, &closestsol, heur) );
936  }
937 
938  /* start diving */
939  SCIP_CALL( SCIPstartDive(scip) );
940 
941  /* lp was solved optimal */
942  lperror = FALSE;
943  lpsolstat = SCIP_LPSOLSTAT_OPTIMAL;
944 
945  /* pumping rounds */
946  nsolsfound = SCIPgetNBestSolsFound(scip);
947  if( heurdata->objfactor == 1.0 )
948  objfactor = MIN(1.0 - 0.1 / (SCIP_Real)(1 + nsolsfound), 0.999);
949  else
950  objfactor = heurdata->objfactor;
951 
952  /* scale distance function and original objective to the same norm */
953  objnorm = SCIPgetObjNorm(scip);
954  objnorm = MAX(objnorm, 1.0);
955  scalingfactor = SQRT((SCIP_Real)(nbinvars + nintvars)) / objnorm;
956 
957  /* data initialization */
958  alpha = heurdata->alpha;
959  nloops = 0;
960  nstallloops = 0;
961  nbestsolsfound = SCIPgetNBestSolsFound(scip);
962  bestnfracs = INT_MAX;
963  mindistance = SCIPinfinity(scip);
964 
965  /* pumping loop */
966  while( nfracs > 0
967  && heurdata->nlpiterations < adjustedMaxNLPIterations(maxnlpiterations, nsolsfound, nstallloops)
968  && nloops < maxloops && nstallloops < maxstallloops
969  && !SCIPisStopped(scip) )
970  {
971  int minimum;
972  SCIP_Real* pseudocandsfrac;
973  SCIP_Longint nlpiterationsleft;
974  int iterlimit;
975 
976  /* decrease convex combination scalar */
977  nloops++;
978  alpha *= objfactor;
979 
980  SCIPdebugMsg(scip, "feasibility pump loop %d: %d fractional variables (alpha: %.4f, stall: %d/%d)\n",
981  nloops, nfracs, alpha, nstallloops, maxstallloops);
982 
983  success = FALSE;
984 
985  SCIP_CALL( SCIPlinkLPSol(scip, heurdata->roundedsol) );
986 
987  /* randomly choose maximum number of variables to flip in current pumping round in case of a 1-cycle */
988  maxnflipcands = SCIPrandomGetInt(heurdata->randnumgen, MIN(nfracs/2+1, heurdata->minflips), MIN(nfracs, maxflips));
989  nflipcands = 0;
990 
991  /* get all unfixed integer variables */
992  SCIP_CALL( SCIPgetPseudoBranchCands(scip, &tmppseudocands, &npseudocands, NULL) );
993  SCIP_CALL( SCIPduplicateBufferArray(scip, &pseudocands, tmppseudocands, npseudocands) );
994 
995  /* get array of all fractional variables and sort it w.r.t. their fractionalities */
996  if( heurdata->usefp20 )
997  {
998  SCIP_CALL( SCIPallocBufferArray(scip, &pseudocandsfrac, npseudocands) );
999 
1000  for( i = 0; i < npseudocands; i++ )
1001  {
1002  frac = SCIPfeasFrac(scip, SCIPvarGetLPSol(pseudocands[i]));
1003  pseudocandsfrac[i] = MIN(frac, 1.0-frac); /* always a number between 0 and 0.5 */
1004  if( SCIPvarGetType(pseudocands[i]) == SCIP_VARTYPE_BINARY )
1005  pseudocandsfrac[i] -= 10.0; /* binaries always come first */
1006  }
1007  SCIPsortRealPtr(pseudocandsfrac, (void**)pseudocands, npseudocands);
1008  SCIPfreeBufferArray(scip, &pseudocandsfrac);
1009 
1010  SCIPdebugMsg(scip, "iteratively fix and propagate variables\n");
1011  }
1012 
1013  for( i = 0; i < npseudocands; i++ )
1014  {
1015  SCIP_VAR* probingvar;
1016  SCIP_Bool infeasible;
1017  SCIP_Longint ndomreds;
1018 
1019  var = pseudocands[i];
1020 
1021  /* round the LP solution */
1022  solval = SCIPvarGetLPSol(var);
1023  frac = SCIPfeasFrac(scip, solval);
1024 
1025  /* round randomly if the value is close to 0.5 */
1026  if( SCIPisEQ(scip, frac, 0.5) )
1027  {
1028  if( SCIPrandomGetReal(heurdata->randnumgen, 0.0, 1.0) <= 0.5 )
1029  solval = SCIPfloor(scip, solval);
1030  else
1031  solval = SCIPceil(scip, solval);
1032  }
1033  else
1034  solval = SCIPfloor(scip, solval + 0.5);
1035 
1036  /* ensure, that the fixing value is inside the local domains */
1037  if( heurdata->usefp20 )
1038  {
1039  SCIP_Real lbprobing;
1040  SCIP_Real ubprobing;
1041 
1042  probingvar = (SCIP_VAR*) SCIPhashmapGetImage(varmapfw, var);
1043  /* skip if variable does not exist in probingscip */
1044  if( probingvar != NULL )
1045  {
1046  lbprobing = SCIPvarGetLbLocal(probingvar);
1047  ubprobing = SCIPvarGetUbLocal(probingvar);
1048 
1049  solval = MAX(solval, lbprobing);
1050  solval = MIN(solval, ubprobing);
1051 
1052  /* fix the variable and propagate the domain change */
1053  if( !SCIPisFeasEQ(probingscip, lbprobing, ubprobing) && SCIPvarIsActive(SCIPvarGetTransVar(probingvar)) )
1054  {
1055  assert(SCIPisFeasLE(probingscip, lbprobing, ubprobing));
1056  SCIPdebugMsg(scip, "try to fix variable <%s> (domain [%f,%f] to %f\n", SCIPvarGetName(probingvar), lbprobing, ubprobing,
1057  solval);
1058  SCIP_CALL( SCIPfixVarProbing(probingscip, probingvar, solval) );
1059  SCIP_CALL( SCIPpropagateProbing(probingscip, -1, &infeasible, &ndomreds) );
1060  SCIPdebugMsg(scip, " -> reduced %" SCIP_LONGINT_FORMAT " domains\n", ndomreds);
1061 
1062  if( infeasible )
1063  {
1064  SCIPdebugMsg(scip, " -> infeasible!\n");
1065  SCIP_CALL( SCIPbacktrackProbing(probingscip, 0) );
1066  }
1067  }
1068  else
1069  {
1070  SCIPdebugMsg(scip, "variable <%s> is already fixed to %f\n", SCIPvarGetName(probingvar), solval);
1071  }
1072  }
1073  }
1074 
1075  SCIP_CALL( updateVariableRounding(scip, heurdata, var, solval, alpha, scalingfactor) );
1076 
1077  /* check whether the variable is one of the most fractionals and label if so */
1078  if( SCIPisFeasPositive(scip, frac) )
1079  insertFlipCand(mostfracvars, mostfracvals, &nflipcands, maxnflipcands, var, frac);
1080  }
1081 
1082  if( heurdata->usefp20 )
1083  {
1084  SCIP_CALL( SCIPbacktrackProbing(probingscip, 0) );
1085  }
1086 
1087  /* change objective coefficients for continuous variables */
1088  for( i = nbinvars+nintvars; i < nvars; i++ )
1089  {
1090  SCIP_CALL( SCIPchgVarObjDive(scip, vars[i], alpha * SCIPvarGetObj(vars[i])) );
1091  }
1092 
1093  SCIPfreeBufferArray(scip, &pseudocands);
1094 
1095  /* initialize cycle check */
1096  minimum = MIN(heurdata->cyclelength, nloops-1);
1097  for( j = 0; j < heurdata->cyclelength; j++ )
1098  cycles[j] = (nloops > j+1) && (REALABS(lastalphas[j] - alpha) < heurdata->alphadiff);
1099 
1100  /* check for j-cycles */
1101  for( i = 0; i < nbinvars+nintvars; i++ )
1102  {
1103  solval = SCIPgetSolVal(scip, heurdata->roundedsol, vars[i]);
1104 
1105  /* cycles exist, iff all solution values are equal */
1106  for( j = 0; j < minimum; j++ )
1107  {
1108  oldsolval = SCIPgetSolVal(scip, lastroundedsols[j], vars[i]);
1109  cycles[j] = cycles[j] && SCIPisFeasEQ(scip, solval, oldsolval);
1110  }
1111  }
1112 
1113  /* force to flip variables at random after a couple of pumping rounds,
1114  * or if a new best solution in the current region has been found
1115  */
1116  assert(heurdata->perturbfreq > 0);
1117  if( nloops % heurdata->perturbfreq == 0 || (heurdata->pertsolfound && SCIPgetNBestSolsFound(scip) > nbestsolsfound) )
1118  {
1119  SCIPdebugMsg(scip, " -> random perturbation\n");
1120  SCIP_CALL( handleCycle(scip, heurdata, vars, nintvars+nbinvars, alpha, scalingfactor) );
1121  nbestsolsfound = SCIPgetNBestSolsFound(scip);
1122  }
1123  else
1124  {
1125  minimum = MIN(heurdata->cyclelength, nloops-1);
1126 
1127  for( j = 0; j < minimum; j++ )
1128  {
1129  /* if we got the same rounded solution as in some step before, we have to flip some variables */
1130  if( cycles[j] )
1131  {
1132  /* 1-cycles have a special flipping rule (flip most fractional variables) */
1133  if( j == 0 )
1134  {
1135  SCIPdebugMsg(scip, " -> avoiding 1-cycle: flipping %d candidates\n", nflipcands);
1136  SCIP_CALL( handle1Cycle(scip, heurdata, mostfracvars, nflipcands, alpha, scalingfactor) );
1137  }
1138  else
1139  {
1140  SCIPdebugMsg(scip, " -> avoiding %d-cycle by random flip\n", j+1);
1141  SCIP_CALL( handleCycle(scip, heurdata, vars, nintvars+nbinvars, alpha, scalingfactor) );
1142  }
1143  break;
1144  }
1145  }
1146  }
1147 
1148  /* the LP with the new (distance) objective is solved */
1149  nlpiterations = SCIPgetNLPIterations(scip);
1150  nlpiterationsleft = adjustedMaxNLPIterations(maxnlpiterations, nsolsfound, nstallloops) - heurdata->nlpiterations;
1151  iterlimit = MAX((int)nlpiterationsleft, MINLPITER);
1152  SCIPdebugMsg(scip, " -> solve LP with iteration limit %d\n", iterlimit);
1153 
1154  if( heurdata->stage3 )
1155  {
1156  SCIP_CALL( SCIPunlinkSol(scip, heurdata->roundedsol) );
1157  }
1158 
1159  retcode = SCIPsolveDiveLP(scip, iterlimit, &lperror, NULL);
1160  lpsolstat = SCIPgetLPSolstat(scip);
1161 
1162  /* Errors in the LP solver should not kill the overall solving process, if the LP is just needed for a heuristic.
1163  * Hence in optimized mode, the return code is caught and a warning is printed, only in debug mode, SCIP will stop.
1164  */
1165  if( retcode != SCIP_OKAY )
1166  {
1167 #ifndef NDEBUG
1168  if( lpsolstat != SCIP_LPSOLSTAT_UNBOUNDEDRAY )
1169  {
1170  SCIP_CALL( retcode );
1171  }
1172 #endif
1173  SCIPwarningMessage(scip, "Error while solving LP in Feaspump heuristic; LP solve terminated with code <%d>\n", retcode);
1174  SCIPwarningMessage(scip, "This does not affect the remaining solution procedure --> continue\n");
1175  }
1176 
1177  /* update iteration count */
1178  heurdata->nlpiterations += SCIPgetNLPIterations(scip) - nlpiterations;
1179  SCIPdebugMsg(scip, " -> number of iterations: %" SCIP_LONGINT_FORMAT "/%" SCIP_LONGINT_FORMAT ", lperror=%u, lpsolstat=%d\n",
1180  heurdata->nlpiterations, adjustedMaxNLPIterations(maxnlpiterations, nsolsfound, nstallloops), lperror, lpsolstat);
1181 
1182  /* check whether LP was solved optimal */
1183  if( lperror || lpsolstat != SCIP_LPSOLSTAT_OPTIMAL )
1184  break;
1185 
1186  if( heurdata->stage3 )
1187  {
1188  SCIP_Real distance; /* distance of the current rounded solution from the LP solution */
1189 
1190  assert(closestsol != NULL);
1191 
1192  /* calculate distance */
1193  distance = 0.0;
1194  for( i = 0; i < nbinvars+nintvars; i++ )
1195  {
1196  SCIP_Real roundedval;
1197  SCIP_Real lpval;
1198 
1199  roundedval = SCIPgetSolVal(scip, heurdata->roundedsol, vars[i]);
1200  lpval = SCIPvarGetLPSol(vars[i]);
1201  distance += REALABS(roundedval - lpval);
1202  }
1203 
1204  /* copy solution and update minimum distance */
1205  if( SCIPisLT(scip, distance, mindistance) )
1206  {
1207  for( i = 0; i < nbinvars+nintvars; i++ )
1208  {
1209  assert(SCIPisIntegral(scip,SCIPgetSolVal(scip, heurdata->roundedsol, vars[i])));
1210  SCIP_CALL( SCIPsetSolVal(scip, closestsol, vars[i], SCIPgetSolVal(scip, heurdata->roundedsol, vars[i])) );
1211  }
1212  mindistance = distance;
1213  }
1214  }
1215 
1216  /* swap the last solutions */
1217  SCIP_CALL( SCIPunlinkSol(scip, heurdata->roundedsol) );
1218  tmpsol = lastroundedsols[heurdata->cyclelength-1];
1219  for( j = heurdata->cyclelength-1; j > 0; j-- )
1220  {
1221  lastroundedsols[j] = lastroundedsols[j-1];
1222  lastalphas[j] = lastalphas[j-1];
1223  }
1224  lastroundedsols[0] = heurdata->roundedsol;
1225  lastalphas[0] = alpha;
1226  heurdata->roundedsol = tmpsol;
1227 
1228  /* check for improvement in number of fractionals */
1229  nfracs = SCIPgetNLPBranchCands(scip);
1230  if( nfracs < bestnfracs )
1231  {
1232  bestnfracs = nfracs;
1233  nstallloops = 0;
1234  }
1235  else
1236  nstallloops++;
1237 
1238  SCIPdebugMsg(scip, " -> loop finished: %d fractional variables (stall: %d/%d, iterations: %" SCIP_LONGINT_FORMAT "/%" SCIP_LONGINT_FORMAT ")\n",
1239  nfracs, nstallloops, maxstallloops, heurdata->nlpiterations, adjustedMaxNLPIterations(maxnlpiterations, nsolsfound, nstallloops));
1240  }
1241 
1242  /* try final solution, if no more fractional variables are left */
1243  if( nfracs == 0 && !lperror && lpsolstat == SCIP_LPSOLSTAT_OPTIMAL )
1244  {
1245  success = FALSE;
1246 
1247  SCIP_CALL( SCIPlinkLPSol(scip, heurdata->sol) );
1248  SCIPdebugMsg(scip, "feasibility pump found solution (%d fractional variables)\n", nfracs);
1249  SCIP_CALL( SCIPtrySol(scip, heurdata->sol, FALSE, FALSE, FALSE, FALSE, FALSE, &success) );
1250  if( success )
1251  *result = SCIP_FOUNDSOL;
1252  }
1253 
1254  /* end diving */
1255  SCIP_CALL( SCIPendDive(scip) );
1256 
1257  /* end probing in order to be able to apply stage 3 */
1258  if( heurdata->usefp20 )
1259  {
1260  SCIP_CALL( SCIPendProbing(probingscip) );
1261  }
1262 
1263  /* only do stage 3 if we have not found a solution yet */
1264  /* only do stage 3 if the distance of the closest infeasible solution to the polyhedron is below a certain threshold */
1265  if( heurdata->stage3 && (*result != SCIP_FOUNDSOL) && SCIPisLE(scip, mindistance, (SCIP_Real) heurdata->neighborhoodsize) )
1266  {
1267  SCIP_Bool cancopy;
1268  assert(closestsol != NULL);
1269  assert(!SCIPisInfinity(scip, mindistance) || nloops == 0);
1270 
1271  /* if we do not use feasibility pump 2.0, we have not created a copied SCIP instance yet */
1272  if( heurdata->usefp20 )
1273  {
1274  assert(probingscip != NULL);
1275  SCIP_CALL( SCIPfreeTransform(probingscip) );
1276  }
1277  else
1278  {
1279  assert(probingscip == NULL);
1280  SCIP_CALL( setupProbingSCIP(scip, &probingscip, &varmapfw, heurdata->copycuts, &success) );
1281  }
1282 
1283  /* check whether there is enough time and memory left */
1284  SCIP_CALL( SCIPcheckCopyLimits(scip, &cancopy) );
1285 
1286  if( cancopy )
1287  {
1288  SCIP_CALL( setupSCIPparamsStage3(scip, probingscip) );
1289 
1290  /* the neighborhood size is double the distance plus another ten percent */
1291  mindistance = SCIPceil(scip, 2.2*mindistance);
1292 
1293  SCIP_CALL( addLocalBranchingConstraint(scip, probingscip, varmapfw, closestsol, mindistance) );
1294 
1295  /* Errors in the LP solver should not kill the overall solving process, if the LP is just needed for a heuristic.
1296  * Hence in optimized mode, the return code is caught and a warning is printed, only in debug mode, SCIP will stop.
1297  */
1298  SCIP_CALL_ABORT( SCIPsolve(probingscip) );
1299 
1300  /* check, whether a solution was found */
1301  if( SCIPgetNSols(probingscip) > 0 )
1302  {
1303  success = FALSE;
1304  SCIP_CALL( createNewSols(scip, probingscip, varmapfw, heur, &success) );
1305  if( success )
1306  *result = SCIP_FOUNDSOL;
1307  }
1308  }
1309  }
1310 
1311  if( *result == SCIP_FOUNDSOL )
1312  heurdata->nsuccess++;
1313 
1314  /* free hash map and copied SCIP */
1315  if( varmapfw != NULL )
1316  SCIPhashmapFree(&varmapfw);
1317 
1318  if( probingscip != NULL )
1319  {
1320  SCIP_CALL( SCIPfree(&probingscip) );
1321  }
1322 
1323  if( heurdata->stage3 )
1324  {
1325  SCIP_CALL( SCIPfreeSol(scip, &closestsol) );
1326  }
1327 
1328  /* free memory */
1329  for( j = 0; j < heurdata->cyclelength; j++ )
1330  {
1331  SCIP_CALL( SCIPfreeSol(scip, &lastroundedsols[j]) );
1332  }
1333 
1334  SCIPfreeBufferArray(scip, &cycles);
1335  SCIPfreeBufferArray(scip, &lastalphas);
1336  SCIPfreeBufferArray(scip, &lastroundedsols);
1337  SCIPfreeBufferArray(scip, &mostfracvals);
1338  SCIPfreeBufferArray(scip, &mostfracvars);
1339 
1340  SCIPdebugMsg(scip, "feasibility pump finished [%d iterations done].\n", nloops);
1341 
1342 #ifdef SCIP_STATISTIC
1343  if( nfracs == 0 )
1344  {
1345  double objval;
1346  double primalBound;
1347 
1348  objval = SCIPgetSolOrigObj(scip, heurdata->sol);
1349  primalBound = SCIPgetPrimalbound(scip);
1350  SCIPstatisticMessage("feasibility pump found: 1, objval: %f, iterations: %d, primal bound: %f\n", objval, nloops, primalBound);
1351  }
1352  else
1353  {
1354  double primalBound;
1355 
1356  primalBound = SCIPgetPrimalbound(scip);
1357  SCIPstatisticMessage("feasibility pump found: 0, objval: +inf, iterations: %d, primal bound: %f\n", nloops, primalBound);
1358  }
1359 
1360 #endif /* SCIP_STATISTIC */
1361 
1362  return SCIP_OKAY;
1363 }
1364 
1365 
1366 /*
1367  * primal heuristic specific interface methods
1368  */
1369 
1370 /** creates the feaspump primal heuristic and includes it in SCIP */
1372  SCIP* scip /**< SCIP data structure */
1373  )
1374 {
1375  SCIP_HEURDATA* heurdata;
1376  SCIP_HEUR* heur;
1377 
1378  /* create Feaspump primal heuristic data */
1379  SCIP_CALL( SCIPallocBlockMemory(scip, &heurdata) );
1380 
1381  /* include primal heuristic */
1382  SCIP_CALL( SCIPincludeHeurBasic(scip, &heur,
1384  HEUR_MAXDEPTH, HEUR_TIMING, HEUR_USESSUBSCIP, heurExecFeaspump, heurdata) );
1385 
1386  assert(heur != NULL);
1387 
1388  /* set non-NULL pointers to callback methods */
1389  SCIP_CALL( SCIPsetHeurCopy(scip, heur, heurCopyFeaspump) );
1390  SCIP_CALL( SCIPsetHeurFree(scip, heur, heurFreeFeaspump) );
1391  SCIP_CALL( SCIPsetHeurInit(scip, heur, heurInitFeaspump) );
1392  SCIP_CALL( SCIPsetHeurExit(scip, heur, heurExitFeaspump) );
1393  SCIP_CALL( SCIPsetHeurInitsol(scip, heur, heurInitsolFeaspump) );
1394  SCIP_CALL( SCIPsetHeurExitsol(scip, heur, heurExitsolFeaspump) );
1395 
1396  /* add feaspump primal heuristic parameters */
1398  "heuristics/" HEUR_NAME "/maxlpiterquot",
1399  "maximal fraction of diving LP iterations compared to node LP iterations",
1400  &heurdata->maxlpiterquot, FALSE, DEFAULT_MAXLPITERQUOT, 0.0, SCIP_REAL_MAX, NULL, NULL) );
1402  "heuristics/" HEUR_NAME "/objfactor",
1403  "factor by which the regard of the objective is decreased in each round, 1.0 for dynamic",
1404  &heurdata->objfactor, FALSE, DEFAULT_OBJFACTOR, 0.0, 1.0, NULL, NULL) );
1406  "heuristics/" HEUR_NAME "/alpha",
1407  "initial weight of the objective function in the convex combination",
1408  &heurdata->alpha, FALSE, DEFAULT_ALPHA, 0.0, 1.0, NULL, NULL) );
1410  "heuristics/" HEUR_NAME "/alphadiff",
1411  "threshold difference for the convex parameter to perform perturbation",
1412  &heurdata->alphadiff, FALSE, DEFAULT_ALPHADIFF, 0.0, 1.0, NULL, NULL) );
1413 
1414  SCIP_CALL( SCIPaddIntParam(scip,
1415  "heuristics/" HEUR_NAME "/maxlpiterofs",
1416  "additional number of allowed LP iterations",
1417  &heurdata->maxlpiterofs, FALSE, DEFAULT_MAXLPITEROFS, 0, INT_MAX, NULL, NULL) );
1418  SCIP_CALL( SCIPaddIntParam(scip,
1419  "heuristics/" HEUR_NAME "/maxsols",
1420  "total number of feasible solutions found up to which heuristic is called (-1: no limit)",
1421  &heurdata->maxsols, TRUE, DEFAULT_MAXSOLS, -1, INT_MAX, NULL, NULL) );
1422  SCIP_CALL( SCIPaddIntParam(scip,
1423  "heuristics/" HEUR_NAME "/maxloops",
1424  "maximal number of pumping loops (-1: no limit)",
1425  &heurdata->maxloops, TRUE, DEFAULT_MAXLOOPS, -1, INT_MAX, NULL, NULL) );
1426  SCIP_CALL( SCIPaddIntParam(scip,
1427  "heuristics/" HEUR_NAME "/maxstallloops",
1428  "maximal number of pumping rounds without fractionality improvement (-1: no limit)",
1429  &heurdata->maxstallloops, TRUE, DEFAULT_MAXSTALLLOOPS, -1, INT_MAX, NULL, NULL) );
1430  SCIP_CALL( SCIPaddIntParam(scip,
1431  "heuristics/" HEUR_NAME "/minflips",
1432  "minimum number of random variables to flip, if a 1-cycle is encountered",
1433  &heurdata->minflips, TRUE, DEFAULT_MINFLIPS, 1, INT_MAX, NULL, NULL) );
1434  SCIP_CALL( SCIPaddIntParam(scip,
1435  "heuristics/" HEUR_NAME "/cyclelength",
1436  "maximum length of cycles to be checked explicitly in each round",
1437  &heurdata->cyclelength, TRUE, DEFAULT_CYCLELENGTH, 1, 100, NULL, NULL) );
1438  SCIP_CALL( SCIPaddIntParam(scip,
1439  "heuristics/" HEUR_NAME "/perturbfreq",
1440  "number of iterations until a random perturbation is forced",
1441  &heurdata->perturbfreq, TRUE, DEFAULT_PERTURBFREQ, 1, INT_MAX, NULL, NULL) );
1442  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/" HEUR_NAME "/neighborhoodsize",
1443  "radius (using Manhattan metric) of the neighborhood to be searched in stage 3",
1444  &heurdata->neighborhoodsize, FALSE, DEFAULT_NEIGHBORHOODSIZE, 1, INT_MAX, NULL, NULL) );
1445 
1447  "heuristics/" HEUR_NAME "/beforecuts",
1448  "should the feasibility pump be called at root node before cut separation?",
1449  &heurdata->beforecuts, FALSE, DEFAULT_BEFORECUTS, NULL, NULL) );
1451  "heuristics/" HEUR_NAME "/usefp20",
1452  "should an iterative round-and-propagate scheme be used to find the integral points?",
1453  &heurdata->usefp20, FALSE, DEFAULT_USEFP20, NULL, NULL) );
1455  "heuristics/" HEUR_NAME "/pertsolfound",
1456  "should a random perturbation be performed if a feasible solution was found?",
1457  &heurdata->pertsolfound, FALSE, DEFAULT_PERTSOLFOUND, NULL, NULL) );
1459  "heuristics/" HEUR_NAME "/stage3",
1460  "should we solve a local branching sub-MIP if no solution could be found?",
1461  &heurdata->stage3, FALSE, DEFAULT_STAGE3, NULL, NULL) );
1462  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/copycuts",
1463  "should all active cuts from cutpool be copied to constraints in subproblem?",
1464  &heurdata->copycuts, TRUE, DEFAULT_COPYCUTS, NULL, NULL) );
1465 
1466  return SCIP_OKAY;
1467 }
SCIP_RETCODE SCIPsetHeurExitsol(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEUREXITSOL((*heurexitsol)))
Definition: scip_heur.c:242
void SCIPfreeRandom(SCIP *scip, SCIP_RANDNUMGEN **randnumgen)
#define DEFAULT_NEIGHBORHOODSIZE
Definition: heur_feaspump.c:93
#define SCIP_HEURTIMING_DURINGLPLOOP
Definition: type_timing.h:80
SCIP_RETCODE SCIPlinkLPSol(SCIP *scip, SCIP_SOL *sol)
Definition: scip_sol.c:1026
SCIP_RETCODE SCIPsetSeparating(SCIP *scip, SCIP_PARAMSETTING paramsetting, SCIP_Bool quiet)
Definition: scip_param.c:958
#define DEFAULT_MAXSOLS
Definition: heur_feaspump.c:76
SCIP_Bool SCIPisFeasEQ(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
public methods for SCIP parameter handling
SCIP_STAGE SCIPgetStage(SCIP *scip)
Definition: scip_general.c:365
SCIP_RETCODE SCIPbacktrackProbing(SCIP *scip, int probingdepth)
Definition: scip_probing.c:225
SCIP_Longint SCIPgetNLPIterations(SCIP *scip)
#define HEUR_NAME
Definition: heur_feaspump.c:64
static SCIP_DECL_HEURINIT(heurInitFeaspump)
public methods for node selector plugins
#define HEUR_FREQOFS
Definition: heur_feaspump.c:69
public methods for memory management
#define DEFAULT_ALPHA
Definition: heur_feaspump.c:87
#define DEFAULT_CYCLELENGTH
Definition: heur_feaspump.c:82
SCIP_Real SCIPgetCutoffbound(SCIP *scip)
SCIP_Real SCIPgetPrimalbound(SCIP *scip)
#define DEFAULT_MAXLPITEROFS
Definition: heur_feaspump.c:75
static SCIP_RETCODE addLocalBranchingConstraint(SCIP *scip, SCIP *probingscip, SCIP_HASHMAP *varmapfw, SCIP_SOL *bestsol, SCIP_Real neighborhoodsize)
static SCIP_DECL_HEUREXITSOL(heurExitsolFeaspump)
#define SCIP_MAXSTRLEN
Definition: def.h:302
SCIP_Longint SCIPheurGetNBestSolsFound(SCIP_HEUR *heur)
Definition: heur.c:1596
static SCIP_RETCODE handleCycle(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_VAR **vars, int nbinandintvars, SCIP_Real alpha, SCIP_Real scalingfactor)
SCIP_Longint SCIPgetNSolsFound(SCIP *scip)
SCIP_RETCODE SCIPsetHeurExit(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEUREXIT((*heurexit)))
Definition: scip_heur.c:210
SCIP_Real SCIPvarGetLbLocal(SCIP_VAR *var)
Definition: var.c:17957
SCIP_Bool SCIPisGE(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
public solving methods
SCIP_RETCODE SCIPgetVarsData(SCIP *scip, SCIP_VAR ***vars, int *nvars, int *nbinvars, int *nintvars, int *nimplvars, int *ncontvars)
Definition: scip_prob.c:1866
SCIP_RETCODE SCIPsetHeuristics(SCIP *scip, SCIP_PARAMSETTING paramsetting, SCIP_Bool quiet)
Definition: scip_param.c:906
#define FALSE
Definition: def.h:96
SCIP_RETCODE SCIPhashmapCreate(SCIP_HASHMAP **hashmap, BMS_BLKMEM *blkmem, int mapsize)
Definition: misc.c:3024
SCIP_RETCODE SCIPcopyLimits(SCIP *sourcescip, SCIP *targetscip)
Definition: scip_copy.c:3292
#define HEUR_FREQ
Definition: heur_feaspump.c:68
SCIP_Real SCIPinfinity(SCIP *scip)
int SCIPsnprintf(char *t, int len, const char *s,...)
Definition: misc.c:10788
#define TRUE
Definition: def.h:95
enum SCIP_Retcode SCIP_RETCODE
Definition: type_retcode.h:63
#define SCIPstatisticMessage
Definition: pub_message.h:123
SCIP_RETCODE SCIPsetPresolving(SCIP *scip, SCIP_PARAMSETTING paramsetting, SCIP_Bool quiet)
Definition: scip_param.c:932
#define DEFAULT_COPYCUTS
Definition: heur_feaspump.c:94
SCIP_BRANCHRULE * SCIPfindBranchrule(SCIP *scip, const char *name)
Definition: scip_branch.c:297
int SCIPheurGetFreqofs(SCIP_HEUR *heur)
Definition: heur.c:1556
struct SCIP_HeurData SCIP_HEURDATA
Definition: type_heur.h:76
public methods for problem variables
#define SCIPfreeBlockMemory(scip, ptr)
Definition: scip_mem.h:108
SCIP_RETCODE SCIPincludeHeurBasic(SCIP *scip, SCIP_HEUR **heur, const char *name, const char *desc, char dispchar, int priority, int freq, int freqofs, int maxdepth, SCIP_HEURTIMING timingmask, SCIP_Bool usessubscip, SCIP_DECL_HEUREXEC((*heurexec)), SCIP_HEURDATA *heurdata)
Definition: scip_heur.c:117
SCIP_RETCODE SCIPtranslateSubSols(SCIP *scip, SCIP *subscip, SCIP_HEUR *heur, SCIP_VAR **subvars, SCIP_Bool *success, int *solindex)
Definition: scip_copy.c:1448
int SCIPrandomGetInt(SCIP_RANDNUMGEN *randnumgen, int minrandval, int maxrandval)
Definition: misc.c:10019
#define SCIPduplicateBufferArray(scip, ptr, source, num)
Definition: scip_mem.h:132
void * SCIPhashmapGetImage(SCIP_HASHMAP *hashmap, void *origin)
Definition: misc.c:3211
SCIP_Real SCIPfeasFrac(SCIP *scip, SCIP_Real val)
SCIP_Bool SCIPisEQ(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
#define SCIPfreeBufferArray(scip, ptr)
Definition: scip_mem.h:136
enum SCIP_LPSolStat SCIP_LPSOLSTAT
Definition: type_lp.h:51
SCIP_RETCODE SCIPcreate(SCIP **scip)
Definition: scip_general.c:292
void SCIPheurSetData(SCIP_HEUR *heur, SCIP_HEURDATA *heurdata)
Definition: heur.c:1371
#define SCIPallocBlockMemory(scip, ptr)
Definition: scip_mem.h:89
int SCIPgetNLPBranchCands(SCIP *scip)
Definition: scip_branch.c:428
public methods for SCIP variables
void SCIPwarningMessage(SCIP *scip, const char *formatstr,...)
Definition: scip_message.c:120
#define SCIPdebugMsg
Definition: scip_message.h:78
SCIP_RETCODE SCIPaddIntParam(SCIP *scip, const char *name, const char *desc, int *valueptr, SCIP_Bool isadvanced, int defaultvalue, int minvalue, int maxvalue, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:83
SCIP_RETCODE SCIPcopyParamSettings(SCIP *sourcescip, SCIP *targetscip)
Definition: scip_copy.c:2564
SCIP_Real SCIPgetObjNorm(SCIP *scip)
Definition: scip_prob.c:1641
SCIP_Real SCIPfeasCeil(SCIP *scip, SCIP_Real val)
public methods for numerical tolerances
SCIP_Real SCIPfeasFloor(SCIP *scip, SCIP_Real val)
public methods for querying solving statistics
const char * SCIPgetProbName(SCIP *scip)
Definition: scip_prob.c:1067
public methods for the branch-and-bound tree
#define DEFAULT_MAXSTALLLOOPS
Definition: heur_feaspump.c:80
SCIP_Bool SCIPisLPSolBasic(SCIP *scip)
Definition: scip_lp.c:667
#define MINLPITER
#define DEFAULT_RANDSEED
#define HEUR_USESSUBSCIP
Definition: heur_feaspump.c:72
SCIP_RETCODE SCIPsetHeurInitsol(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURINITSOL((*heurinitsol)))
Definition: scip_heur.c:226
SCIP_RETCODE SCIPsolve(SCIP *scip)
Definition: scip_solve.c:2631
const char * SCIPheurGetName(SCIP_HEUR *heur)
Definition: heur.c:1450
#define HEUR_DESC
Definition: heur_feaspump.c:65
#define DEFAULT_ALPHADIFF
Definition: heur_feaspump.c:88
SCIP_Bool SCIPisParamFixed(SCIP *scip, const char *name)
Definition: scip_param.c:219
SCIP_RETCODE SCIPaddCons(SCIP *scip, SCIP_CONS *cons)
Definition: scip_prob.c:2770
SCIP_RETCODE SCIPsolveDiveLP(SCIP *scip, int itlim, SCIP_Bool *lperror, SCIP_Bool *cutoff)
Definition: scip_lp.c:2678
SCIP_RETCODE SCIPsetHeurFree(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURFREE((*heurfree)))
Definition: scip_heur.c:178
SCIP_Bool SCIPisLT(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
static SCIP_DECL_HEURFREE(heurFreeFeaspump)
SCIP_RETCODE SCIPpropagateProbing(SCIP *scip, int maxproprounds, SCIP_Bool *cutoff, SCIP_Longint *ndomredsfound)
Definition: scip_probing.c:580
SCIP_RETCODE SCIPfixVarProbing(SCIP *scip, SCIP_VAR *var, SCIP_Real fixedval)
Definition: scip_probing.c:418
SCIP_RETCODE SCIPsetBoolParam(SCIP *scip, const char *name, SCIP_Bool value)
Definition: scip_param.c:429
SCIP_STATUS SCIPgetStatus(SCIP *scip)
Definition: scip_general.c:483
SCIP_RETCODE SCIPcopyCuts(SCIP *sourcescip, SCIP *targetscip, SCIP_HASHMAP *varmap, SCIP_HASHMAP *consmap, SCIP_Bool global, int *ncutsadded)
Definition: scip_copy.c:2130
BMS_BLKMEM * SCIPblkmem(SCIP *scip)
Definition: scip_mem.c:57
SCIP_RETCODE SCIPendProbing(SCIP *scip)
Definition: scip_probing.c:260
const char * SCIPvarGetName(SCIP_VAR *var)
Definition: var.c:17242
static SCIP_RETCODE handle1Cycle(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_VAR **mostfracvars, int nflipcands, SCIP_Real alpha, SCIP_Real scalingfactor)
void SCIPhashmapFree(SCIP_HASHMAP **hashmap)
Definition: misc.c:3058
#define DEFAULT_STAGE3
Definition: heur_feaspump.c:92
#define NULL
Definition: lpi_spx1.cpp:164
#define REALABS(x)
Definition: def.h:210
SCIP_Real SCIPvarGetLPSol(SCIP_VAR *var)
Definition: var.c:18275
SCIP_RETCODE SCIPunfixParam(SCIP *scip, const char *name)
Definition: scip_param.c:385
public methods for problem copies
#define SCIP_CALL(x)
Definition: def.h:394
static SCIP_DECL_HEURCOPY(heurCopyFeaspump)
SCIP_Bool SCIPisFeasLE(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
#define DEFAULT_BEFORECUTS
Definition: heur_feaspump.c:89
SCIP_Longint SCIPheurGetNCalls(SCIP_HEUR *heur)
Definition: heur.c:1576
SCIP_RETCODE SCIPgetPseudoBranchCands(SCIP *scip, SCIP_VAR ***pseudocands, int *npseudocands, int *npriopseudocands)
Definition: scip_branch.c:733
SCIP_Bool SCIPhasCurrentNodeLP(SCIP *scip)
Definition: scip_lp.c:83
public methods for primal heuristic plugins and divesets
SCIP_Longint SCIPgetLastDivenode(SCIP *scip)
Definition: scip_lp.c:2745
public methods for constraint handler plugins and constraints
#define DEFAULT_OBJFACTOR
Definition: heur_feaspump.c:84
SCIP_RETCODE SCIPchgVarObj(SCIP *scip, SCIP_VAR *var, SCIP_Real newobj)
Definition: scip_var.c:4513
SCIP_RETCODE SCIPcreateRandom(SCIP *scip, SCIP_RANDNUMGEN **randnumgen, unsigned int initialseed, SCIP_Bool useglobalseed)
static SCIP_RETCODE updateVariableRounding(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_VAR *var, SCIP_Real solval, SCIP_Real alpha, SCIP_Real scalingfactor)
SCIP_RETCODE SCIPcopyConsCompression(SCIP *sourcescip, SCIP *targetscip, SCIP_HASHMAP *varmap, SCIP_HASHMAP *consmap, const char *suffix, SCIP_VAR **fixedvars, SCIP_Real *fixedvals, int nfixedvars, SCIP_Bool global, SCIP_Bool enablepricing, SCIP_Bool threadsafe, SCIP_Bool passmessagehdlr, SCIP_Bool *valid)
Definition: scip_copy.c:2965
#define SCIPallocBufferArray(scip, ptr, num)
Definition: scip_mem.h:124
SCIP_RETCODE SCIPsetSolVal(SCIP *scip, SCIP_SOL *sol, SCIP_VAR *var, SCIP_Real val)
Definition: scip_sol.c:1221
public data structures and miscellaneous methods
#define DEFAULT_MINFLIPS
Definition: heur_feaspump.c:81
SCIP_RETCODE SCIPfreeTransform(SCIP *scip)
Definition: scip_solve.c:3476
#define SCIP_Bool
Definition: def.h:93
SCIP_LPSOLSTAT SCIPgetLPSolstat(SCIP *scip)
Definition: scip_lp.c:168
enum SCIP_Status SCIP_STATUS
Definition: type_stat.h:67
int SCIPgetNPricers(SCIP *scip)
Definition: scip_pricer.c:337
int SCIPgetDepth(SCIP *scip)
Definition: scip_tree.c:670
#define DEFAULT_MAXLOOPS
Definition: heur_feaspump.c:79
#define MAX(x, y)
Definition: tclique_def.h:92
Objective Feasibility Pump 2.0.
SCIP_RETCODE SCIPsetIntParam(SCIP *scip, const char *name, int value)
Definition: scip_param.c:487
SCIP_RETCODE SCIPfreeSol(SCIP *scip, SCIP_SOL **sol)
Definition: scip_sol.c:985
SCIP_Real SCIPvarGetObj(SCIP_VAR *var)
Definition: var.c:17749
#define HEUR_PRIORITY
Definition: heur_feaspump.c:67
int SCIPgetNSols(SCIP *scip)
Definition: scip_sol.c:2214
SCIP_Real SCIPgetSolOrigObj(SCIP *scip, SCIP_SOL *sol)
Definition: scip_sol.c:1444
Constraint handler for linear constraints in their most general form, .
static SCIP_DECL_HEUREXEC(heurExecFeaspump)
SCIP_Bool SCIPisInfinity(SCIP *scip, SCIP_Real val)
SCIP_RETCODE SCIPtrySol(SCIP *scip, SCIP_SOL *sol, SCIP_Bool printreason, SCIP_Bool completely, SCIP_Bool checkbounds, SCIP_Bool checkintegrality, SCIP_Bool checklprows, SCIP_Bool *stored)
Definition: scip_sol.c:3098
#define SCIP_MAXTREEDEPTH
Definition: def.h:330
static SCIP_DECL_HEUREXIT(heurExitFeaspump)
SCIP_Real SCIPrandomGetReal(SCIP_RANDNUMGEN *randnumgen, SCIP_Real minrandval, SCIP_Real maxrandval)
Definition: misc.c:10041
public methods for the LP relaxation, rows and columns
public methods for variable pricer plugins
int SCIPgetNVars(SCIP *scip)
Definition: scip_prob.c:1992
#define SCIP_REAL_MAX
Definition: def.h:187
methods for sorting joint arrays of various types
SCIP_RETCODE SCIPcreateConsLinear(SCIP *scip, SCIP_CONS **cons, const char *name, int nvars, SCIP_VAR **vars, SCIP_Real *vals, SCIP_Real lhs, SCIP_Real rhs, SCIP_Bool initial, SCIP_Bool separate, SCIP_Bool enforce, SCIP_Bool check, SCIP_Bool propagate, SCIP_Bool local, SCIP_Bool modifiable, SCIP_Bool dynamic, SCIP_Bool removable, SCIP_Bool stickingatnode)
static void insertFlipCand(SCIP_VAR **mostfracvars, SCIP_Real *mostfracvals, int *nflipcands, int maxnflipcands, SCIP_VAR *var, SCIP_Real frac)
public methods for branching rule plugins and branching
general public methods
SCIP_Real SCIPgetLPObjval(SCIP *scip)
Definition: scip_lp.c:247
SCIP_Bool SCIPisIntegral(SCIP *scip, SCIP_Real val)
SCIP_RETCODE SCIPchgVarObjDive(SCIP *scip, SCIP_VAR *var, SCIP_Real newobj)
Definition: scip_lp.c:2378
public methods for solutions
public methods for random numbers
public methods for the probing mode
SCIP_RETCODE SCIPreleaseCons(SCIP *scip, SCIP_CONS **cons)
Definition: scip_cons.c:1119
public methods for message output
#define DEFAULT_PERTURBFREQ
Definition: heur_feaspump.c:83
SCIP_RETCODE SCIPsetHeurInit(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURINIT((*heurinit)))
Definition: scip_heur.c:194
SCIP_Bool SCIPisFeasPositive(SCIP *scip, SCIP_Real val)
SCIP_NODESEL * SCIPfindNodesel(SCIP *scip, const char *name)
Definition: scip_nodesel.c:234
SCIP_Longint SCIPgetNBestSolsFound(SCIP *scip)
#define SCIP_Real
Definition: def.h:186
#define DEFAULT_PERTSOLFOUND
Definition: heur_feaspump.c:91
SCIP_Bool SCIPisStopped(SCIP *scip)
Definition: scip_general.c:703
#define DEFAULT_MAXLPITERQUOT
Definition: heur_feaspump.c:74
public methods for message handling
void SCIPheurSetTimingmask(SCIP_HEUR *heur, SCIP_HEURTIMING timingmask)
Definition: heur.c:1490
#define HEUR_MAXDEPTH
Definition: heur_feaspump.c:70
SCIP_RETCODE SCIPincludeHeurFeaspump(SCIP *scip)
#define SCIP_Longint
Definition: def.h:171
void SCIPsortRealPtr(SCIP_Real *realarray, void **ptrarray, int len)
SCIP_RETCODE SCIPunlinkSol(SCIP *scip, SCIP_SOL *sol)
Definition: scip_sol.c:1190
SCIP_RETCODE SCIPcheckCopyLimits(SCIP *sourcescip, SCIP_Bool *success)
Definition: scip_copy.c:3249
SCIP_VARTYPE SCIPvarGetType(SCIP_VAR *var)
Definition: var.c:17407
SCIP_RETCODE SCIPstartDive(SCIP *scip)
Definition: scip_lp.c:2242
SCIP_Bool SCIPisLE(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
SCIP_RETCODE SCIPsetHeurCopy(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURCOPY((*heurcopy)))
Definition: scip_heur.c:162
SCIP_Real SCIPvarGetUbLocal(SCIP_VAR *var)
Definition: var.c:17967
SCIP_RETCODE SCIPnewProbingNode(SCIP *scip)
Definition: scip_probing.c:165
SCIP_Bool SCIPisFeasIntegral(SCIP *scip, SCIP_Real val)
#define DEFAULT_USEFP20
Definition: heur_feaspump.c:90
SCIP_RETCODE SCIPstartProbing(SCIP *scip)
Definition: scip_probing.c:119
static SCIP_RETCODE setupSCIPparamsStage3(SCIP *scip, SCIP *probingscip)
public methods for primal heuristics
static SCIP_DECL_HEURINITSOL(heurInitsolFeaspump)
SCIP_RETCODE SCIPendDive(SCIP *scip)
Definition: scip_lp.c:2291
#define HEUR_DISPCHAR
Definition: heur_feaspump.c:66
#define SCIP_CALL_ABORT(x)
Definition: def.h:373
SCIP_Real SCIPceil(SCIP *scip, SCIP_Real val)
SCIP_HEURDATA * SCIPheurGetData(SCIP_HEUR *heur)
Definition: heur.c:1361
SCIP_VAR * SCIPvarGetTransVar(SCIP_VAR *var)
Definition: var.c:17601
SCIP_Longint SCIPgetNNodes(SCIP *scip)
public methods for global and local (sub)problems
static SCIP_RETCODE setupSCIPparamsFP2(SCIP *scip, SCIP *probingscip)
static SCIP_Longint adjustedMaxNLPIterations(SCIP_Longint maxnlpiterations, SCIP_Longint nsolsfound, int nstallloops)
SCIP_Real SCIPgetSolVal(SCIP *scip, SCIP_SOL *sol, SCIP_VAR *var)
Definition: scip_sol.c:1361
static SCIP_RETCODE createNewSols(SCIP *scip, SCIP *subscip, SCIP_HASHMAP *varmapfw, SCIP_HEUR *heur, SCIP_Bool *success)
SCIP_RETCODE SCIPaddRealParam(SCIP *scip, const char *name, const char *desc, SCIP_Real *valueptr, SCIP_Bool isadvanced, SCIP_Real defaultvalue, SCIP_Real minvalue, SCIP_Real maxvalue, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:139
static SCIP_RETCODE setupProbingSCIP(SCIP *scip, SCIP **probingscip, SCIP_HASHMAP **varmapfw, SCIP_Bool copycuts, SCIP_Bool *success)
SCIP_RETCODE SCIPsetSubscipsOff(SCIP *scip, SCIP_Bool quiet)
Definition: scip_param.c:883
SCIP_Real SCIPfloor(SCIP *scip, SCIP_Real val)
SCIP_RETCODE SCIPsetLongintParam(SCIP *scip, const char *name, SCIP_Longint value)
Definition: scip_param.c:545
#define HEUR_TIMING
Definition: heur_feaspump.c:71
SCIP_RETCODE SCIPaddBoolParam(SCIP *scip, const char *name, const char *desc, SCIP_Bool *valueptr, SCIP_Bool isadvanced, SCIP_Bool defaultvalue, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:57
SCIP_Bool SCIPvarIsActive(SCIP_VAR *var)
Definition: var.c:17571
SCIP_RETCODE SCIPfree(SCIP **scip)
Definition: scip_general.c:324
SCIP_RETCODE SCIPcreateSol(SCIP *scip, SCIP_SOL **sol, SCIP_HEUR *heur)
Definition: scip_sol.c:328
memory allocation routines