Scippy

SCIP

Solving Constraint Integer Programs

heur_alns.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_alns.c
26  * @ingroup DEFPLUGINS_HEUR
27  * @brief Adaptive large neighborhood search heuristic that orchestrates popular LNS heuristics
28  * @author Gregor Hendel
29  */
30 
31 /*---+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8----+----9----+----0----+----1----+----2*/
32 
33 #include "blockmemshell/memory.h"
34 #include "scip/cons_linear.h"
35 #include "scip/heur_alns.h"
36 #include "scip/heuristics.h"
38 #include "scip/pub_bandit_exp3.h"
39 #include "scip/pub_bandit.h"
40 #include "scip/pub_bandit_ucb.h"
41 #include "scip/pub_cons.h"
42 #include "scip/pub_event.h"
43 #include "scip/pub_heur.h"
44 #include "scip/pub_message.h"
45 #include "scip/pub_misc.h"
46 #include "scip/pub_misc_select.h"
47 #include "scip/pub_sol.h"
48 #include "scip/pub_var.h"
49 #include "scip/scip_bandit.h"
50 #include "scip/scip_branch.h"
51 #include "scip/scip_cons.h"
52 #include "scip/scip_copy.h"
53 #include "scip/scip_event.h"
54 #include "scip/scip_general.h"
55 #include "scip/scip_heur.h"
56 #include "scip/scip_lp.h"
57 #include "scip/scip_mem.h"
58 #include "scip/scip_message.h"
59 #include "scip/scip_nodesel.h"
60 #include "scip/scip_numerics.h"
61 #include "scip/scip_param.h"
62 #include "scip/scip_prob.h"
63 #include "scip/scip_randnumgen.h"
64 #include "scip/scip_sol.h"
65 #include "scip/scip_solve.h"
66 #include "scip/scip_solvingstats.h"
67 #include "scip/scip_table.h"
68 #include "scip/scip_timing.h"
69 #include "scip/scip_tree.h"
70 #include "scip/scip_var.h"
71 #include <string.h>
72 
73 #if !defined(_WIN32) && !defined(_WIN64)
74 #include <strings.h> /*lint --e{766}*/ /* needed for strncasecmp() */
75 #endif
76 
77 #define HEUR_NAME "alns"
78 #define HEUR_DESC "Large neighborhood search heuristic that orchestrates the popular neighborhoods Local Branching, RINS, RENS, DINS etc."
79 #define HEUR_DISPCHAR SCIP_HEURDISPCHAR_LNS
80 #define HEUR_PRIORITY -1100500
81 #define HEUR_FREQ 20
82 #define HEUR_FREQOFS 0
83 #define HEUR_MAXDEPTH -1
84 #define HEUR_TIMING SCIP_HEURTIMING_AFTERNODE | SCIP_HEURTIMING_DURINGLPLOOP
85 #define HEUR_USESSUBSCIP TRUE /**< does the heuristic use a secondary SCIP instance? */
86 
87 #define NNEIGHBORHOODS 9
88 
89 #define DEFAULT_SHOWNBSTATS FALSE /**< show statistics on neighborhoods? */
90 
91 /*
92  * limit parameters for sub-SCIPs
93  */
94 #define DEFAULT_NODESQUOT 0.1
95 #define DEFAULT_NODESQUOTMIN 0.0
96 #define DEFAULT_NODESOFFSET 500LL
97 #define DEFAULT_NSOLSLIM 3
98 #define DEFAULT_MINNODES 50LL
99 #define DEFAULT_MAXNODES 5000LL
100 #define DEFAULT_WAITINGNODES 25LL /**< number of nodes since last incumbent solution that the heuristic should wait */
101 #define DEFAULT_TARGETNODEFACTOR 1.05
102 #define LRATEMIN 0.01 /**< lower bound for learning rate for target nodes and minimum improvement */
103 #define LPLIMFAC 4.0
104 #define DEFAULT_INITDURINGROOT FALSE
105 #define DEFAULT_MAXCALLSSAMESOL -1 /**< number of allowed executions of the heuristic on the same incumbent solution */
106 
107 /*
108  * parameters for the minimum improvement
109  */
110 #define DEFAULT_MINIMPROVELOW 0.01
111 #define DEFAULT_MINIMPROVEHIGH 0.01
112 #define MINIMPROVEFAC 1.5
113 #define DEFAULT_STARTMINIMPROVE 0.01
114 #define DEFAULT_ADJUSTMINIMPROVE FALSE
115 #define DEFAULT_ADJUSTTARGETNODES TRUE /**< should the target nodes be dynamically adjusted? */
116 
117 /*
118  * bandit algorithm parameters
119  */
120 #define DEFAULT_BESTSOLWEIGHT 1
121 #define DEFAULT_BANDITALGO 'u' /**< the default bandit algorithm: (u)pper confidence bounds, (e)xp.3, epsilon (g)reedy */
122 #define DEFAULT_REWARDCONTROL 0.8 /**< reward control to increase the weight of the simple solution indicator and decrease the weight of the closed gap reward */
123 #define DEFAULT_SCALEBYEFFORT TRUE /**< should the reward be scaled by the effort? */
124 #define DEFAULT_RESETWEIGHTS TRUE /**< should the bandit algorithms be reset when a new problem is read? */
125 #define DEFAULT_SUBSCIPRANDSEEDS FALSE /**< should random seeds of sub-SCIPs be altered to increase diversification? */
126 #define DEFAULT_REWARDBASELINE 0.5 /**< the reward baseline to separate successful and failed calls */
127 #define DEFAULT_FIXTOL 0.1 /**< tolerance by which the fixing rate may be missed without generic fixing */
128 #define DEFAULT_UNFIXTOL 0.1 /**< tolerance by which the fixing rate may be exceeded without generic unfixing */
129 #define DEFAULT_USELOCALREDCOST FALSE /**< should local reduced costs be used for generic (un)fixing? */
130 #define DEFAULT_BETA 0.0 /**< default reward offset between 0 and 1 at every observation for exp3 */
131 
132 /*
133  * the following 3 parameters have been tuned by a simulation experiment
134  * as described in the paper.
135  */
136 #define DEFAULT_EPS 0.4685844 /**< increase exploration in epsilon-greedy bandit algorithm */
137 #define DEFAULT_ALPHA 0.0016 /**< parameter to increase the confidence width in UCB */
138 #define DEFAULT_GAMMA 0.07041455 /**< default weight between uniform (gamma ~ 1) and weight driven (gamma ~ 0) probability distribution for exp3 */
139 /*
140  * parameters to control variable fixing
141  */
142 #define DEFAULT_USEREDCOST TRUE /**< should reduced cost scores be used for variable priorization? */
143 #define DEFAULT_USEPSCOST TRUE /**< should pseudo cost scores be used for variable priorization? */
144 #define DEFAULT_USEDISTANCES TRUE /**< should distances from fixed variables be used for variable priorization */
145 #define DEFAULT_DOMOREFIXINGS TRUE /**< should the ALNS heuristic do more fixings by itself based on variable prioritization
146  * until the target fixing rate is reached? */
147 #define DEFAULT_ADJUSTFIXINGRATE TRUE /**< should the heuristic adjust the target fixing rate based on the success? */
148 #define FIXINGRATE_DECAY 0.75 /**< geometric decay for fixing rate adjustments */
149 #define FIXINGRATE_STARTINC 0.2 /**< initial increment value for fixing rate */
150 #define DEFAULT_USESUBSCIPHEURS FALSE /**< should the heuristic activate other sub-SCIP heuristics during its search? */
151 #define DEFAULT_COPYCUTS FALSE /**< should cutting planes be copied to the sub-SCIP? */
152 #define DEFAULT_REWARDFILENAME "-" /**< file name to store all rewards and the selection of the bandit */
154 /* individual random seeds */
155 #define DEFAULT_SEED 113
156 #define MUTATIONSEED 121
157 #define CROSSOVERSEED 321
159 /* individual neighborhood parameters */
160 #define DEFAULT_MINFIXINGRATE_RENS 0.3
161 #define DEFAULT_MAXFIXINGRATE_RENS 0.9
162 #define DEFAULT_ACTIVE_RENS TRUE
163 #define DEFAULT_PRIORITY_RENS 1.0
165 #define DEFAULT_MINFIXINGRATE_RINS 0.3
166 #define DEFAULT_MAXFIXINGRATE_RINS 0.9
167 #define DEFAULT_ACTIVE_RINS TRUE
168 #define DEFAULT_PRIORITY_RINS 1.0
170 #define DEFAULT_MINFIXINGRATE_MUTATION 0.3
171 #define DEFAULT_MAXFIXINGRATE_MUTATION 0.9
172 #define DEFAULT_ACTIVE_MUTATION TRUE
173 #define DEFAULT_PRIORITY_MUTATION 1.0
175 #define DEFAULT_MINFIXINGRATE_LOCALBRANCHING 0.3
176 #define DEFAULT_MAXFIXINGRATE_LOCALBRANCHING 0.9
177 #define DEFAULT_ACTIVE_LOCALBRANCHING TRUE
178 #define DEFAULT_PRIORITY_LOCALBRANCHING 1.0
180 #define DEFAULT_MINFIXINGRATE_PROXIMITY 0.3
181 #define DEFAULT_MAXFIXINGRATE_PROXIMITY 0.9
182 #define DEFAULT_ACTIVE_PROXIMITY TRUE
183 #define DEFAULT_PRIORITY_PROXIMITY 1.0
185 #define DEFAULT_MINFIXINGRATE_CROSSOVER 0.3
186 #define DEFAULT_MAXFIXINGRATE_CROSSOVER 0.9
187 #define DEFAULT_ACTIVE_CROSSOVER TRUE
188 #define DEFAULT_PRIORITY_CROSSOVER 1.0
190 #define DEFAULT_MINFIXINGRATE_ZEROOBJECTIVE 0.3
191 #define DEFAULT_MAXFIXINGRATE_ZEROOBJECTIVE 0.9
192 #define DEFAULT_ACTIVE_ZEROOBJECTIVE TRUE
193 #define DEFAULT_PRIORITY_ZEROOBJECTIVE 1.0
195 #define DEFAULT_MINFIXINGRATE_DINS 0.3
196 #define DEFAULT_MAXFIXINGRATE_DINS 0.9
197 #define DEFAULT_ACTIVE_DINS TRUE
198 #define DEFAULT_PRIORITY_DINS 1.0
200 #define DEFAULT_MINFIXINGRATE_TRUSTREGION 0.3
201 #define DEFAULT_MAXFIXINGRATE_TRUSTREGION 0.9
202 #define DEFAULT_ACTIVE_TRUSTREGION FALSE
203 #define DEFAULT_PRIORITY_TRUSTREGION 1.0
205 
206 #define DEFAULT_NSOLS_CROSSOVER 2 /**< parameter for the number of solutions that crossover should combine */
207 #define DEFAULT_NPOOLSOLS_DINS 5 /**< number of pool solutions where binary solution values must agree */
208 #define DEFAULT_VIOLPENALTY_TRUSTREGION 100.0 /**< the penalty for violating the trust region */
210 /* event handler properties */
211 #define EVENTHDLR_NAME "Alns"
212 #define EVENTHDLR_DESC "LP event handler for " HEUR_NAME " heuristic"
213 #define SCIP_EVENTTYPE_ALNS (SCIP_EVENTTYPE_LPSOLVED | SCIP_EVENTTYPE_SOLFOUND | SCIP_EVENTTYPE_BESTSOLFOUND)
215 /* properties of the ALNS neighborhood statistics table */
216 #define TABLE_NAME_NEIGHBORHOOD "neighborhood"
217 #define TABLE_DESC_NEIGHBORHOOD "ALNS neighborhood statistics"
218 #define TABLE_POSITION_NEIGHBORHOOD 12500 /**< the position of the statistics table */
219 #define TABLE_EARLIEST_STAGE_NEIGHBORHOOD SCIP_STAGE_TRANSFORMED /**< output of the statistics table is only printed from this stage onwards */
221 
222 /** reward types of ALNS */
223 enum RewardType {
224  REWARDTYPE_TOTAL, /**< combination of the other rewards */
225  REWARDTYPE_BESTSOL, /**< 1, if a new solution was found, 0 otherwise */
226  REWARDTYPE_CLOSEDGAP, /**< 0 if no solution was found, closed gap otherwise */
227  REWARDTYPE_NOSOLPENALTY, /**< 1 if a solution was found, otherwise between 0 and 1 depending on the effort spent */
229 };
230 
231 /*
232  * Data structures
233  */
234 
235 /*
236  * additional neighborhood data structures
237  */
238 
239 
240 typedef struct data_crossover DATA_CROSSOVER; /**< crossover neighborhood data structure */
242 typedef struct data_mutation DATA_MUTATION; /**< mutation neighborhood data structure */
244 typedef struct data_dins DATA_DINS; /**< dins neighborhood data structure */
246 typedef struct data_trustregion DATA_TRUSTREGION; /**< trustregion neighborhood data structure */
248 typedef struct NH_FixingRate NH_FIXINGRATE; /** fixing rate data structure */
250 typedef struct NH_Stats NH_STATS; /**< neighborhood statistics data structure */
252 typedef struct Nh NH; /**< neighborhood data structure */
254 
255 /*
256  * variable priorization data structure for sorting
257  */
258 typedef struct VarPrio VARPRIO;
260 /** callback to collect variable fixings of neighborhood */
261  #define DECL_VARFIXINGS(x) SCIP_RETCODE x ( \
262  SCIP* scip, /**< SCIP data structure */ \
263  NH* neighborhood, /**< ALNS neighborhood data structure */ \
264  SCIP_VAR** varbuf, /**< buffer array to collect variables to fix */\
265  SCIP_Real* valbuf, /**< buffer array to collect fixing values */ \
266  int* nfixings, /**< pointer to store the number of fixings */ \
267  SCIP_RESULT* result /**< result pointer */ \
268  )
269 
270 /** callback for subproblem changes other than variable fixings
271  *
272  * this callback can be used to further modify the subproblem by changes other than variable fixings.
273  * Typical modifications include restrictions of variable domains, the formulation of additional constraints,
274  * or changed objective coefficients.
275  *
276  * The callback should set the \p success pointer to indicate whether it was successful with its modifications or not.
277  */
278 #define DECL_CHANGESUBSCIP(x) SCIP_RETCODE x ( \
279  SCIP* sourcescip, /**< source SCIP data structure */\
280  SCIP* targetscip, /**< target SCIP data structure */\
281  NH* neighborhood, /**< ALNS neighborhood data structure */\
282  SCIP_VAR** subvars, /**< array of targetscip variables in the same order as the source SCIP variables */\
283  int* ndomchgs, /**< pointer to store the number of performed domain changes */\
284  int* nchgobjs, /**< pointer to store the number of changed objective coefficients */ \
285  int* naddedconss, /**< pointer to store the number of additional constraints */\
286  SCIP_Bool* success /**< pointer to store if the sub-MIP was successfully adjusted */\
287  )
288 
289 /** optional initialization callback for neighborhoods when a new problem is read */
290 #define DECL_NHINIT(x) SCIP_RETCODE x ( \
291  SCIP* scip, /**< SCIP data structure */ \
292  NH* neighborhood /**< neighborhood data structure */ \
293  )
294 
295 /** deinitialization callback for neighborhoods when exiting a problem */
296 #define DECL_NHEXIT(x) SCIP_RETCODE x ( \
297  SCIP* scip, /**< SCIP data structure */ \
298  NH* neighborhood /**< neighborhood data structure */ \
299  )
300 
301 /** deinitialization callback for neighborhoods before SCIP is freed */
302 #define DECL_NHFREE(x) SCIP_RETCODE x ( \
303  SCIP* scip, /**< SCIP data structure */ \
304  NH* neighborhood /**< neighborhood data structure */ \
305  )
306 
307 /** callback function to return a feasible reference solution for further fixings
308  *
309  * The reference solution should be stored in the \p solptr.
310  * The \p result pointer can be used to indicate either
311  *
312  * - SCIP_SUCCESS or
313  * - SCIP_DIDNOTFIND
314  */
315 #define DECL_NHREFSOL(x) SCIP_RETCODE x ( \
316  SCIP* scip, /**< SCIP data structure */ \
317  NH* neighborhood, /**< neighborhood data structure */ \
318  SCIP_SOL** solptr, /**< pointer to store the reference solution */ \
319  SCIP_RESULT* result /**< pointer to indicate the callback success whether a reference solution is available */ \
320  )
321 
322 /** callback function to deactivate neighborhoods on problems where they are irrelevant */
323 #define DECL_NHDEACTIVATE(x) SCIP_RETCODE x (\
324  SCIP* scip, /**< SCIP data structure */ \
325  SCIP_Bool* deactivate /**< pointer to store whether the neighborhood should be deactivated (TRUE) for an instance */ \
326  )
327 
328 /** sub-SCIP status code enumerator */
329 enum HistIndex
330 {
331  HIDX_OPT = 0, /**< sub-SCIP was solved to optimality */
332  HIDX_USR = 1, /**< sub-SCIP was user interrupted */
333  HIDX_NODELIM = 2, /**< sub-SCIP reached the node limit */
334  HIDX_STALLNODE = 3, /**< sub-SCIP reached the stall node limit */
335  HIDX_INFEAS = 4, /**< sub-SCIP was infeasible */
336  HIDX_SOLLIM = 5, /**< sub-SCIP reached the solution limit */
337  HIDX_OTHER = 6 /**< sub-SCIP reached none of the above codes */
338 };
339 typedef enum HistIndex HISTINDEX;
340 #define NHISTENTRIES 7
342 
343 /** statistics for a neighborhood */
344 struct NH_Stats
345 {
346  SCIP_CLOCK* setupclock; /**< clock for sub-SCIP setup time */
347  SCIP_CLOCK* submipclock; /**< clock for the sub-SCIP solve */
348  SCIP_Longint usednodes; /**< total number of used nodes */
349  SCIP_Real oldupperbound; /**< upper bound before the sub-SCIP started */
350  SCIP_Real newupperbound; /**< new upper bound for allrewards mode to work correctly */
351  int nruns; /**< number of runs of a neighborhood */
352  int nrunsbestsol; /**< number of runs that produced a new incumbent */
353  SCIP_Longint nsolsfound; /**< the total number of solutions found */
354  SCIP_Longint nbestsolsfound; /**< the total number of improving solutions found */
355  int nfixings; /**< the number of fixings in one run */
356  int statushist[NHISTENTRIES]; /**< array to count sub-SCIP statuses */
357 };
358 
359 
360 /** fixing rate data structure to control the amount of target fixings of a neighborhood */
361 struct NH_FixingRate
362 {
363  SCIP_Real minfixingrate; /**< the minimum fixing rate */
364  SCIP_Real targetfixingrate; /**< the current target fixing rate */
365  SCIP_Real increment; /**< the current increment by which the target fixing rate is in-/decreased */
366  SCIP_Real maxfixingrate; /**< the maximum fixing rate */
367 };
368 
369 /** neighborhood data structure with callbacks, statistics, fixing rate */
370 struct Nh
371 {
372  char* name; /**< the name of this neighborhood */
373  NH_FIXINGRATE fixingrate; /**< fixing rate for this neighborhood */
374  NH_STATS stats; /**< statistics for this neighborhood */
375  DECL_VARFIXINGS ((*varfixings)); /**< variable fixings callback for this neighborhood */
376  DECL_CHANGESUBSCIP ((*changesubscip)); /**< callback for subproblem changes other than variable fixings */
377  DECL_NHINIT ((*nhinit)); /**< initialization callback when a new problem is read */
378  DECL_NHEXIT ((*nhexit)); /**< deinitialization callback when exiting a problem */
379  DECL_NHFREE ((*nhfree)); /**< deinitialization callback before SCIP is freed */
380  DECL_NHREFSOL ((*nhrefsol)); /**< callback function to return a reference solution for further fixings, or NULL */
381  DECL_NHDEACTIVATE ((*nhdeactivate)); /**< callback function to deactivate neighborhoods on problems where they are irrelevant, or NULL if it is always active */
382  SCIP_Bool active; /**< is this neighborhood active or not? */
383  SCIP_Real priority; /**< positive call priority to initialize bandit algorithms */
384  union
385  {
386  DATA_MUTATION* mutation; /**< mutation data */
387  DATA_CROSSOVER* crossover; /**< crossover data */
388  DATA_DINS* dins; /**< dins data */
389  DATA_TRUSTREGION* trustregion; /**< trustregion data */
390  } data; /**< data object for neighborhood specific data */
391 };
392 
393 /** mutation neighborhood data structure */
394 struct data_mutation
395 {
396  SCIP_RANDNUMGEN* rng; /**< random number generator */
397 };
398 
399 /** crossover neighborhood data structure */
400 struct data_crossover
401 {
402  int nsols; /**< the number of solutions that crossover should combine */
403  SCIP_RANDNUMGEN* rng; /**< random number generator to draw from the solution pool */
404  SCIP_SOL* selsol; /**< best selected solution by crossover as reference point */
405 };
406 
407 /** dins neighborhood data structure */
408 struct data_dins
409 {
410  int npoolsols; /**< number of pool solutions where binary solution values must agree */
411 };
412 
413 struct data_trustregion
414 {
415  SCIP_Real violpenalty; /**< the penalty for violating the trust region */
416 };
417 
418 /** primal heuristic data */
419 struct SCIP_HeurData
420 {
421  NH** neighborhoods; /**< array of neighborhoods */
422  SCIP_BANDIT* bandit; /**< bandit algorithm */
423  SCIP_SOL* lastcallsol; /**< incumbent when the heuristic was last called */
424  char* rewardfilename; /**< file name to store all rewards and the selection of the bandit */
425  FILE* rewardfile; /**< reward file pointer, or NULL */
426  SCIP_Longint nodesoffset; /**< offset added to the nodes budget */
427  SCIP_Longint maxnodes; /**< maximum number of nodes in a single sub-SCIP */
428  SCIP_Longint targetnodes; /**< targeted number of nodes to start a sub-SCIP */
429  SCIP_Longint minnodes; /**< minimum number of nodes required to start a sub-SCIP */
430  SCIP_Longint usednodes; /**< total number of nodes already spent in sub-SCIPs */
431  SCIP_Longint waitingnodes; /**< number of nodes since last incumbent solution that the heuristic should wait */
432  SCIP_Real nodesquot; /**< fraction of nodes compared to the main SCIP for budget computation */
433  SCIP_Real nodesquotmin; /**< lower bound on fraction of nodes compared to the main SCIP for budget computation */
434  SCIP_Real startminimprove; /**< initial factor by which ALNS should at least improve the incumbent */
435  SCIP_Real minimprovelow; /**< lower threshold for the minimal improvement over the incumbent */
436  SCIP_Real minimprovehigh; /**< upper bound for the minimal improvement over the incumbent */
437  SCIP_Real minimprove; /**< factor by which ALNS should at least improve the incumbent */
438  SCIP_Real lplimfac; /**< limit fraction of LPs per node to interrupt sub-SCIP */
439  SCIP_Real exp3_gamma; /**< weight between uniform (gamma ~ 1) and weight driven (gamma ~ 0) probability distribution for exp3 */
440  SCIP_Real exp3_beta; /**< reward offset between 0 and 1 at every observation for exp3 */
441  SCIP_Real epsgreedy_eps; /**< increase exploration in epsilon-greedy bandit algorithm */
442  SCIP_Real ucb_alpha; /**< parameter to increase the confidence width in UCB */
443  SCIP_Real rewardcontrol; /**< reward control to increase the weight of the simple solution indicator
444  * and decrease the weight of the closed gap reward */
445  SCIP_Real targetnodefactor; /**< factor by which target node number is eventually increased */
446  SCIP_Real rewardbaseline; /**< the reward baseline to separate successful and failed calls */
447  SCIP_Real fixtol; /**< tolerance by which the fixing rate may be missed without generic fixing */
448  SCIP_Real unfixtol; /**< tolerance by which the fixing rate may be exceeded without generic unfixing */
449  int nneighborhoods; /**< number of neighborhoods */
450  int nactiveneighborhoods;/**< number of active neighborhoods */
451  int ninitneighborhoods; /**< neighborhoods that were used at least one time */
452  int nsolslim; /**< limit on the number of improving solutions in a sub-SCIP call */
453  int seed; /**< initial random seed for bandit algorithms and random decisions by neighborhoods */
454  int currneighborhood; /**< index of currently selected neighborhood */
455  int ndelayedcalls; /**< the number of delayed calls */
456  int maxcallssamesol; /**< number of allowed executions of the heuristic on the same incumbent solution
457  * (-1: no limit, 0: number of active neighborhoods) */
458  SCIP_Longint firstcallthissol; /**< counter for the number of calls on this incumbent */
459  char banditalgo; /**< the bandit algorithm: (u)pper confidence bounds, (e)xp.3, epsilon (g)reedy */
460  SCIP_Bool useredcost; /**< should reduced cost scores be used for variable prioritization? */
461  SCIP_Bool usedistances; /**< should distances from fixed variables be used for variable prioritization */
462  SCIP_Bool usepscost; /**< should pseudo cost scores be used for variable prioritization? */
463  SCIP_Bool domorefixings; /**< should the ALNS heuristic do more fixings by itself based on variable prioritization
464  * until the target fixing rate is reached? */
465  SCIP_Bool adjustfixingrate; /**< should the heuristic adjust the target fixing rate based on the success? */
466  SCIP_Bool usesubscipheurs; /**< should the heuristic activate other sub-SCIP heuristics during its search? */
467  SCIP_Bool adjustminimprove; /**< should the factor by which the minimum improvement is bound be dynamically updated? */
468  SCIP_Bool adjusttargetnodes; /**< should the target nodes be dynamically adjusted? */
469  SCIP_Bool resetweights; /**< should the bandit algorithms be reset when a new problem is read? */
470  SCIP_Bool subsciprandseeds; /**< should random seeds of sub-SCIPs be altered to increase diversification? */
471  SCIP_Bool scalebyeffort; /**< should the reward be scaled by the effort? */
472  SCIP_Bool copycuts; /**< should cutting planes be copied to the sub-SCIP? */
473  SCIP_Bool uselocalredcost; /**< should local reduced costs be used for generic (un)fixing? */
474  SCIP_Bool initduringroot; /**< should the heuristic be executed multiple times during the root node? */
475  SCIP_Bool shownbstats; /**< show statistics on neighborhoods? */
476 };
477 
478 /** event handler data */
479 struct SCIP_EventData
480 {
481  SCIP_VAR** subvars; /**< the variables of the subproblem */
482  SCIP* sourcescip; /**< original SCIP data structure */
483  SCIP_HEUR* heur; /**< alns heuristic structure */
484  SCIP_Longint nodelimit; /**< node limit of the run */
485  SCIP_Real lplimfac; /**< limit fraction of LPs per node to interrupt sub-SCIP */
486  NH_STATS* runstats; /**< run statistics for the current neighborhood */
487  SCIP_Bool allrewardsmode; /**< true if solutions should only be checked for reward comparisons */
488 };
489 
490 /** represents limits for the sub-SCIP solving process */
491 struct SolveLimits
492 {
493  SCIP_Longint nodelimit; /**< maximum number of solving nodes for the sub-SCIP */
494  SCIP_Real memorylimit; /**< memory limit for the sub-SCIP */
495  SCIP_Real timelimit; /**< time limit for the sub-SCIP */
496  SCIP_Longint stallnodes; /**< maximum number of nodes without (primal) stalling */
497 };
498 
499 typedef struct SolveLimits SOLVELIMITS;
501 /** data structure that can be used for variable prioritization for additional fixings */
502 struct VarPrio
503 {
504  SCIP* scip; /**< SCIP data structure */
505  SCIP_Real* randscores; /**< random scores for prioritization */
506  int* distances; /**< breadth-first distances from already fixed variables */
507  SCIP_Real* redcostscores; /**< reduced cost scores for fixing a variable to a reference value */
508  SCIP_Real* pscostscores; /**< pseudocost scores for fixing a variable to a reference value */
509  unsigned int useredcost:1; /**< should reduced cost scores be used for variable prioritization? */
510  unsigned int usedistances:1; /**< should distances from fixed variables be used for variable prioritization */
511  unsigned int usepscost:1; /**< should pseudo cost scores be used for variable prioritization? */
512 };
513 
514 /*
515  * Local methods
516  */
517 
518 /** Reset target fixing rate */
519 static
521  SCIP* scip, /**< SCIP data structure */
522  NH_FIXINGRATE* fixingrate /**< heuristic fixing rate */
523  )
524 {
525  assert(scip != NULL);
526  assert(fixingrate != NULL);
527  fixingrate->increment = FIXINGRATE_STARTINC;
528 
529  /* always start with the most conservative value */
530  fixingrate->targetfixingrate = fixingrate->maxfixingrate;
531 
532  return SCIP_OKAY;
533 }
534 
535 /** reset the currently active neighborhood */
536 static
538  SCIP_HEURDATA* heurdata
539  )
540 {
541  assert(heurdata != NULL);
542  heurdata->currneighborhood = -1;
543  heurdata->ndelayedcalls = 0;
544 }
545 
546 /** update increment for fixing rate */
547 static
549  NH_FIXINGRATE* fx /**< fixing rate */
550  )
551 {
553  fx->increment = MAX(fx->increment, LRATEMIN);
554 }
555 
556 
557 /** increase fixing rate
558  *
559  * decrease also the rate by which the target fixing rate is adjusted
560  */
561 static
562 void increaseFixingRate(
563  NH_FIXINGRATE* fx /**< fixing rate */
564  )
565 {
566  fx->targetfixingrate += fx->increment;
567  fx->targetfixingrate = MIN(fx->targetfixingrate, fx->maxfixingrate);
568 }
569 
570 /** decrease fixing rate
571  *
572  * decrease also the rate by which the target fixing rate is adjusted
573  */
574 static
575 void decreaseFixingRate(
576  NH_FIXINGRATE* fx /**< fixing rate */
577  )
578 {
579  fx->targetfixingrate -= fx->increment;
581 }
582 
583 /** update fixing rate based on the results of the current run */
584 static
585 void updateFixingRate(
586  NH* neighborhood, /**< neighborhood */
587  SCIP_STATUS subscipstatus, /**< status of the sub-SCIP run */
588  NH_STATS* runstats /**< run statistics for this run */
589  )
590 {
591  NH_FIXINGRATE* fx;
592 
593  fx = &neighborhood->fixingrate;
594 
595  switch (subscipstatus)
596  {
597  case SCIP_STATUS_OPTIMAL:
602  /* decrease the fixing rate (make subproblem harder) */
603  decreaseFixingRate(fx);
604  break;
609  /* increase the fixing rate (make the subproblem easier) only if no solution was found */
610  if( runstats->nbestsolsfound <= 0 )
611  increaseFixingRate(fx);
612  break;
613  /* fall through cases to please lint */
614  case SCIP_STATUS_UNKNOWN:
621  default:
622  break;
623  }
624 
626 }
627 
628 /** increase target node limit */
629 static
631  SCIP_HEURDATA* heurdata /**< heuristic data */
632  )
633 {
634  heurdata->targetnodes = (SCIP_Longint)(heurdata->targetnodes * heurdata->targetnodefactor) + 1;
635 
636  /* respect upper and lower parametrized bounds on targetnodes */
637  if( heurdata->targetnodes > heurdata->maxnodes )
638  heurdata->targetnodes = heurdata->maxnodes;
639 }
640 
641 /** reset target node limit */
642 static
644  SCIP_HEURDATA* heurdata /**< heuristic data */
645  )
646 {
647  heurdata->targetnodes = heurdata->minnodes;
648 }
649 
650 /** update target node limit based on the current run results */
651 static
653  SCIP_HEURDATA* heurdata, /**< heuristic data */
654  NH_STATS* runstats, /**< statistics of the run */
655  SCIP_STATUS subscipstatus /**< status of the sub-SCIP run */
656  )
657 {
658  switch (subscipstatus)
659  {
662  /* the subproblem could be explored more */
663  if( runstats->nbestsolsfound == 0 )
664  increaseTargetNodeLimit(heurdata);
665  break;
666  case SCIP_STATUS_OPTIMAL:
671  break;
674  case SCIP_STATUS_UNKNOWN:
681  break;
682  default:
683  break;
684  }
685 }
686 
687 /** reset the minimum improvement for the sub-SCIPs */
688 static
690  SCIP_HEURDATA* heurdata /**< heuristic data */
691  )
692 {
693  assert(heurdata != NULL);
694  heurdata->minimprove = heurdata->startminimprove;
695 }
696 
697 /** increase minimum improvement for the sub-SCIPs */
698 static
700  SCIP_HEURDATA* heurdata /**< heuristic data */
701  )
702 {
703  assert(heurdata != NULL);
704 
705  heurdata->minimprove *= MINIMPROVEFAC;
706  heurdata->minimprove = MIN(heurdata->minimprove, heurdata->minimprovehigh);
707 }
708 
709 /** decrease the minimum improvement for the sub-SCIPs */
710 static
712  SCIP_HEURDATA* heurdata /**< heuristic data */
713  )
714 {
715  assert(heurdata != NULL);
716 
717  heurdata->minimprove /= MINIMPROVEFAC;
718  SCIPdebugMessage("%.4f", heurdata->minimprovelow);
719  heurdata->minimprove = MAX(heurdata->minimprove, heurdata->minimprovelow);
720 }
721 
722 /** update the minimum improvement based on the status of the sub-SCIP */
723 static
725  SCIP_HEURDATA* heurdata, /**< heuristic data */
726  SCIP_STATUS subscipstatus, /**< status of the sub-SCIP run */
727  NH_STATS* runstats /**< run statistics for this run */
728  )
729 {
730  assert(heurdata != NULL);
731 
732  /* if the sub-SCIP status was infeasible, we rather want to make the sub-SCIP easier
733  * with a smaller minimum improvement.
734  *
735  * If a solution limit was reached, we may, set it higher.
736  */
737  switch (subscipstatus)
738  {
741  /* subproblem was infeasible, probably due to the minimum improvement -> decrease minimum improvement */
742  decreaseMinimumImprovement(heurdata);
743 
744  break;
747  case SCIP_STATUS_OPTIMAL:
748  /* subproblem could be optimally solved -> try higher minimum improvement */
749  increaseMinimumImprovement(heurdata);
750  break;
754  /* subproblem was too hard, decrease minimum improvement */
755  if( runstats->nbestsolsfound <= 0 )
756  decreaseMinimumImprovement(heurdata);
757  break;
758  case SCIP_STATUS_UNKNOWN:
766  default:
767  break;
768  }
769 }
770 
771 /** Reset neighborhood statistics */
772 static
774  SCIP* scip, /**< SCIP data structure */
775  NH_STATS* stats /**< neighborhood statistics */
776  )
777 {
778  assert(scip != NULL);
779  assert(stats != NULL);
780 
781  stats->nbestsolsfound = 0;
782  stats->nruns = 0;
783  stats->nrunsbestsol = 0;
784  stats->nsolsfound = 0;
785  stats->usednodes = 0L;
786  stats->nfixings = 0L;
787 
789 
790  SCIP_CALL( SCIPresetClock(scip, stats->setupclock) );
791  SCIP_CALL( SCIPresetClock(scip, stats->submipclock) );
792 
793  return SCIP_OKAY;
794 }
795 
796 /** create a neighborhood of the specified name and include it into the ALNS heuristic */
797 static
799  SCIP* scip, /**< SCIP data structure */
800  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS heuristic */
801  NH** neighborhood, /**< pointer to store the neighborhood */
802  const char* name, /**< name for this neighborhood */
803  SCIP_Real minfixingrate, /**< default value for minfixingrate parameter of this neighborhood */
804  SCIP_Real maxfixingrate, /**< default value for maxfixingrate parameter of this neighborhood */
805  SCIP_Bool active, /**< default value for active parameter of this neighborhood */
806  SCIP_Real priority, /**< positive call priority to initialize bandit algorithms */
807  DECL_VARFIXINGS ((*varfixings)), /**< variable fixing callback for this neighborhood, or NULL */
808  DECL_CHANGESUBSCIP ((*changesubscip)), /**< subscip changes callback for this neighborhood, or NULL */
809  DECL_NHINIT ((*nhinit)), /**< initialization callback for neighborhood, or NULL */
810  DECL_NHEXIT ((*nhexit)), /**< deinitialization callback for neighborhood, or NULL */
811  DECL_NHFREE ((*nhfree)), /**< deinitialization callback before SCIP is freed, or NULL */
812  DECL_NHREFSOL ((*nhrefsol)), /**< callback function to return a reference solution for further fixings, or NULL */
813  DECL_NHDEACTIVATE ((*nhdeactivate)) /**< callback function to deactivate neighborhoods on problems where they are irrelevant, or NULL if neighborhood is always active */
814  )
815 {
817 
818  assert(scip != NULL);
819  assert(heurdata != NULL);
820  assert(neighborhood != NULL);
821  assert(name != NULL);
822 
823  SCIP_CALL( SCIPallocBlockMemory(scip, neighborhood) );
824  assert(*neighborhood != NULL);
825 
826  SCIP_ALLOC( BMSduplicateMemoryArray(&(*neighborhood)->name, name, strlen(name)+1) );
827 
828  SCIP_CALL( SCIPcreateClock(scip, &(*neighborhood)->stats.setupclock) );
829  SCIP_CALL( SCIPcreateClock(scip, &(*neighborhood)->stats.submipclock) );
830 
831  (*neighborhood)->changesubscip = changesubscip;
832  (*neighborhood)->varfixings = varfixings;
833  (*neighborhood)->nhinit = nhinit;
834  (*neighborhood)->nhexit = nhexit;
835  (*neighborhood)->nhfree = nhfree;
836  (*neighborhood)->nhrefsol = nhrefsol;
837  (*neighborhood)->nhdeactivate = nhdeactivate;
838 
839  /* add parameters for this neighborhood */
840  (void) SCIPsnprintf(paramname, SCIP_MAXSTRLEN, "heuristics/alns/%s/minfixingrate", name);
841  SCIP_CALL( SCIPaddRealParam(scip, paramname, "minimum fixing rate for this neighborhood",
842  &(*neighborhood)->fixingrate.minfixingrate, TRUE, minfixingrate, 0.0, 1.0, NULL, NULL) );
843  (void) SCIPsnprintf(paramname, SCIP_MAXSTRLEN, "heuristics/alns/%s/maxfixingrate", name);
844  SCIP_CALL( SCIPaddRealParam(scip, paramname, "maximum fixing rate for this neighborhood",
845  &(*neighborhood)->fixingrate.maxfixingrate, TRUE, maxfixingrate, 0.0, 1.0, NULL, NULL) );
846  (void) SCIPsnprintf(paramname, SCIP_MAXSTRLEN, "heuristics/alns/%s/active", name);
847  SCIP_CALL( SCIPaddBoolParam(scip, paramname, "is this neighborhood active?",
848  &(*neighborhood)->active, TRUE, active, NULL, NULL) );
849  (void) SCIPsnprintf(paramname, SCIP_MAXSTRLEN, "heuristics/alns/%s/priority", name);
850  SCIP_CALL( SCIPaddRealParam(scip, paramname, "positive call priority to initialize bandit algorithms",
851  &(*neighborhood)->priority, TRUE, priority, 1e-2, 1.0, NULL, NULL) );
852 
853  /* add the neighborhood to the ALNS heuristic */
854  heurdata->neighborhoods[heurdata->nneighborhoods++] = (*neighborhood);
855 
856  return SCIP_OKAY;
857 }
858 
859 /** release all data and free neighborhood */
860 static
862  SCIP* scip, /**< SCIP data structure */
863  NH** neighborhood /**< pointer to neighborhood that should be freed */
864  )
865 {
866  NH* nhptr;
867  assert(scip != NULL);
868  assert(neighborhood != NULL);
869 
870  nhptr = *neighborhood;
871  assert(nhptr != NULL);
872 
873  BMSfreeMemoryArray(&nhptr->name);
874 
875  /* release further, neighborhood specific data structures */
876  if( nhptr->nhfree != NULL )
877  {
878  SCIP_CALL( nhptr->nhfree(scip, nhptr) );
879  }
880 
881  SCIP_CALL( SCIPfreeClock(scip, &nhptr->stats.setupclock) );
882  SCIP_CALL( SCIPfreeClock(scip, &nhptr->stats.submipclock) );
883 
884  SCIPfreeBlockMemory(scip, neighborhood);
885  *neighborhood = NULL;
886 
887  return SCIP_OKAY;
888 }
889 
890 /** initialize neighborhood specific data */
891 static
893  SCIP* scip, /**< SCIP data structure */
894  NH* neighborhood /**< neighborhood to initialize */
895  )
896 {
897  assert(scip != NULL);
898  assert(neighborhood != NULL);
899 
900  /* call the init callback of the neighborhood */
901  if( neighborhood->nhinit != NULL )
902  {
903  SCIP_CALL( neighborhood->nhinit(scip, neighborhood) );
904  }
905 
906  return SCIP_OKAY;
907 }
908 
909 /** deinitialize neighborhood specific data */
910 static
912  SCIP* scip, /**< SCIP data structure */
913  NH* neighborhood /**< neighborhood to initialize */
914  )
915 {
916  assert(scip != NULL);
917  assert(neighborhood != NULL);
918 
919  if( neighborhood->nhexit != NULL )
920  {
921  SCIP_CALL( neighborhood->nhexit(scip, neighborhood) );
922  }
923 
924  return SCIP_OKAY;
925 }
926 
927 /** creates a new solution for the original problem by copying the solution of the subproblem */
928 static
930  SCIP* subscip, /**< SCIP data structure of the subproblem */
931  SCIP_EVENTDATA* eventdata /**< event handler data */
932  )
933 {
934  SCIP* sourcescip; /* original SCIP data structure */
935  SCIP_VAR** subvars; /* the variables of the subproblem */
936  SCIP_HEUR* heur; /* alns heuristic structure */
937  SCIP_SOL* subsol; /* solution of the subproblem */
938  SCIP_SOL* newsol; /* solution to be created for the original problem */
939  SCIP_Bool success;
940  NH_STATS* runstats;
941  SCIP_SOL* oldbestsol;
942 
943  assert(subscip != NULL);
944 
945  subsol = SCIPgetBestSol(subscip);
946  assert(subsol != NULL);
947 
948  sourcescip = eventdata->sourcescip;
949  subvars = eventdata->subvars;
950  heur = eventdata->heur;
951  runstats = eventdata->runstats;
952  assert(sourcescip != NULL);
953  assert(sourcescip != subscip);
954  assert(heur != NULL);
955  assert(subvars != NULL);
956  assert(runstats != NULL);
957 
958  SCIP_CALL( SCIPtranslateSubSol(sourcescip, subscip, subsol, heur, subvars, &newsol) );
959 
960  oldbestsol = SCIPgetBestSol(sourcescip);
961 
962  /* in the special, experimental all rewards mode, the solution is only checked for feasibility
963  * but not stored
964  */
965  if( eventdata->allrewardsmode )
966  {
967  SCIP_CALL( SCIPcheckSol(sourcescip, newsol, FALSE, FALSE, TRUE, TRUE, TRUE, &success) );
968 
969  if( success )
970  {
971  runstats->nsolsfound++;
972  if( SCIPgetSolTransObj(sourcescip, newsol) < SCIPgetCutoffbound(sourcescip) )
973  runstats->nbestsolsfound++;
974  }
975 
976  SCIP_CALL( SCIPfreeSol(sourcescip, &newsol) );
977  }
978  else
979  {
980  /* try to add new solution to scip and free it immediately */
981  SCIP_CALL( SCIPtrySolFree(sourcescip, &newsol, FALSE, FALSE, TRUE, TRUE, TRUE, &success) );
982 
983  if( success )
984  {
985  runstats->nsolsfound++;
986  if( SCIPgetBestSol(sourcescip) != oldbestsol )
987  runstats->nbestsolsfound++;
988  }
989  }
990 
991  /* update new upper bound for reward later */
992  runstats->newupperbound = SCIPgetUpperbound(sourcescip);
993 
994  return SCIP_OKAY;
995 }
996 
997 
998 /* ---------------- Callback methods of event handler ---------------- */
999 
1000 /** execution callback of the event handler
1001  *
1002  * transfer new solutions or interrupt the solving process manually
1003  */
1004 static
1005 SCIP_DECL_EVENTEXEC(eventExecAlns)
1007  assert(eventhdlr != NULL);
1008  assert(eventdata != NULL);
1009  assert(strcmp(SCIPeventhdlrGetName(eventhdlr), EVENTHDLR_NAME) == 0);
1010  assert(event != NULL);
1011  assert(SCIPeventGetType(event) & SCIP_EVENTTYPE_ALNS);
1012  assert(eventdata != NULL);
1013 
1014  /* treat the different atomic events */
1015  switch( SCIPeventGetType(event) )
1016  {
1019  /* try to transfer the solution to the original SCIP */
1020  SCIP_CALL( transferSolution(scip, eventdata) );
1021  break;
1023  /* interrupt solution process of sub-SCIP */
1024  if( SCIPgetNLPs(scip) > eventdata->lplimfac * eventdata->nodelimit )
1025  {
1026  SCIPdebugMsg(scip, "interrupt after %" SCIP_LONGINT_FORMAT " LPs\n", SCIPgetNLPs(scip));
1027  SCIP_CALL( SCIPinterruptSolve(scip) );
1028  }
1029  break;
1030  default:
1031  break;
1032  }
1033 
1034  return SCIP_OKAY;
1035 }
1036 
1037 /** initialize neighborhood statistics before the next run */
1038 static
1039 void initRunStats(
1040  SCIP* scip, /**< SCIP data structure */
1041  NH_STATS* stats /**< run statistics */
1042  )
1043 {
1044  stats->nbestsolsfound = 0;
1045  stats->nsolsfound = 0;
1046  stats->usednodes = 0L;
1047  stats->nfixings = 0;
1048  stats->oldupperbound = SCIPgetUpperbound(scip);
1049  stats->newupperbound = SCIPgetUpperbound(scip);
1050 }
1051 
1052 /** update run stats after the sub SCIP was solved */
1053 static
1054 void updateRunStats(
1055  NH_STATS* stats, /**< run statistics */
1056  SCIP* subscip /**< sub-SCIP instance, or NULL */
1057  )
1058 {
1059  /* treat an untransformed subscip as if none was created */
1060  if( subscip != NULL && ! SCIPisTransformed(subscip) )
1061  subscip = NULL;
1062 
1063  stats->usednodes = subscip != NULL ? SCIPgetNNodes(subscip) : 0L;
1064 }
1065 
1066 /** get the histogram index for this status */
1067 static
1068 int getHistIndex(
1069  SCIP_STATUS subscipstatus /**< sub-SCIP status */
1070  )
1071 {
1072  switch (subscipstatus)
1073  {
1074  case SCIP_STATUS_OPTIMAL:
1075  return (int)HIDX_OPT;
1077  return (int)HIDX_INFEAS;
1078  case SCIP_STATUS_NODELIMIT:
1079  return (int)HIDX_NODELIM;
1081  return (int)HIDX_STALLNODE;
1082  case SCIP_STATUS_SOLLIMIT:
1084  return (int)HIDX_SOLLIM;
1086  return (int)HIDX_USR;
1087  default:
1088  return (int)HIDX_OTHER;
1089  } /*lint !e788*/
1090 }
1091 
1092 /** print neighborhood statistics */
1093 static
1095  SCIP* scip, /**< SCIP data structure */
1096  SCIP_HEURDATA* heurdata, /**< heuristic data */
1097  FILE* file /**< file handle, or NULL for standard out */
1098  )
1099 {
1100  int i;
1101  int j;
1103 
1104  if( ! heurdata->shownbstats )
1105  return;
1106 
1107  SCIPinfoMessage(scip, file, "Neighborhoods : %10s %10s %10s %10s %10s %10s %10s %10s %10s %10s %4s %4s %4s %4s %4s %4s %4s %4s\n",
1108  "Calls", "SetupTime", "SolveTime", "SolveNodes", "Sols", "Best", "Exp3", "EpsGreedy", "UCB", "TgtFixRate",
1109  "Opt", "Inf", "Node", "Stal", "Sol", "Usr", "Othr", "Actv");
1110 
1111  /* loop over neighborhoods and fill in statistics */
1112  for( i = 0; i < heurdata->nneighborhoods; ++i )
1113  {
1114  NH* neighborhood;
1115  SCIP_Real proba;
1116  SCIP_Real ucb;
1117  SCIP_Real epsgreedyweight;
1118 
1119  neighborhood = heurdata->neighborhoods[i];
1120  SCIPinfoMessage(scip, file, " %-17s:", neighborhood->name);
1121  SCIPinfoMessage(scip, file, " %10d", neighborhood->stats.nruns);
1122  SCIPinfoMessage(scip, file, " %10.2f", SCIPgetClockTime(scip, neighborhood->stats.setupclock) );
1123  SCIPinfoMessage(scip, file, " %10.2f", SCIPgetClockTime(scip, neighborhood->stats.submipclock) );
1124  SCIPinfoMessage(scip, file, " %10" SCIP_LONGINT_FORMAT, neighborhood->stats.usednodes );
1125  SCIPinfoMessage(scip, file, " %10" SCIP_LONGINT_FORMAT, neighborhood->stats.nsolsfound);
1126  SCIPinfoMessage(scip, file, " %10" SCIP_LONGINT_FORMAT, neighborhood->stats.nbestsolsfound);
1127 
1128  proba = 0.0;
1129  ucb = 1.0;
1130  epsgreedyweight = -1.0;
1131 
1132  if( heurdata->bandit != NULL && i < heurdata->nactiveneighborhoods )
1133  {
1134  switch (heurdata->banditalgo)
1135  {
1136  case 'u':
1137  ucb = SCIPgetConfidenceBoundUcb(heurdata->bandit, i);
1138  break;
1139  case 'g':
1140  epsgreedyweight = SCIPgetWeightsEpsgreedy(heurdata->bandit)[i];
1141  break;
1142  case 'e':
1143  proba = SCIPgetProbabilityExp3(heurdata->bandit, i);
1144  break;
1145  default:
1146  break;
1147  }
1148  }
1149 
1150  SCIPinfoMessage(scip, file, " %10.5f", proba);
1151  SCIPinfoMessage(scip, file, " %10.5f", epsgreedyweight);
1152  SCIPinfoMessage(scip, file, " %10.5f", ucb);
1153  SCIPinfoMessage(scip, file, " %10.3f", neighborhood->fixingrate.targetfixingrate);
1154 
1155  /* loop over status histogram */
1156  for( j = 0; j < NHISTENTRIES; ++j )
1157  SCIPinfoMessage(scip, file, " %4d", neighborhood->stats.statushist[statusses[j]]);
1158 
1159  SCIPinfoMessage(scip, file, " %4d", i < heurdata->nactiveneighborhoods ? 1 : 0);
1160  SCIPinfoMessage(scip, file, "\n");
1161  }
1162 }
1163 
1164 /** update the statistics of the neighborhood based on the sub-SCIP run */
1165 static
1167  NH_STATS* runstats, /**< run statistics */
1168  NH* neighborhood, /**< the selected neighborhood */
1169  SCIP_STATUS subscipstatus /**< status of the sub-SCIP solve */
1170  )
1171 { /*lint --e{715}*/
1172  NH_STATS* stats;
1173  stats = &neighborhood->stats;
1174 
1175  /* copy run statistics into neighborhood statistics */
1176  stats->nbestsolsfound += runstats->nbestsolsfound;
1177  stats->nsolsfound += runstats->nsolsfound;
1178  stats->usednodes += runstats->usednodes;
1179  stats->nruns += 1;
1180 
1181  if( runstats->nbestsolsfound > 0 )
1183  else if( runstats->nsolsfound > 0 )
1184  stats->nrunsbestsol++;
1185 
1186  /* update the counter for the subscip status */
1187  ++stats->statushist[getHistIndex(subscipstatus)];
1188 }
1189 
1190 /** sort callback for variable pointers using the ALNS variable prioritization
1191  *
1192  * the variable prioritization works hierarchically as follows. A variable
1193  * a has the higher priority over b iff
1194  *
1195  * - variable distances should be used and a has a smaller distance than b
1196  * - variable reduced costs should be used and a has a smaller score than b
1197  * - variable pseudo costs should be used and a has a smaller score than b
1198  * - based on previously assigned random scores
1199  *
1200  * @note: distances are context-based. For fixing more variables,
1201  * distances are initialized from the already fixed variables.
1202  * For unfixing variables, distances are initialized starting
1203  * from the unfixed variables
1204  */
1205 static
1206 SCIP_DECL_SORTINDCOMP(sortIndCompAlns)
1207 { /*lint --e{715}*/
1208  VARPRIO* varprio;
1209 
1210  varprio = (VARPRIO*)dataptr;
1211  assert(varprio != NULL);
1212  assert(varprio->randscores != NULL);
1213 
1214  if( ind1 == ind2 )
1215  return 0;
1216 
1217  /* priority is on distances, if enabled. The variable which is closer in a breadth-first search sense to
1218  * the already fixed variables has precedence */
1219  if( varprio->usedistances )
1220  {
1221  int dist1;
1222  int dist2;
1223 
1224  dist1 = varprio->distances[ind1];
1225  dist2 = varprio->distances[ind2];
1226 
1227  if( dist1 < 0 )
1228  dist1 = INT_MAX;
1229 
1230  if( dist2 < 0 )
1231  dist2 = INT_MAX;
1232 
1233  assert(varprio->distances != NULL);
1234  if( dist1 < dist2 )
1235  return -1;
1236  else if( dist1 > dist2 )
1237  return 1;
1238  }
1239 
1240  assert(! varprio->usedistances || varprio->distances[ind1] == varprio->distances[ind2]);
1241 
1242  /* if the indices tie considering distances or distances are disabled -> use reduced cost information instead */
1243  if( varprio->useredcost )
1244  {
1245  assert(varprio->redcostscores != NULL);
1246 
1247  if( varprio->redcostscores[ind1] < varprio->redcostscores[ind2] )
1248  return -1;
1249  else if( varprio->redcostscores[ind1] > varprio->redcostscores[ind2] )
1250  return 1;
1251  }
1252 
1253  /* use pseudo cost scores if reduced costs are disabled or a tie was found */
1254  if( varprio->usepscost )
1255  {
1256  assert(varprio->pscostscores != NULL);
1257 
1258  /* prefer the variable with smaller pseudocost score */
1259  if( varprio->pscostscores[ind1] < varprio->pscostscores[ind2] )
1260  return -1;
1261  else if( varprio->pscostscores[ind1] > varprio->pscostscores[ind2] )
1262  return 1;
1263  }
1264 
1265  if( varprio->randscores[ind1] < varprio->randscores[ind2] )
1266  return -1;
1267  else if( varprio->randscores[ind1] > varprio->randscores[ind2] )
1268  return 1;
1269 
1270  return ind1 - ind2;
1271 }
1272 
1273 /** Compute the reduced cost score for this variable in the reference solution */
1274 static
1276  SCIP* scip, /**< SCIP data structure */
1277  SCIP_VAR* var, /**< the variable for which the score should be computed */
1278  SCIP_Real refsolval, /**< solution value in reference solution */
1279  SCIP_Bool uselocalredcost /**< should local reduced costs be used for generic (un)fixing? */
1280  )
1281 {
1282  SCIP_Real bestbound;
1283  SCIP_Real redcost;
1284  SCIP_Real score;
1285  assert(scip != NULL);
1286  assert(var != NULL);
1287 
1288  /* prefer column variables */
1290  return SCIPinfinity(scip);
1291 
1292  if( ! uselocalredcost )
1293  {
1294  redcost = SCIPvarGetBestRootRedcost(var);
1295 
1296  bestbound = SCIPvarGetBestRootSol(var);
1297 
1298  /* using global reduced costs, the two factors yield a nonnegative score within tolerances */
1299  assert(SCIPisDualfeasZero(scip, redcost)
1300  || (SCIPisDualfeasNegative(scip, redcost) && ! SCIPisFeasPositive(scip, refsolval - bestbound))
1301  || (SCIPisDualfeasPositive(scip, redcost) && ! SCIPisFeasNegative(scip, refsolval - bestbound)));
1302  }
1303  else
1304  {
1305  /* this can be safely asserted here, since the heuristic would not reach this point, otherwise */
1306  assert(SCIPhasCurrentNodeLP(scip));
1307  assert(SCIPgetLPSolstat(scip) == SCIP_LPSOLSTAT_OPTIMAL);
1308 
1309  redcost = SCIPgetVarRedcost(scip, var);
1310 
1311  bestbound = SCIPvarGetLPSol(var);
1312  }
1313 
1314  assert(! SCIPisInfinity(scip, REALABS(bestbound)));
1315  assert(SCIPisDualfeasZero(scip, redcost) || SCIPisFeasIntegral(scip, bestbound));
1316 
1317  score = redcost * (refsolval - bestbound);
1318 
1319  /* max out numerical inaccuracies from global scores */
1320  if( ! uselocalredcost )
1321  score = MAX(score, 0.0);
1322 
1323  return score;
1324 }
1325 
1326 /** get the pseudo cost score of this variable with respect to the reference solution */
1327 static
1329  SCIP* scip, /**< SCIP data structure */
1330  SCIP_VAR* var, /**< the variable for which the score should be computed */
1331  SCIP_Real refsolval, /**< solution value in reference solution */
1332  SCIP_Bool uselocallpsol /**< should local LP solution be used? */
1333  )
1334 {
1335  SCIP_Real lpsolval;
1336 
1337  assert(scip != NULL);
1338  assert(var != NULL);
1339 
1340  /* variables that aren't LP columns have no pseudocost score */
1342  return 0.0;
1343 
1344  lpsolval = uselocallpsol ? SCIPvarGetLPSol(var) : SCIPvarGetRootSol(var);
1345 
1346  /* the score is 0.0 if the values are equal */
1347  if( SCIPisEQ(scip, lpsolval, refsolval) )
1348  return 0.0;
1349  else
1350  return SCIPgetVarPseudocostVal(scip, var, refsolval - lpsolval);
1351 }
1352 
1353 /** add variable and solution value to buffer data structure for variable fixings. The method checks if
1354  * the value still lies within the variable bounds. The value stays unfixed otherwise.
1355  */
1356 static
1358  SCIP* scip, /**< SCIP data structure */
1359  SCIP_VAR* var, /**< (source) SCIP variable that should be added to the buffer */
1360  SCIP_Real val, /**< fixing value for this variable */
1361  SCIP_VAR** varbuf, /**< variable buffer to store variables that should be fixed */
1362  SCIP_Real* valbuf, /**< value buffer to store fixing values */
1363  int* nfixings, /**< pointer to number of fixed buffer variables, will be increased by 1 */
1364  SCIP_Bool integer /**< is this an integer variable? */
1365  )
1366 {
1367  /* todo: this assert can fail when there was a dual reduction that changed a variable to
1368  * an integral type after the reference solution was found and the variable has a fractional
1369  * value in this solution, e.g., for boxQP instances (spar*)
1370  * implicit integer variables could also be an issue, as they can take fractional values in feasible solutions
1371  */
1372  assert(SCIPisFeasIntegral(scip, val) || ! SCIPvarIsIntegral(var));
1373  assert(*nfixings < SCIPgetNVars(scip));
1374 
1375  /* round the value to its nearest integer */
1376  if( integer )
1377  val = SCIPfloor(scip, val + 0.5);
1378 
1379  /* only add fixing if it is still valid within the global variable bounds. Invalidity
1380  * of this solution value may come from a dual reduction that was performed after the solution from which
1381  * this value originated was found
1382  */
1383  if( SCIPvarGetLbGlobal(var) <= val && val <= SCIPvarGetUbGlobal(var) )
1384  {
1385  varbuf[*nfixings] = var;
1386  valbuf[*nfixings] = val;
1387  ++(*nfixings);
1388  }
1389 }
1390 
1391 /** query neighborhood for a reference solution for further fixings */
1392 static
1394  SCIP* scip, /**< SCIP data structure */
1395  NH* neighborhood, /**< ALNS neighborhood data structure */
1396  SCIP_SOL** solptr /**< solution pointer */
1397  )
1398 {
1399  assert(solptr != NULL);
1400  assert(scip != NULL);
1401  assert(neighborhood != NULL);
1402 
1403  *solptr = NULL;
1404  if( neighborhood->nhrefsol != NULL )
1405  {
1406  SCIP_RESULT result;
1407  SCIP_CALL( neighborhood->nhrefsol(scip, neighborhood, solptr, &result) );
1408 
1409  if( result == SCIP_DIDNOTFIND )
1410  *solptr = NULL;
1411  else
1412  assert(*solptr != NULL);
1413  }
1414 
1415  return SCIP_OKAY;
1416 }
1417 
1418 /** fix additional variables found in feasible reference solution if the ones that the neighborhood found were not enough
1419  *
1420  * use not always the best solution for the values, but a reference solution provided by the neighborhood itself
1421  *
1422  * @note it may happen that the target fixing rate is not completely reached. This is the case if intermediate,
1423  * dual reductions render the solution values of the reference solution infeasible for
1424  * the current, global variable bounds.
1425  */
1426 static
1428  SCIP* scip, /**< SCIP data structure */
1429  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
1430  SCIP_SOL* refsol, /**< feasible reference solution for more variable fixings */
1431  SCIP_VAR** varbuf, /**< buffer array to store variables to fix */
1432  SCIP_Real* valbuf, /**< buffer array to store fixing values */
1433  int* nfixings, /**< pointer to store the number of fixings */
1434  int ntargetfixings, /**< number of required target fixings */
1435  SCIP_Bool* success /**< pointer to store whether the target fixings have been successfully reached */
1436  )
1437 {
1438  VARPRIO varprio;
1439  SCIP_VAR** vars;
1440  SCIP_Real* redcostscores;
1441  SCIP_Real* pscostscores;
1442  SCIP_Real* solvals;
1443  SCIP_RANDNUMGEN* rng;
1444  SCIP_VAR** unfixedvars;
1445  SCIP_Bool* isfixed;
1446  int* distances;
1447  int* perm;
1448  SCIP_Real* randscores;
1449  int nbinvars;
1450  int nintvars;
1451  int nbinintvars;
1452  int nvars;
1453  int b;
1454  int nvarstoadd;
1455  int nunfixedvars;
1456 
1457  assert(scip != NULL);
1458  assert(varbuf != NULL);
1459  assert(nfixings != NULL);
1460  assert(success != NULL);
1461  assert(heurdata != NULL);
1462  assert(refsol != NULL);
1463 
1464  *success = FALSE;
1465 
1466  /* if the user parameter forbids more fixings, return immediately */
1467  if( ! heurdata->domorefixings )
1468  return SCIP_OKAY;
1469 
1470  SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, &nbinvars, &nintvars, NULL, NULL) );
1471 
1472  nbinintvars = nbinvars + nintvars;
1473 
1474  if( ntargetfixings >= nbinintvars )
1475  return SCIP_OKAY;
1476 
1477  /* determine the number of required additional fixings */
1478  nvarstoadd = ntargetfixings - *nfixings;
1479  if( nvarstoadd == 0 )
1480  return SCIP_OKAY;
1481 
1482  varprio.usedistances = heurdata->usedistances && (*nfixings >= 1);
1483  varprio.useredcost = heurdata->useredcost;
1484  varprio.usepscost = heurdata->usepscost;
1485  varprio.scip = scip;
1486  rng = SCIPbanditGetRandnumgen(heurdata->bandit);
1487  assert(rng != NULL);
1488 
1489  SCIP_CALL( SCIPallocBufferArray(scip, &randscores, nbinintvars) );
1490  SCIP_CALL( SCIPallocBufferArray(scip, &perm, nbinintvars) );
1491  SCIP_CALL( SCIPallocBufferArray(scip, &distances, nvars) );
1492  SCIP_CALL( SCIPallocBufferArray(scip, &redcostscores, nbinintvars) );
1493  SCIP_CALL( SCIPallocBufferArray(scip, &solvals, nbinintvars) );
1494  SCIP_CALL( SCIPallocBufferArray(scip, &isfixed, nbinintvars) );
1495  SCIP_CALL( SCIPallocBufferArray(scip, &unfixedvars, nbinintvars) );
1496  SCIP_CALL( SCIPallocBufferArray(scip, &pscostscores, nbinintvars) );
1497 
1498  /* initialize variable graph distances from already fixed variables */
1499  if( varprio.usedistances )
1500  {
1501  SCIP_CALL( SCIPvariablegraphBreadthFirst(scip, NULL, varbuf, *nfixings, distances, INT_MAX, INT_MAX, ntargetfixings) );
1502  }
1503  else
1504  {
1505  /* initialize all equal distances to make them irrelevant */
1506  BMSclearMemoryArray(distances, nbinintvars);
1507  }
1508 
1509  BMSclearMemoryArray(isfixed, nbinintvars);
1510 
1511  /* mark binary and integer variables if they are fixed */
1512  for( b = 0; b < *nfixings; ++b )
1513  {
1514  int probindex;
1515 
1516  assert(varbuf[b] != NULL);
1517  probindex = SCIPvarGetProbindex(varbuf[b]);
1518  assert(probindex >= 0);
1519 
1520  if( probindex < nbinintvars )
1521  isfixed[probindex] = TRUE;
1522  }
1523 
1524  SCIP_CALL( SCIPgetSolVals(scip, refsol, nbinintvars, vars, solvals) );
1525 
1526  /* assign scores to unfixed every discrete variable of the problem */
1527  nunfixedvars = 0;
1528  for( b = 0; b < nbinintvars; ++b )
1529  {
1530  SCIP_VAR* var = vars[b];
1531 
1532  /* filter fixed variables */
1533  if( isfixed[b] )
1534  continue;
1535 
1536  /* filter variables with a solution value outside its global bounds */
1537  if( solvals[b] < SCIPvarGetLbGlobal(var) - 0.5 || solvals[b] > SCIPvarGetUbGlobal(var) + 0.5 )
1538  continue;
1539 
1540  redcostscores[nunfixedvars] = getVariableRedcostScore(scip, var, solvals[b], heurdata->uselocalredcost);
1541  pscostscores[nunfixedvars] = getVariablePscostScore(scip, var, solvals[b], heurdata->uselocalredcost);
1542 
1543  unfixedvars[nunfixedvars] = var;
1544  perm[nunfixedvars] = nunfixedvars;
1545  randscores[nunfixedvars] = SCIPrandomGetReal(rng, 0.0, 1.0);
1546 
1547  /* these assignments are based on the fact that nunfixedvars <= b */
1548  solvals[nunfixedvars] = solvals[b];
1549  distances[nunfixedvars] = distances[b];
1550 
1551  SCIPdebugMsg(scip, "Var <%s> scores: dist %3d, red cost %15.9g, pscost %15.9g rand %6.4f\n",
1552  SCIPvarGetName(var), distances[nunfixedvars], redcostscores[nunfixedvars],
1553  pscostscores[nunfixedvars], randscores[nunfixedvars]);
1554 
1555  nunfixedvars++;
1556  }
1557 
1558  /* use selection algorithm (order of the variables does not matter) for quickly completing the fixing */
1559  varprio.randscores = randscores;
1560  varprio.distances = distances;
1561  varprio.redcostscores = redcostscores;
1562  varprio.pscostscores = pscostscores;
1563 
1564  nvarstoadd = MIN(nunfixedvars, nvarstoadd);
1565 
1566  /* select the first nvarstoadd many variables according to the score */
1567  if( nvarstoadd < nunfixedvars )
1568  SCIPselectInd(perm, sortIndCompAlns, &varprio, nvarstoadd, nunfixedvars);
1569 
1570  /* loop over the first elements of the selection defined in permutation. They represent the best variables */
1571  for( b = 0; b < nvarstoadd; ++b )
1572  {
1573  int permindex = perm[b];
1574  assert(permindex >= 0);
1575  assert(permindex < nunfixedvars);
1576 
1577  tryAdd2variableBuffer(scip, unfixedvars[permindex], solvals[permindex], varbuf, valbuf, nfixings, TRUE);
1578  }
1579 
1580  *success = TRUE;
1581 
1582  /* free buffer arrays */
1583  SCIPfreeBufferArray(scip, &pscostscores);
1584  SCIPfreeBufferArray(scip, &unfixedvars);
1585  SCIPfreeBufferArray(scip, &isfixed);
1586  SCIPfreeBufferArray(scip, &solvals);
1587  SCIPfreeBufferArray(scip, &redcostscores);
1588  SCIPfreeBufferArray(scip, &distances);
1589  SCIPfreeBufferArray(scip, &perm);
1590  SCIPfreeBufferArray(scip, &randscores);
1591 
1592  return SCIP_OKAY;
1593 }
1594 
1595 /** create the bandit algorithm for the heuristic depending on the user parameter */
1596 static
1598  SCIP* scip, /**< SCIP data structure */
1599  SCIP_HEURDATA* heurdata, /**< heuristic data structure */
1600  SCIP_Real* priorities, /**< call priorities for active neighborhoods */
1601  unsigned int initseed /**< initial random seed */
1602  )
1603 {
1604  switch (heurdata->banditalgo)
1605  {
1606  case 'u':
1607  SCIP_CALL( SCIPcreateBanditUcb(scip, &heurdata->bandit, priorities,
1608  heurdata->ucb_alpha, heurdata->nactiveneighborhoods, initseed) );
1609  break;
1610 
1611  case 'e':
1612  SCIP_CALL( SCIPcreateBanditExp3(scip, &heurdata->bandit, priorities,
1613  heurdata->exp3_gamma, heurdata->exp3_beta, heurdata->nactiveneighborhoods, initseed) );
1614  break;
1615 
1616  case 'g':
1617  SCIP_CALL( SCIPcreateBanditEpsgreedy(scip, &heurdata->bandit, priorities,
1618  heurdata->epsgreedy_eps, FALSE, 0.9, 0, heurdata->nactiveneighborhoods, initseed) );
1619  break;
1620 
1621  default:
1622  SCIPerrorMessage("Unknown bandit parameter %c\n", heurdata->banditalgo);
1623  return SCIP_INVALIDDATA;
1624  }
1625 
1626  return SCIP_OKAY;
1627 }
1628 
1629 /*
1630  * Callback methods of primal heuristic
1631  */
1632 
1633 /** copy method for primal heuristic plugins (called when SCIP copies plugins) */
1634 static
1635 SCIP_DECL_HEURCOPY(heurCopyAlns)
1636 { /*lint --e{715}*/
1637  assert(scip != NULL);
1638  assert(heur != NULL);
1639  assert(strcmp(SCIPheurGetName(heur), HEUR_NAME) == 0);
1640 
1641  /* call inclusion method of primal heuristic */
1642  SCIP_CALL( SCIPincludeHeurAlns(scip) );
1643 
1644  return SCIP_OKAY;
1645 }
1646 
1647 /** unfix some of the variables because there are too many fixed
1648  *
1649  * a variable is ideally unfixed if it is close to other unfixed variables
1650  * and fixing it has a high reduced cost impact
1651  */
1652 static
1654  SCIP* scip, /**< SCIP data structure */
1655  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
1656  SCIP_VAR** varbuf, /**< buffer array to store variables to fix */
1657  SCIP_Real* valbuf, /**< buffer array to store fixing values */
1658  int* nfixings, /**< pointer to store the number of fixings */
1659  int ntargetfixings, /**< number of required target fixings */
1660  SCIP_Bool* success /**< pointer to store whether the target fixings have been successfully reached */
1661  )
1662 {
1663  VARPRIO varprio;
1664  SCIP_Real* redcostscores;
1665  SCIP_Real* pscostscores;
1666  SCIP_Real* randscores;
1667  SCIP_VAR** unfixedvars;
1668  SCIP_VAR** varbufcpy;
1669  SCIP_Real* valbufcpy;
1670  SCIP_Bool* isfixedvar;
1671  SCIP_VAR** vars;
1672  SCIP_RANDNUMGEN* rng;
1673  int* distances;
1674  int* fixeddistances;
1675  int* perm;
1676  int nvars;
1677  int i;
1678  int nbinintvars;
1679  int nunfixed;
1680 
1681  *success = FALSE;
1682 
1683  nbinintvars = SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip);
1684  if( nbinintvars == 0 )
1685  return SCIP_OKAY;
1686 
1687  assert(*nfixings > 0);
1688 
1689  nvars = SCIPgetNVars(scip);
1690  SCIP_CALL( SCIPallocBufferArray(scip, &isfixedvar, nvars) );
1691  SCIP_CALL( SCIPallocBufferArray(scip, &unfixedvars, nbinintvars) );
1692  SCIP_CALL( SCIPallocBufferArray(scip, &distances, nvars) );
1693  SCIP_CALL( SCIPallocBufferArray(scip, &fixeddistances, *nfixings) );
1694  SCIP_CALL( SCIPallocBufferArray(scip, &redcostscores, *nfixings) );
1695  SCIP_CALL( SCIPallocBufferArray(scip, &randscores, *nfixings) );
1696  SCIP_CALL( SCIPallocBufferArray(scip, &perm, *nfixings) );
1697  SCIP_CALL( SCIPallocBufferArray(scip, &pscostscores, *nfixings) );
1698 
1699  SCIP_CALL( SCIPduplicateBufferArray(scip, &varbufcpy, varbuf, *nfixings) );
1700  SCIP_CALL( SCIPduplicateBufferArray(scip, &valbufcpy, valbuf, *nfixings) );
1701 
1702  /*
1703  * collect the unfixed binary and integer variables
1704  */
1705  BMSclearMemoryArray(isfixedvar, nvars);
1706  /* loop over fixed variables and mark their respective positions as fixed */
1707  for( i = 0; i < *nfixings; ++i )
1708  {
1709  int probindex = SCIPvarGetProbindex(varbuf[i]);
1710 
1711  assert(probindex >= 0);
1712 
1713  isfixedvar[probindex] = TRUE;
1714  }
1715 
1716  nunfixed = 0;
1717  vars = SCIPgetVars(scip);
1718  /* collect unfixed binary and integer variables */
1719  for( i = 0; i < nbinintvars; ++i )
1720  {
1721  if( ! isfixedvar[i] )
1722  unfixedvars[nunfixed++] = vars[i];
1723  }
1724 
1725  varprio.usedistances = heurdata->usedistances && nunfixed > 0;
1726 
1727  /* collect distances of all fixed variables from those that are not fixed */
1728  if( varprio.usedistances )
1729  {
1730  SCIP_CALL( SCIPvariablegraphBreadthFirst(scip, NULL, unfixedvars, nunfixed, distances, INT_MAX, INT_MAX, INT_MAX) );
1731 
1732  for( i = 0; i < *nfixings; ++i )
1733  {
1734  int probindex = SCIPvarGetProbindex(varbuf[i]);
1735  if( probindex >= 0 )
1736  fixeddistances[i] = distances[probindex];
1737  }
1738  }
1739  else
1740  {
1741  BMSclearMemoryArray(fixeddistances, *nfixings);
1742  }
1743 
1744  /* collect reduced cost scores of the fixings and assign random scores */
1745  rng = SCIPbanditGetRandnumgen(heurdata->bandit);
1746  for( i = 0; i < *nfixings; ++i )
1747  {
1748  SCIP_VAR* fixedvar = varbuf[i];
1749  SCIP_Real fixval = valbuf[i];
1750 
1751  /* use negative reduced cost and pseudo cost scores to prefer variable fixings with small score */
1752  redcostscores[i] = - getVariableRedcostScore(scip, fixedvar, fixval, heurdata->uselocalredcost);
1753  pscostscores[i] = - getVariablePscostScore(scip, fixedvar, fixval, heurdata->uselocalredcost);
1754  randscores[i] = SCIPrandomGetReal(rng, 0.0, 1.0);
1755  perm[i] = i;
1756 
1757  SCIPdebugMsg(scip, "Var <%s> scores: dist %3d, red cost %15.9g, pscost %15.9g rand %6.4f\n",
1758  SCIPvarGetName(fixedvar), fixeddistances[i], redcostscores[i], pscostscores[i], randscores[i]);
1759  }
1760 
1761  varprio.distances = fixeddistances;
1762  varprio.randscores = randscores;
1763  varprio.redcostscores = redcostscores;
1764  varprio.pscostscores = pscostscores;
1765  varprio.useredcost = heurdata->useredcost;
1766  varprio.usepscost = heurdata->usepscost;
1767  varprio.scip = scip;
1768 
1769  /* scores are assigned in such a way that variables with a smaller score should be fixed last */
1770  SCIPselectDownInd(perm, sortIndCompAlns, &varprio, ntargetfixings, *nfixings);
1771 
1772  /* bring the desired variables to the front of the array */
1773  for( i = 0; i < ntargetfixings; ++i )
1774  {
1775  valbuf[i] = valbufcpy[perm[i]];
1776  varbuf[i] = varbufcpy[perm[i]];
1777  }
1778 
1779  *nfixings = ntargetfixings;
1780 
1781  /* free the buffer arrays in reverse order of allocation */
1782  SCIPfreeBufferArray(scip, &valbufcpy);
1783  SCIPfreeBufferArray(scip, &varbufcpy);
1784  SCIPfreeBufferArray(scip, &pscostscores);
1785  SCIPfreeBufferArray(scip, &perm);
1786  SCIPfreeBufferArray(scip, &randscores);
1787  SCIPfreeBufferArray(scip, &redcostscores);
1788  SCIPfreeBufferArray(scip, &fixeddistances);
1789  SCIPfreeBufferArray(scip, &distances);
1790  SCIPfreeBufferArray(scip, &unfixedvars);
1791  SCIPfreeBufferArray(scip, &isfixedvar);
1792 
1793  *success = TRUE;
1794 
1795  return SCIP_OKAY;
1796 }
1797 
1798 /** call variable fixing callback for this neighborhood and orchestrate additional variable fixings, if necessary */
1799 static
1801  SCIP* scip, /**< SCIP data structure */
1802  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
1803  NH* neighborhood, /**< neighborhood data structure */
1804  SCIP_VAR** varbuf, /**< buffer array to keep variables that should be fixed */
1805  SCIP_Real* valbuf, /**< buffer array to keep fixing values */
1806  int* nfixings, /**< pointer to store the number of variable fixings */
1807  SCIP_RESULT* result /**< pointer to store the result of the fixing operation */
1808  )
1809 {
1810  int ntargetfixings;
1811  int nmaxfixings;
1812  int nminfixings;
1813  int nbinintvars;
1814 
1815  assert(scip != NULL);
1816  assert(neighborhood != NULL);
1817  assert(varbuf != NULL);
1818  assert(valbuf != NULL);
1819  assert(nfixings != NULL);
1820  assert(result != NULL);
1821 
1822  *nfixings = 0;
1823 
1824  *result = SCIP_DIDNOTRUN;
1825  ntargetfixings = (int)(neighborhood->fixingrate.targetfixingrate * (SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip)));
1826 
1827  if( neighborhood->varfixings != NULL )
1828  {
1829  SCIP_CALL( neighborhood->varfixings(scip, neighborhood, varbuf, valbuf, nfixings, result) );
1830 
1831  if( *result != SCIP_SUCCESS )
1832  return SCIP_OKAY;
1833  }
1834  else if( ntargetfixings == 0 )
1835  {
1836  *result = SCIP_SUCCESS;
1837 
1838  return SCIP_OKAY;
1839  }
1840 
1841  /* compute upper and lower target fixing limits using tolerance parameters */
1842  assert(neighborhood->varfixings == NULL || *result != SCIP_DIDNOTRUN);
1843  nbinintvars = SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip);
1844  ntargetfixings = (int)(neighborhood->fixingrate.targetfixingrate * nbinintvars);
1845  nminfixings = (int)((neighborhood->fixingrate.targetfixingrate - heurdata->fixtol) * nbinintvars);
1846  nminfixings = MAX(nminfixings, 0);
1847  nmaxfixings = (int)((neighborhood->fixingrate.targetfixingrate + heurdata->unfixtol) * nbinintvars);
1848  nmaxfixings = MIN(nmaxfixings, nbinintvars);
1849 
1850  SCIPdebugMsg(scip, "Neighborhood Fixings/Target: %d / %d <= %d <= %d\n",*nfixings, nminfixings, ntargetfixings, nmaxfixings);
1851 
1852  /* if too few fixings, use a strategy to select more variable fixings: randomized, LP graph, ReducedCost based, mix */
1853  if( (*result == SCIP_SUCCESS || *result == SCIP_DIDNOTRUN) && (*nfixings < nminfixings) )
1854  {
1855  SCIP_Bool success;
1856  SCIP_SOL* refsol;
1857 
1858  /* get reference solution from neighborhood */
1859  SCIP_CALL( neighborhoodGetRefsol(scip, neighborhood, &refsol) );
1860 
1861  /* try to fix more variables based on the reference solution */
1862  if( refsol != NULL )
1863  {
1864  SCIP_CALL( alnsFixMoreVariables(scip, heurdata, refsol, varbuf, valbuf, nfixings, ntargetfixings, &success) );
1865  }
1866  else
1867  success = FALSE;
1868 
1869  if( success )
1870  *result = SCIP_SUCCESS;
1871  else if( *result == SCIP_SUCCESS )
1872  *result = SCIP_DIDNOTFIND;
1873  else
1874  *result = SCIP_DIDNOTRUN;
1875 
1876  SCIPdebugMsg(scip, "After additional fixings: %d / %d\n",*nfixings, ntargetfixings);
1877  }
1878  else if( (SCIP_Real)(*nfixings) > nmaxfixings )
1879  {
1880  SCIP_Bool success;
1881 
1882  SCIP_CALL( alnsUnfixVariables(scip, heurdata, varbuf, valbuf, nfixings, ntargetfixings, &success) );
1883 
1884  assert(success);
1885  *result = SCIP_SUCCESS;
1886  SCIPdebugMsg(scip, "Unfixed variables, fixed variables remaining: %d\n", ntargetfixings);
1887  }
1888  else
1889  {
1890  SCIPdebugMsg(scip, "No additional fixings performed\n");
1891  }
1892 
1893  return SCIP_OKAY;
1894 }
1895 
1896 /** change the sub-SCIP by restricting variable domains, changing objective coefficients, or adding constraints */
1897 static
1899  SCIP* sourcescip, /**< source SCIP data structure */
1900  SCIP* targetscip, /**< target SCIP data structure */
1901  NH* neighborhood, /**< neighborhood */
1902  SCIP_VAR** targetvars, /**< array of target SCIP variables aligned with source SCIP variables */
1903  int* ndomchgs, /**< pointer to store the number of variable domain changes */
1904  int* nchgobjs, /**< pointer to store the number of changed objective coefficients */
1905  int* naddedconss, /**< pointer to store the number of added constraints */
1906  SCIP_Bool* success /**< pointer to store whether the sub-SCIP has been successfully modified */
1907  )
1908 {
1909  assert(sourcescip != NULL);
1910  assert(targetscip != NULL);
1911  assert(neighborhood != NULL);
1912  assert(targetvars != NULL);
1913  assert(ndomchgs != NULL);
1914  assert(nchgobjs != NULL);
1915  assert(naddedconss != NULL);
1916  assert(success != NULL);
1917 
1918  *success = FALSE;
1919  *ndomchgs = 0;
1920  *nchgobjs = 0;
1921  *naddedconss = 0;
1922 
1923  /* call the change sub-SCIP callback of the neighborhood */
1924  if( neighborhood->changesubscip != NULL )
1925  {
1926  SCIP_CALL( neighborhood->changesubscip(sourcescip, targetscip, neighborhood, targetvars, ndomchgs, nchgobjs, naddedconss, success) );
1927  }
1928  else
1929  {
1930  *success = TRUE;
1931  }
1932 
1933  return SCIP_OKAY;
1934 }
1935 
1936 /** set sub-SCIP solving limits */
1937 static
1939  SCIP* subscip, /**< SCIP data structure */
1940  SOLVELIMITS* solvelimits /**< pointer to solving limits data structure */
1941  )
1942 {
1943  assert(subscip != NULL);
1944  assert(solvelimits != NULL);
1945 
1946  assert(solvelimits->nodelimit >= solvelimits->stallnodes);
1947 
1948  SCIP_CALL( SCIPsetLongintParam(subscip, "limits/nodes", solvelimits->nodelimit) );
1949  SCIP_CALL( SCIPsetLongintParam(subscip, "limits/stallnodes", solvelimits->stallnodes));
1950  SCIP_CALL( SCIPsetRealParam(subscip, "limits/time", solvelimits->timelimit) );
1951  SCIP_CALL( SCIPsetRealParam(subscip, "limits/memory", solvelimits->memorylimit) );
1952 
1953  return SCIP_OKAY;
1954 }
1955 
1956 /** determine limits for a sub-SCIP */
1957 static
1959  SCIP* scip, /**< SCIP data structure */
1960  SCIP_HEUR* heur, /**< this heuristic */
1961  SOLVELIMITS* solvelimits, /**< pointer to solving limits data structure */
1962  SCIP_Bool* runagain /**< can we solve another sub-SCIP with these limits */
1963  )
1964 {
1965  SCIP_HEURDATA* heurdata;
1966  SCIP_Real initfactor;
1967  SCIP_Real nodesquot;
1968  SCIP_Bool avoidmemout;
1969 
1970  assert(scip != NULL);
1971  assert(heur != NULL);
1972  assert(solvelimits != NULL);
1973  assert(runagain != NULL);
1974 
1975  heurdata = SCIPheurGetData(heur);
1976 
1977  /* check whether there is enough time and memory left */
1978  SCIP_CALL( SCIPgetRealParam(scip, "limits/time", &solvelimits->timelimit) );
1979  if( ! SCIPisInfinity(scip, solvelimits->timelimit) )
1980  solvelimits->timelimit -= SCIPgetSolvingTime(scip);
1981  SCIP_CALL( SCIPgetRealParam(scip, "limits/memory", &solvelimits->memorylimit) );
1982  SCIP_CALL( SCIPgetBoolParam(scip, "misc/avoidmemout", &avoidmemout) );
1983 
1984  /* substract the memory already used by the main SCIP and the estimated memory usage of external software */
1985  if( ! SCIPisInfinity(scip, solvelimits->memorylimit) )
1986  {
1987  solvelimits->memorylimit -= SCIPgetMemUsed(scip)/1048576.0;
1988  solvelimits->memorylimit -= SCIPgetMemExternEstim(scip)/1048576.0;
1989  }
1990 
1991  /* abort if no time is left or not enough memory (we don't abort in this case if misc_avoidmemout == FALSE)
1992  * to create a copy of SCIP, including external memory usage */
1993  if( solvelimits->timelimit <= 0.0 || (avoidmemout && solvelimits->memorylimit <= 2.0*SCIPgetMemExternEstim(scip)/1048576.0) )
1994  *runagain = FALSE;
1995 
1996  nodesquot = heurdata->nodesquot;
1997 
1998  /* if the heuristic is used to measure all rewards, it will always be penalized here */
1999  if( heurdata->rewardfile == NULL )
2000  nodesquot *= (SCIPheurGetNBestSolsFound(heur) + 1.0)/(SCIPheurGetNCalls(heur) + 1.0);
2001 
2002  nodesquot = MAX(nodesquot, heurdata->nodesquotmin);
2003 
2004  /* calculate the search node limit of the heuristic */
2005  solvelimits->stallnodes = (SCIP_Longint)(nodesquot * SCIPgetNNodes(scip));
2006  solvelimits->stallnodes += heurdata->nodesoffset;
2007  solvelimits->stallnodes -= heurdata->usednodes;
2008  solvelimits->stallnodes -= 100 * SCIPheurGetNCalls(heur);
2009  solvelimits->stallnodes = MIN(heurdata->maxnodes, solvelimits->stallnodes);
2010 
2011  /* use a smaller budget if not all neighborhoods have been initialized yet */
2012  assert(heurdata->ninitneighborhoods >= 0);
2013  initfactor = (heurdata->nactiveneighborhoods - heurdata->ninitneighborhoods + 1.0) / (heurdata->nactiveneighborhoods + 1.0);
2014  solvelimits->stallnodes = (SCIP_Longint)(solvelimits->stallnodes * initfactor);
2015  solvelimits->nodelimit = (SCIP_Longint)(heurdata->maxnodes);
2016 
2017  /* check whether we have enough nodes left to call subproblem solving */
2018  if( solvelimits->stallnodes < heurdata->targetnodes )
2019  *runagain = FALSE;
2020 
2021  return SCIP_OKAY;
2022 }
2023 
2024 /** return the bandit algorithm that should be used */
2025 static
2027  SCIP_HEURDATA* heurdata /**< heuristic data of the ALNS neighborhood */
2028  )
2029 {
2030  assert(heurdata != NULL);
2031  return heurdata->bandit;
2032 }
2033 
2034 /** select a neighborhood depending on the selected bandit algorithm */
2035 static
2037  SCIP* scip, /**< SCIP data structure */
2038  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
2039  int* neighborhoodidx /**< pointer to store the selected neighborhood index */
2040  )
2041 {
2042  SCIP_BANDIT* bandit;
2043  assert(scip != NULL);
2044  assert(heurdata != NULL);
2045  assert(neighborhoodidx != NULL);
2046 
2047  *neighborhoodidx = -1;
2048 
2049  bandit = getBandit(heurdata);
2050 
2051  SCIP_CALL( SCIPbanditSelect(bandit, neighborhoodidx) );
2052  assert(*neighborhoodidx >= 0);
2053 
2054  return SCIP_OKAY;
2055 }
2056 
2057 /** Calculate reward based on the selected reward measure */
2058 static
2060  SCIP* scip, /**< SCIP data structure */
2061  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
2062  NH_STATS* runstats, /**< run statistics */
2063  SCIP_Real* rewardptr /**< array to store the computed rewards, total and individual */
2064  )
2065 {
2066  SCIP_Real reward = 0.0;
2067  SCIP_Real effort;
2068  int ndiscretevars;
2069 
2070  memset(rewardptr, 0, sizeof(*rewardptr)*(int)NREWARDTYPES);
2071 
2072  assert(rewardptr != NULL);
2073  assert(runstats->usednodes >= 0);
2074  assert(runstats->nfixings >= 0);
2075 
2076  effort = runstats->usednodes / 100.0;
2077 
2078  ndiscretevars = SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip);
2079  /* assume that every fixed variable linearly reduces the subproblem complexity */
2080  if( ndiscretevars > 0 )
2081  {
2082  effort = (1.0 - (runstats->nfixings / (SCIP_Real)ndiscretevars)) * effort;
2083  }
2084  assert(rewardptr != NULL);
2085 
2086  /* a positive reward is only assigned if a new incumbent solution was found */
2087  if( runstats->nbestsolsfound > 0 )
2088  {
2089  SCIP_Real rewardcontrol = heurdata->rewardcontrol;
2090 
2091  SCIP_Real lb;
2092  SCIP_Real ub;
2093 
2094  /* the indicator function is simply 1.0 */
2095  rewardptr[REWARDTYPE_BESTSOL] = 1.0;
2096  rewardptr[REWARDTYPE_NOSOLPENALTY] = 1.0;
2097 
2098  ub = runstats->newupperbound;
2099  lb = SCIPgetLowerbound(scip);
2100 
2101  /* compute the closed gap reward */
2102  if( SCIPisEQ(scip, ub, lb) || SCIPisInfinity(scip, runstats->oldupperbound) )
2103  rewardptr[REWARDTYPE_CLOSEDGAP] = 1.0;
2104  else
2105  {
2106  rewardptr[REWARDTYPE_CLOSEDGAP] = (runstats->oldupperbound - ub) / (runstats->oldupperbound - lb);
2107  }
2108 
2109  /* the reward is a convex combination of the best solution reward and the reward for the closed gap */
2110  reward = rewardcontrol * rewardptr[REWARDTYPE_BESTSOL] + (1.0 - rewardcontrol) * rewardptr[REWARDTYPE_CLOSEDGAP];
2111 
2112  /* optionally, scale the reward by the involved effort */
2113  if( heurdata->scalebyeffort )
2114  reward /= (effort + 1.0);
2115 
2116  /* add the baseline and rescale the reward into the interval [baseline, 1.0] */
2117  reward = heurdata->rewardbaseline + (1.0 - heurdata->rewardbaseline) * reward;
2118  }
2119  else
2120  {
2121  /* linearly decrease the reward based on the number of nodes spent */
2122  SCIP_Real maxeffort = heurdata->targetnodes;
2123  SCIP_Real usednodes = runstats->usednodes;
2124 
2125  if( ndiscretevars > 0 )
2126  usednodes *= (1.0 - (runstats->nfixings / (SCIP_Real)ndiscretevars));
2127 
2128  rewardptr[REWARDTYPE_NOSOLPENALTY] = 1 - (usednodes / maxeffort);
2129  rewardptr[REWARDTYPE_NOSOLPENALTY] = MAX(0.0, rewardptr[REWARDTYPE_NOSOLPENALTY]);
2130  reward = heurdata->rewardbaseline * rewardptr[REWARDTYPE_NOSOLPENALTY];
2131  }
2132 
2133  rewardptr[REWARDTYPE_TOTAL] = reward;
2134 
2135  return SCIP_OKAY;
2136 }
2137 
2138 /** update internal bandit algorithm statistics for future draws */
2139 static
2141  SCIP* scip, /**< SCIP data structure */
2142  SCIP_HEURDATA* heurdata, /**< heuristic data of the ALNS neighborhood */
2143  SCIP_Real reward, /**< measured reward */
2144  int neighborhoodidx /**< the neighborhood that was chosen */
2145  )
2146 {
2147  SCIP_BANDIT* bandit;
2148  assert(scip != NULL);
2149  assert(heurdata != NULL);
2150  assert(neighborhoodidx >= 0);
2151  assert(neighborhoodidx < heurdata->nactiveneighborhoods);
2152 
2153  bandit = getBandit(heurdata);
2154 
2155  SCIPdebugMsg(scip, "Rewarding bandit algorithm action %d with reward %.2f\n", neighborhoodidx, reward);
2156  SCIP_CALL( SCIPbanditUpdate(bandit, neighborhoodidx, reward) );
2157 
2158  return SCIP_OKAY;
2159 }
2160 
2161 /** set up the sub-SCIP parameters, objective cutoff, and solution limits */
2162 static
2164  SCIP* scip, /**< SCIP data structure */
2165  SCIP* subscip, /**< sub-SCIP data structure */
2166  SCIP_VAR** subvars, /**< array of sub-SCIP variables in the order of the main SCIP */
2167  SOLVELIMITS* solvelimits, /**< pointer to solving limits data structure */
2168  SCIP_HEUR* heur, /**< this heuristic */
2169  SCIP_Bool objchgd /**< did the objective change between the source and the target SCIP? */
2170  )
2171 {
2172  SCIP_HEURDATA* heurdata;
2173  SCIP_Real cutoff;
2174  SCIP_Real upperbound;
2175 
2176  heurdata = SCIPheurGetData(heur);
2177 
2178  /* do not abort subproblem on CTRL-C */
2179  SCIP_CALL( SCIPsetBoolParam(subscip, "misc/catchctrlc", FALSE) );
2180 
2181  /* disable output to console unless we are in debug mode */
2182  SCIP_CALL( SCIPsetIntParam(subscip, "display/verblevel", 0) );
2183 
2184  /* disable statistic timing inside sub SCIP */
2185  SCIP_CALL( SCIPsetBoolParam(subscip, "timing/statistictiming", FALSE) );
2186 
2187 #ifdef ALNS_SUBSCIPOUTPUT
2188  SCIP_CALL( SCIPsetIntParam(subscip, "display/verblevel", 5) );
2189  SCIP_CALL( SCIPsetIntParam(subscip, "display/freq", 1) );
2190  /* enable statistic timing inside sub SCIP */
2191  SCIP_CALL( SCIPsetBoolParam(subscip, "timing/statistictiming", TRUE) );
2192 #endif
2193 
2194  SCIP_CALL( SCIPsetIntParam(subscip, "limits/bestsol", heurdata->nsolslim) );
2195 
2196  /* forbid recursive call of heuristics and separators solving subMIPs */
2197  if( ! heurdata->usesubscipheurs )
2198  {
2199  SCIP_CALL( SCIPsetSubscipsOff(subscip, TRUE) );
2200  }
2201 
2202  /* disable cutting plane separation */
2204 
2205  /* disable expensive presolving */
2207 
2208  /* use best estimate node selection */
2209  if( SCIPfindNodesel(subscip, "estimate") != NULL && ! SCIPisParamFixed(subscip, "nodeselection/estimate/stdpriority") )
2210  {
2211  SCIP_CALL( SCIPsetIntParam(subscip, "nodeselection/estimate/stdpriority", INT_MAX/4) );
2212  }
2213 
2214  /* use inference branching */
2215  if( SCIPfindBranchrule(subscip, "inference") != NULL && ! SCIPisParamFixed(subscip, "branching/inference/priority") )
2216  {
2217  SCIP_CALL( SCIPsetIntParam(subscip, "branching/inference/priority", INT_MAX/4) );
2218  }
2219 
2220  /* enable conflict analysis and restrict conflict pool */
2221  if( ! SCIPisParamFixed(subscip, "conflict/enable") )
2222  {
2223  SCIP_CALL( SCIPsetBoolParam(subscip, "conflict/enable", TRUE) );
2224  }
2225 
2226  if( !SCIPisParamFixed(subscip, "conflict/useboundlp") )
2227  {
2228  SCIP_CALL( SCIPsetCharParam(subscip, "conflict/useboundlp", 'o') );
2229  }
2230 
2231  if( ! SCIPisParamFixed(subscip, "conflict/maxstoresize") )
2232  {
2233  SCIP_CALL( SCIPsetIntParam(subscip, "conflict/maxstoresize", 100) );
2234  }
2235 
2236  /* speed up sub-SCIP by not checking dual LP feasibility */
2237  SCIP_CALL( SCIPsetBoolParam(subscip, "lp/checkdualfeas", FALSE) );
2238 
2239  /* add an objective cutoff */
2240  if( ! SCIPisInfinity(scip, SCIPgetUpperbound(scip)) )
2241  {
2242  upperbound = SCIPgetUpperbound(scip) - SCIPsumepsilon(scip);
2243  if( ! SCIPisInfinity(scip, -1.0 * SCIPgetLowerbound(scip)) )
2244  {
2245  cutoff = (1 - heurdata->minimprove) * SCIPgetUpperbound(scip)
2246  + heurdata->minimprove * SCIPgetLowerbound(scip);
2247  }
2248  else
2249  {
2250  if( SCIPgetUpperbound(scip) >= 0 )
2251  cutoff = (1 - heurdata->minimprove) * SCIPgetUpperbound(scip);
2252  else
2253  cutoff = (1 + heurdata->minimprove) * SCIPgetUpperbound(scip);
2254  }
2255  cutoff = MIN(upperbound, cutoff);
2256 
2257  if( SCIPisObjIntegral(scip) )
2258  cutoff = SCIPfloor(scip, cutoff);
2259 
2260  SCIPdebugMsg(scip, "Sub-SCIP cutoff: %15.9" SCIP_REAL_FORMAT " (%15.9" SCIP_REAL_FORMAT " in original space)\n",
2261  cutoff, SCIPretransformObj(scip, cutoff));
2262 
2263  /* if the objective changed between the source and the target SCIP, encode the cutoff as a constraint */
2264  if( ! objchgd )
2265  {
2266  SCIP_CALL(SCIPsetObjlimit(subscip, cutoff));
2267 
2268  SCIPdebugMsg(scip, "Cutoff added as Objective Limit\n");
2269  }
2270  else
2271  {
2272  SCIP_CONS* objcons;
2273  int nvars;
2274  SCIP_VAR** vars;
2275  int i;
2276 
2277  vars = SCIPgetVars(scip);
2278  nvars = SCIPgetNVars(scip);
2279 
2280  SCIP_CALL( SCIPcreateConsLinear(subscip, &objcons, "objbound_of_origscip", 0, NULL, NULL, -SCIPinfinity(subscip), cutoff,
2281  TRUE, TRUE, TRUE, TRUE, TRUE, FALSE, FALSE, FALSE, FALSE, FALSE) );
2282  for( i = 0; i < nvars; ++i)
2283  {
2284  if( ! SCIPisFeasZero(subscip, SCIPvarGetObj(vars[i])) )
2285  {
2286  assert(subvars[i] != NULL);
2287  SCIP_CALL( SCIPaddCoefLinear(subscip, objcons, subvars[i], SCIPvarGetObj(vars[i])) );
2288  }
2289  }
2290  SCIP_CALL( SCIPaddCons(subscip, objcons) );
2291  SCIP_CALL( SCIPreleaseCons(subscip, &objcons) );
2292 
2293  SCIPdebugMsg(scip, "Cutoff added as constraint\n");
2294  }
2295  }
2296 
2297  /* set solve limits for sub-SCIP */
2298  SCIP_CALL( setLimits(subscip, solvelimits) );
2299 
2300  /* change random seed of sub-SCIP */
2301  if( heurdata->subsciprandseeds )
2302  {
2303  SCIP_CALL( SCIPsetIntParam(subscip, "randomization/randomseedshift", (int)SCIPheurGetNCalls(heur)) );
2304  }
2305 
2306  SCIPdebugMsg(scip, "Solve Limits: %lld (%lld) nodes (stall nodes), %.1f sec., %d sols\n",
2307  solvelimits->nodelimit, solvelimits->stallnodes, solvelimits->timelimit, heurdata->nsolslim);
2308 
2309  return SCIP_OKAY;
2310 }
2311 
2312 /** execution method of primal heuristic */
2313 static
2314 SCIP_DECL_HEUREXEC(heurExecAlns)
2315 { /*lint --e{715}*/
2316  SCIP_HEURDATA* heurdata;
2317  SCIP_VAR** varbuf;
2318  SCIP_Real* valbuf;
2319  SCIP_VAR** vars;
2320  SCIP_VAR** subvars;
2321  NH_STATS runstats[NNEIGHBORHOODS];
2322  SCIP_STATUS subscipstatus[NNEIGHBORHOODS];
2323  SCIP* subscip = NULL;
2324 
2325  int nfixings;
2326  int nvars;
2327  int neighborhoodidx;
2328  int ntries;
2329  SCIP_Bool tryagain;
2330  NH* neighborhood;
2331  SOLVELIMITS solvelimits;
2332  SCIP_Bool success;
2333  SCIP_Bool run;
2334  SCIP_Bool allrewardsmode;
2335  SCIP_Real rewards[NNEIGHBORHOODS][NREWARDTYPES] = {{0}};
2336  int banditidx;
2337 
2338  int i;
2339 
2340  heurdata = SCIPheurGetData(heur);
2341  assert(heurdata != NULL);
2342 
2343  *result = SCIP_DIDNOTRUN;
2344 
2345  if( heurdata->nactiveneighborhoods == 0 )
2346  return SCIP_OKAY;
2347 
2348  /* we only allow to run multiple times at a node during the root */
2349  if( (heurtiming & SCIP_HEURTIMING_DURINGLPLOOP) && (SCIPgetDepth(scip) > 0 || !heurdata->initduringroot) )
2350  return SCIP_OKAY;
2351 
2352  /* update internal incumbent solution */
2353  if( SCIPgetBestSol(scip) != heurdata->lastcallsol )
2354  {
2355  heurdata->lastcallsol = SCIPgetBestSol(scip);
2356  heurdata->firstcallthissol = SCIPheurGetNCalls(heur);
2357  }
2358 
2359  /* do not run more than a user-defined number of times on each incumbent (-1: no limit) */
2360  if( heurdata->maxcallssamesol != -1 )
2361  {
2362  SCIP_Longint samesollimit = (heurdata->maxcallssamesol > 0) ?
2363  heurdata->maxcallssamesol :
2364  heurdata->nactiveneighborhoods;
2365 
2366  if( SCIPheurGetNCalls(heur) - heurdata->firstcallthissol >= samesollimit )
2367  {
2368  SCIPdebugMsg(scip, "Heuristic already called %" SCIP_LONGINT_FORMAT " times on current incumbent\n", SCIPheurGetNCalls(heur) - heurdata->firstcallthissol);
2369  return SCIP_OKAY;
2370  }
2371  }
2372 
2373  /* wait for a sufficient number of nodes since last incumbent solution */
2374  if( SCIPgetDepth(scip) > 0 && SCIPgetBestSol(scip) != NULL
2375  && (SCIPgetNNodes(scip) - SCIPsolGetNodenum(SCIPgetBestSol(scip))) < heurdata->waitingnodes )
2376  {
2377  SCIPdebugMsg(scip, "Waiting nodes not satisfied\n");
2378  return SCIP_OKAY;
2379  }
2380 
2381  run = TRUE;
2382  /* check if budget allows a run of the next selected neighborhood */
2383  SCIP_CALL( determineLimits(scip, heur, &solvelimits, &run) );
2384  SCIPdebugMsg(scip, "Budget check: %" SCIP_LONGINT_FORMAT " (%" SCIP_LONGINT_FORMAT ") %s\n", solvelimits.nodelimit, heurdata->targetnodes, run ? "passed" : "must wait");
2385 
2386  if( ! run )
2387  return SCIP_OKAY;
2388 
2389  /* delay the heuristic if local reduced costs should be used for generic variable unfixing */
2390  if( heurdata->uselocalredcost && (nodeinfeasible || ! SCIPhasCurrentNodeLP(scip) || SCIPgetLPSolstat(scip) != SCIP_LPSOLSTAT_OPTIMAL) )
2391  {
2392  *result = SCIP_DELAYED;
2393 
2394  return SCIP_OKAY;
2395  }
2396 
2397  allrewardsmode = heurdata->rewardfile != NULL;
2398 
2399  /* apply some other rules for a fair all rewards mode; in normal execution mode, neighborhoods are iterated through */
2400  if( allrewardsmode )
2401  {
2402  /* most neighborhoods require an incumbent solution */
2403  if( SCIPgetNSols(scip) < 2 )
2404  {
2405  SCIPdebugMsg(scip, "Not enough solutions for all rewards mode\n");
2406  return SCIP_OKAY;
2407  }
2408 
2409  /* if the node is infeasible, or has no LP solution, which is required by some neighborhoods
2410  * if we are not in all rewards mode, the neighborhoods delay themselves individually
2411  */
2412  if( nodeinfeasible || ! SCIPhasCurrentNodeLP(scip) || SCIPgetLPSolstat(scip) != SCIP_LPSOLSTAT_OPTIMAL )
2413  {
2414  SCIPdebugMsg(scip, "Delay ALNS heuristic until a feasible node with optimally solved LP relaxation\n");
2415  *result = SCIP_DELAYED;
2416  return SCIP_OKAY;
2417  }
2418  }
2419 
2420  /* use the neighborhood that requested a delay or select the next neighborhood to run based on the selected bandit algorithm */
2421  if( heurdata->currneighborhood >= 0 )
2422  {
2423  assert(! allrewardsmode);
2424  banditidx = heurdata->currneighborhood;
2425  SCIPdebugMsg(scip, "Select delayed neighborhood %d (was delayed %d times)\n", banditidx, heurdata->ndelayedcalls);
2426  }
2427  else
2428  {
2429  SCIP_CALL( selectNeighborhood(scip, heurdata, &banditidx) );
2430  SCIPdebugMsg(scip, "Selected neighborhood %d with bandit algorithm\n", banditidx);
2431  }
2432 
2433  /* in all rewards mode, we simply loop over all heuristics */
2434  if( ! allrewardsmode )
2435  neighborhoodidx = banditidx;
2436  else
2437  neighborhoodidx = 0;
2438 
2439  assert(0 <= neighborhoodidx && neighborhoodidx < NNEIGHBORHOODS);
2440  assert(heurdata->nactiveneighborhoods > neighborhoodidx);
2441 
2442  /* allocate memory for variable fixings buffer */
2443  SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, NULL, NULL, NULL, NULL) );
2444  SCIP_CALL( SCIPallocBufferArray(scip, &varbuf, nvars) );
2445  SCIP_CALL( SCIPallocBufferArray(scip, &valbuf, nvars) );
2446  SCIP_CALL( SCIPallocBufferArray(scip, &subvars, nvars) );
2447 
2448  /* initialize neighborhood statistics for a run */
2449  ntries = 1;
2450  do
2451  {
2452  SCIP_HASHMAP* varmapf;
2453  SCIP_EVENTHDLR* eventhdlr;
2454  SCIP_EVENTDATA eventdata;
2455  char probnamesuffix[SCIP_MAXSTRLEN];
2456  SCIP_Real allfixingrate;
2457  int ndomchgs;
2458  int nchgobjs;
2459  int naddedconss;
2460  int v;
2461  SCIP_RETCODE retcode;
2462  SCIP_RESULT fixresult;
2463 
2464  tryagain = FALSE;
2465  neighborhood = heurdata->neighborhoods[neighborhoodidx];
2466  SCIPdebugMsg(scip, "Running '%s' neighborhood %d\n", neighborhood->name, neighborhoodidx);
2467 
2468  initRunStats(scip, &runstats[neighborhoodidx]);
2469  rewards[neighborhoodidx][REWARDTYPE_TOTAL] = 0.0;
2470 
2471  subscipstatus[neighborhoodidx] = SCIP_STATUS_UNKNOWN;
2472  SCIP_CALL( SCIPstartClock(scip, neighborhood->stats.setupclock) );
2473 
2474  /* determine variable fixings and objective coefficients of this neighborhood */
2475  SCIP_CALL( neighborhoodFixVariables(scip, heurdata, neighborhood, varbuf, valbuf, &nfixings, &fixresult) );
2476 
2477  SCIPdebugMsg(scip, "Fix %d/%d variables, result code %d\n", nfixings, nvars,fixresult);
2478 
2479  /* Fixing was not successful, either because the fixing rate was not reached (and no additional variable
2480  * prioritization was used), or the neighborhood requested a delay, e.g., because no LP relaxation solution exists
2481  * at the current node
2482  *
2483  * The ALNS heuristic keeps a delayed neighborhood active and delays itself.
2484  */
2485  if( fixresult != SCIP_SUCCESS )
2486  {
2487  SCIP_CALL( SCIPstopClock(scip, neighborhood->stats.setupclock) );
2488 
2489  /* to determine all rewards, we cannot delay neighborhoods */
2490  if( allrewardsmode )
2491  {
2492  if( ntries == heurdata->nactiveneighborhoods )
2493  break;
2494 
2495  neighborhoodidx = (neighborhoodidx + 1) % heurdata->nactiveneighborhoods;
2496  ntries++;
2497  tryagain = TRUE;
2498 
2499  continue;
2500  }
2501 
2502  /* delay the heuristic along with the selected neighborhood
2503  *
2504  * if the neighborhood has been delayed for too many consecutive calls, the delay is treated as a failure */
2505  if( fixresult == SCIP_DELAYED )
2506  {
2507  if( heurdata->ndelayedcalls > (SCIPheurGetFreq(heur) / 4 + 1) )
2508  {
2509  resetCurrentNeighborhood(heurdata);
2510 
2511  /* use SCIP_DIDNOTFIND to penalize the neighborhood with a bad reward */
2512  fixresult = SCIP_DIDNOTFIND;
2513  }
2514  else if( heurdata->currneighborhood == -1 )
2515  {
2516  heurdata->currneighborhood = neighborhoodidx;
2517  heurdata->ndelayedcalls = 1;
2518  }
2519  else
2520  {
2521  heurdata->ndelayedcalls++;
2522  }
2523  }
2524 
2525  if( fixresult == SCIP_DIDNOTRUN )
2526  {
2527  if( ntries < heurdata->nactiveneighborhoods )
2528  {
2529  SCIP_CALL( updateBanditAlgorithm(scip, heurdata, 0.0, neighborhoodidx) );
2530  SCIP_CALL( selectNeighborhood(scip, heurdata, &neighborhoodidx) );
2531  ntries++;
2532  tryagain = TRUE;
2533 
2534  SCIPdebugMsg(scip, "Neighborhood cannot run -> try next neighborhood %d\n", neighborhoodidx);
2535  continue;
2536  }
2537  else
2538  break;
2539  }
2540 
2541  assert(fixresult == SCIP_DIDNOTFIND || fixresult == SCIP_DELAYED);
2542  *result = fixresult;
2543  break;
2544  }
2545 
2546  *result = SCIP_DIDNOTFIND;
2547 
2548  neighborhood->stats.nfixings += nfixings;
2549  runstats[neighborhoodidx].nfixings = nfixings;
2550 
2551  SCIP_CALL( SCIPcreate(&subscip) );
2552  SCIP_CALL( SCIPhashmapCreate(&varmapf, SCIPblkmem(scip), nvars) );
2553  (void) SCIPsnprintf(probnamesuffix, SCIP_MAXSTRLEN, "alns_%s", neighborhood->name);
2554 
2555  /* todo later: run global propagation for this set of fixings */
2556  SCIP_CALL( SCIPcopyLargeNeighborhoodSearch(scip, subscip, varmapf, probnamesuffix, varbuf, valbuf, nfixings, FALSE, heurdata->copycuts, &success, NULL) );
2557 
2558  /* store sub-SCIP variables in array for faster access */
2559  for( v = 0; v < nvars; ++v )
2560  {
2561  subvars[v] = (SCIP_VAR*)SCIPhashmapGetImage(varmapf, (void *)vars[v]);
2562  }
2563 
2564  SCIPhashmapFree(&varmapf);
2565 
2566  /* let the neighborhood add additional constraints, or restrict domains */
2567  SCIP_CALL( neighborhoodChangeSubscip(scip, subscip, neighborhood, subvars, &ndomchgs, &nchgobjs, &naddedconss, &success) );
2568 
2569  if( ! success )
2570  {
2571  SCIP_CALL( SCIPstopClock(scip, neighborhood->stats.setupclock) );
2572 
2573  if( ! allrewardsmode || ntries == heurdata->nactiveneighborhoods )
2574  break;
2575 
2576  neighborhoodidx = (neighborhoodidx + 1) % heurdata->nactiveneighborhoods;
2577  ntries++;
2578  tryagain = TRUE;
2579 
2580  SCIP_CALL( SCIPfree(&subscip) );
2581 
2582  continue;
2583  }
2584 
2585  /* set up sub-SCIP parameters */
2586  SCIP_CALL( setupSubScip(scip, subscip, subvars, &solvelimits, heur, nchgobjs > 0) );
2587 
2588  /* copy the necessary data into the event data to create new solutions */
2589  eventdata.nodelimit = solvelimits.nodelimit; /*lint !e644*/
2590  eventdata.lplimfac = heurdata->lplimfac;
2591  eventdata.heur = heur;
2592  eventdata.sourcescip = scip;
2593  eventdata.subvars = subvars;
2594  eventdata.runstats = &runstats[neighborhoodidx];
2595  eventdata.allrewardsmode = allrewardsmode;
2596 
2597  /* include an event handler to transfer solutions into the main SCIP */
2598  SCIP_CALL( SCIPincludeEventhdlrBasic(subscip, &eventhdlr, EVENTHDLR_NAME, EVENTHDLR_DESC, eventExecAlns, NULL) );
2599 
2600  /* transform the problem before catching the events */
2601  SCIP_CALL( SCIPtransformProb(subscip) );
2602  SCIP_CALL( SCIPcatchEvent(subscip, SCIP_EVENTTYPE_ALNS, eventhdlr, &eventdata, NULL) );
2603 
2604  SCIP_CALL( SCIPstopClock(scip, neighborhood->stats.setupclock) );
2605 
2606  SCIP_CALL( SCIPstartClock(scip, neighborhood->stats.submipclock) );
2607 
2608  /* set up sub-SCIP and run presolving */
2609  retcode = SCIPpresolve(subscip);
2610  if( retcode != SCIP_OKAY )
2611  {
2612  SCIPwarningMessage(scip, "Error while presolving subproblem in ALNS heuristic; sub-SCIP terminated with code <%d>\n", retcode);
2613  SCIP_CALL( SCIPstopClock(scip, neighborhood->stats.submipclock) );
2614 
2615  SCIPABORT(); /*lint --e{527}*/
2616  break;
2617  }
2618 
2619  /* was presolving successful enough regarding fixings? otherwise, terminate */
2620  allfixingrate = (SCIPgetNOrigVars(subscip) - SCIPgetNVars(subscip)) / (SCIP_Real)SCIPgetNOrigVars(subscip);
2621 
2622  /* additional variables added in presolving may lead to the subSCIP having more variables than the original */
2623  allfixingrate = MAX(allfixingrate, 0.0);
2624 
2625  if( allfixingrate >= neighborhood->fixingrate.targetfixingrate / 2.0 )
2626  {
2627  /* run sub-SCIP for the given budget, and collect statistics */
2628  SCIP_CALL_ABORT( SCIPsolve(subscip) );
2629  }
2630  else
2631  {
2632  SCIPdebugMsg(scip, "Fixed only %.3f of all variables after presolving -> do not solve sub-SCIP\n", allfixingrate);
2633  }
2634 
2635 #ifdef ALNS_SUBSCIPOUTPUT
2636  SCIP_CALL( SCIPprintStatistics(subscip, NULL) );
2637 #endif
2638 
2639  SCIP_CALL( SCIPstopClock(scip, neighborhood->stats.submipclock) );
2640 
2641  /* update statistics based on the sub-SCIP run results */
2642  updateRunStats(&runstats[neighborhoodidx], subscip);
2643  subscipstatus[neighborhoodidx] = SCIPgetStatus(subscip);
2644  SCIPdebugMsg(scip, "Status of sub-SCIP run: %d\n", subscipstatus[neighborhoodidx]);
2645 
2646  SCIP_CALL( getReward(scip, heurdata, &runstats[neighborhoodidx], rewards[neighborhoodidx]) );
2647 
2648  /* in all rewards mode, continue with the next neighborhood */
2649  if( allrewardsmode && ntries < heurdata->nactiveneighborhoods )
2650  {
2651  neighborhoodidx = (neighborhoodidx + 1) % heurdata->nactiveneighborhoods;
2652  ntries++;
2653  tryagain = TRUE;
2654 
2655  SCIP_CALL( SCIPfree(&subscip) );
2656  }
2657  }
2658  while( tryagain && ! SCIPisStopped(scip) );
2659 
2660  if( subscip != NULL )
2661  {
2662  SCIP_CALL( SCIPfree(&subscip) );
2663  }
2664 
2665  SCIPfreeBufferArray(scip, &subvars);
2666  SCIPfreeBufferArray(scip, &valbuf);
2667  SCIPfreeBufferArray(scip, &varbuf);
2668 
2669  /* update bandit index that may have changed unless we are in all rewards mode */
2670  if( ! allrewardsmode )
2671  banditidx = neighborhoodidx;
2672 
2673  if( *result != SCIP_DELAYED )
2674  {
2675  /* decrease the number of neighborhoods that have not been initialized */
2676  if( neighborhood->stats.nruns == 0 )
2677  --heurdata->ninitneighborhoods;
2678 
2679  heurdata->usednodes += runstats[banditidx].usednodes;
2680 
2681  /* determine the success of this neighborhood, and update the target fixing rate for the next time */
2682  updateNeighborhoodStats(&runstats[banditidx], heurdata->neighborhoods[banditidx], subscipstatus[banditidx]);
2683 
2684  /* adjust the fixing rate for this neighborhood
2685  * make no adjustments in all rewards mode, because this only affects 1 of 8 heuristics
2686  */
2687  if( heurdata->adjustfixingrate && ! allrewardsmode )
2688  {
2689  SCIPdebugMsg(scip, "Update fixing rate: %.2f\n", heurdata->neighborhoods[banditidx]->fixingrate.targetfixingrate);
2690  updateFixingRate(heurdata->neighborhoods[banditidx], subscipstatus[banditidx], &runstats[banditidx]);
2691  SCIPdebugMsg(scip, "New fixing rate: %.2f\n", heurdata->neighborhoods[banditidx]->fixingrate.targetfixingrate);
2692  }
2693  /* similarly, update the minimum improvement for the ALNS heuristic */
2694  if( heurdata->adjustminimprove )
2695  {
2696  SCIPdebugMsg(scip, "Update Minimum Improvement: %.4f\n", heurdata->minimprove);
2697  updateMinimumImprovement(heurdata, subscipstatus[banditidx], &runstats[banditidx]);
2698  SCIPdebugMsg(scip, "--> %.4f\n", heurdata->minimprove);
2699  }
2700 
2701  /* update the target node limit based on the status of the selected algorithm */
2702  if( heurdata->adjusttargetnodes && SCIPheurGetNCalls(heur) >= heurdata->nactiveneighborhoods )
2703  {
2704  updateTargetNodeLimit(heurdata, &runstats[banditidx], subscipstatus[banditidx]);
2705  }
2706 
2707  /* update the bandit algorithm by the measured reward */
2708  SCIP_CALL( updateBanditAlgorithm(scip, heurdata, rewards[banditidx][REWARDTYPE_TOTAL], banditidx) );
2709 
2710  resetCurrentNeighborhood(heurdata);
2711  }
2712 
2713  /* write single, measured rewards and the bandit index to the reward file */
2714  if( allrewardsmode )
2715  {
2716  int j;
2717  for( j = 0; j < (int)NREWARDTYPES; j++ )
2718  for( i = 0; i < heurdata->nactiveneighborhoods; ++i )
2719  fprintf(heurdata->rewardfile, "%.4f,", rewards[i][j]);
2720 
2721  fprintf(heurdata->rewardfile, "%d\n", banditidx);
2722  }
2723 
2724  return SCIP_OKAY;
2725 }
2726 
2727 /** callback to collect variable fixings of RENS */
2728 static
2729 DECL_VARFIXINGS(varFixingsRens)
2730 { /*lint --e{715}*/
2731  int nbinvars;
2732  int nintvars;
2733  SCIP_VAR** vars;
2734  int i;
2735  int *fracidx = NULL;
2736  SCIP_Real* frac = NULL;
2737  int nfracs;
2738 
2739  assert(scip != NULL);
2740  assert(varbuf != NULL);
2741  assert(nfixings != NULL);
2742  assert(valbuf != NULL);
2743 
2744  *result = SCIP_DELAYED;
2745 
2746  if( ! SCIPhasCurrentNodeLP(scip) )
2747  return SCIP_OKAY;
2749  return SCIP_OKAY;
2750 
2751  *result = SCIP_DIDNOTRUN;
2752 
2753  /* get variable information */
2754  SCIP_CALL( SCIPgetVarsData(scip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
2755 
2756  /* return if no binary or integer variables are present */
2757  if( nbinvars + nintvars == 0 )
2758  return SCIP_OKAY;
2759 
2760  SCIP_CALL( SCIPallocBufferArray(scip, &fracidx, nbinvars + nintvars) );
2761  SCIP_CALL( SCIPallocBufferArray(scip, &frac, nbinvars + nintvars) );
2762 
2763  /* loop over binary and integer variables; determine those that should be fixed in the sub-SCIP */
2764  for( nfracs = 0, i = 0; i < nbinvars + nintvars; ++i )
2765  {
2766  SCIP_VAR* var = vars[i];
2767  SCIP_Real lpsolval = SCIPvarGetLPSol(var);
2768  assert((i < nbinvars && SCIPvarIsBinary(var)) || (i >= nbinvars && SCIPvarIsIntegral(var)));
2769 
2770  /* fix all binary and integer variables with integer LP solution value */
2771  if( SCIPisFeasIntegral(scip, lpsolval) )
2772  {
2773  tryAdd2variableBuffer(scip, var, lpsolval, varbuf, valbuf, nfixings, TRUE);
2774  }
2775  else
2776  {
2777  frac[nfracs] = SCIPfrac(scip, lpsolval);
2778  frac[nfracs] = MIN(frac[nfracs], 1.0 - frac[nfracs]);
2779  fracidx[nfracs++] = i;
2780  }
2781  }
2782 
2783  /* do some additional fixing */
2784  if( *nfixings < neighborhood->fixingrate.targetfixingrate * (nbinvars + nintvars) && nfracs > 0 )
2785  {
2786  SCIPsortDownRealInt(frac, fracidx, nfracs);
2787 
2788  /* prefer variables that are almost integer */
2789  for( i = 0; i < nfracs && *nfixings < neighborhood->fixingrate.targetfixingrate * (nbinvars + nintvars); i++ )
2790  {
2791  tryAdd2variableBuffer(scip, vars[fracidx[i]], SCIPround(scip, SCIPvarGetLPSol(vars[fracidx[i]])), varbuf, valbuf, nfixings, TRUE);
2792  }
2793  }
2794 
2795  SCIPfreeBufferArray(scip, &frac);
2796  SCIPfreeBufferArray(scip, &fracidx);
2797 
2798  *result = SCIP_SUCCESS;
2799 
2800  return SCIP_OKAY;
2801 }
2802 
2803 /** callback for RENS subproblem changes */
2804 static
2805 DECL_CHANGESUBSCIP(changeSubscipRens)
2806 { /*lint --e{715}*/
2807  SCIP_VAR** vars;
2808  int nintvars;
2809  int nbinvars;
2810  int i;
2811 
2812  assert(SCIPhasCurrentNodeLP(sourcescip));
2813  assert(SCIPgetLPSolstat(sourcescip) == SCIP_LPSOLSTAT_OPTIMAL);
2814 
2815  /* get variable information */
2816  SCIP_CALL( SCIPgetVarsData(sourcescip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
2817 
2818  /* restrict bounds of integer variables with fractional solution value */
2819  for( i = nbinvars; i < nbinvars + nintvars; ++i )
2820  {
2821  SCIP_VAR* var = vars[i];
2822  SCIP_Real lpsolval = SCIPgetSolVal(sourcescip, NULL, var);
2823 
2824  if( subvars[i] == NULL )
2825  continue;
2826 
2827  if( ! SCIPisFeasIntegral(sourcescip, lpsolval) )
2828  {
2829  SCIP_Real newlb = SCIPfloor(sourcescip, lpsolval);
2830  SCIP_Real newub = newlb + 1.0;
2831 
2832  /* only count this as a domain change if the new lower and upper bound are a further restriction */
2833  if( newlb > SCIPvarGetLbGlobal(subvars[i]) + 0.5 || newub < SCIPvarGetUbGlobal(subvars[i]) - 0.5 )
2834  {
2835  SCIP_CALL( SCIPchgVarLbGlobal(targetscip, subvars[i], newlb) );
2836  SCIP_CALL( SCIPchgVarUbGlobal(targetscip, subvars[i], newub) );
2837  (*ndomchgs)++;
2838  }
2839  }
2840  }
2841 
2842  *success = TRUE;
2843 
2844  return SCIP_OKAY;
2845 }
2846 
2847 /** collect fixings by matching solution values in a collection of solutions for all binary and integer variables,
2848  * or for a custom set of variables
2849  */
2850 static
2852  SCIP* scip, /**< SCIP data structure */
2853  SCIP_SOL** sols, /**< array of 2 or more solutions. It is okay for the array to contain one element
2854  * equal to NULL to represent the current LP solution */
2855  int nsols, /**< number of solutions in the array */
2856  SCIP_VAR** vars, /**< variable array for which solution values must agree */
2857  int nvars, /**< number of variables, or -1 for all binary and integer variables */
2858  SCIP_VAR** varbuf, /**< buffer storage for variable fixings */
2859  SCIP_Real* valbuf, /**< buffer storage for fixing values */
2860  int* nfixings /**< pointer to store the number of fixings */
2861  )
2862 {
2863  int v;
2864  int nbinintvars;
2865  SCIP_SOL* firstsol;
2866 
2867  assert(scip != NULL);
2868  assert(sols != NULL);
2869  assert(nsols >= 2);
2870  assert(varbuf != NULL);
2871  assert(valbuf != NULL);
2872  assert(nfixings != NULL);
2873  assert(*nfixings == 0);
2874 
2875  if( nvars == -1 || vars == NULL )
2876  {
2877  int nbinvars;
2878  int nintvars;
2879  SCIP_CALL( SCIPgetVarsData(scip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
2880  nbinintvars = nbinvars + nintvars;
2881  nvars = nbinintvars;
2882  }
2883  firstsol = sols[0];
2884  assert(nvars > 0);
2885 
2886  /* loop over integer and binary variables and check if their solution values match in all solutions */
2887  for( v = 0; v < nvars; ++v )
2888  {
2889  SCIP_Real solval;
2890  SCIP_VAR* var;
2891  int s;
2892 
2893  var = vars[v];
2894  assert((v < SCIPgetNBinVars(scip) && SCIPvarIsBinary(var)) || (v >= SCIPgetNBinVars(scip) && SCIPvarIsIntegral(var)));
2895  solval = SCIPgetSolVal(scip, firstsol, var);
2896 
2897  /* determine if solution values match in all given solutions */
2898  for( s = 1; s < nsols; ++s )
2899  {
2900  SCIP_Real solval2 = SCIPgetSolVal(scip, sols[s], var);
2901  if( ! SCIPisEQ(scip, solval, solval2) )
2902  break;
2903  }
2904 
2905  /* if we did not break early, all solutions agree on the solution value of this variable */
2906  if( s == nsols )
2907  {
2908  tryAdd2variableBuffer(scip, var, solval, varbuf, valbuf, nfixings, TRUE);
2909  }
2910  }
2911 
2912  return SCIP_OKAY;
2913 }
2914 
2915 /** callback to collect variable fixings of RINS */
2916 static
2917 DECL_VARFIXINGS(varFixingsRins)
2919  /*lint --e{715}*/
2920  int nbinvars;
2921  int nintvars;
2922  SCIP_VAR** vars;
2923  SCIP_SOL* incumbent;
2924  SCIP_SOL* sols[2];
2925  assert(scip != NULL);
2926  assert(varbuf != NULL);
2927  assert(nfixings != NULL);
2928  assert(valbuf != NULL);
2929 
2930  *result = SCIP_DELAYED;
2931 
2932  if( ! SCIPhasCurrentNodeLP(scip) )
2933  return SCIP_OKAY;
2935  return SCIP_OKAY;
2936 
2937  *result = SCIP_DIDNOTRUN;
2938 
2939  incumbent = SCIPgetBestSol(scip);
2940  if( incumbent == NULL )
2941  return SCIP_OKAY;
2942 
2943  if( SCIPsolGetOrigin(incumbent) == SCIP_SOLORIGIN_ORIGINAL )
2944  return SCIP_OKAY;
2945 
2946  /* get variable information */
2947  SCIP_CALL( SCIPgetVarsData(scip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
2948 
2949  /* return if no binary or integer variables are present */
2950  if( nbinvars + nintvars == 0 )
2951  return SCIP_OKAY;
2952 
2953  sols[0] = NULL;
2954  sols[1] = incumbent;
2955 
2956  SCIP_CALL( fixMatchingSolutionValues(scip, sols, 2, vars, nbinvars + nintvars, varbuf, valbuf, nfixings) );
2957 
2958  *result = SCIP_SUCCESS;
2959 
2960  return SCIP_OKAY;
2961 }
2962 
2963 /** initialization callback for crossover when a new problem is read */
2964 static
2965 DECL_NHINIT(nhInitCrossover)
2966 { /*lint --e{715}*/
2967  DATA_CROSSOVER* data;
2968 
2969  data = neighborhood->data.crossover;
2970  assert(data != NULL);
2971 
2972  if( data->rng != NULL )
2973  SCIPfreeRandom(scip, &data->rng);
2974 
2975  data->selsol = NULL;
2976 
2977  SCIP_CALL( SCIPcreateRandom(scip, &data->rng, CROSSOVERSEED + (unsigned int)SCIPgetNVars(scip), TRUE) );
2978 
2979  return SCIP_OKAY;
2980 }
2981 
2982 /** deinitialization callback for crossover when exiting a problem */
2983 static
2984 DECL_NHEXIT(nhExitCrossover)
2985 { /*lint --e{715}*/
2986  DATA_CROSSOVER* data;
2987  data = neighborhood->data.crossover;
2988 
2989  assert(neighborhood != NULL);
2990  assert(data->rng != NULL);
2991 
2992  SCIPfreeRandom(scip, &data->rng);
2993 
2994  return SCIP_OKAY;
2995 }
2996 
2997 /** deinitialization callback for crossover before SCIP is freed */
2998 static
2999 DECL_NHFREE(nhFreeCrossover)
3000 { /*lint --e{715}*/
3001  assert(neighborhood->data.crossover != NULL);
3002  SCIPfreeBlockMemory(scip, &neighborhood->data.crossover);
3003 
3004  return SCIP_OKAY;
3005 }
3006 
3007 /** callback to collect variable fixings of crossover */
3008 static
3009 DECL_VARFIXINGS(varFixingsCrossover)
3010 { /*lint --e{715}*/
3011  DATA_CROSSOVER* data;
3012  SCIP_RANDNUMGEN* rng;
3013  SCIP_SOL** sols;
3014  SCIP_SOL** scipsols;
3015  int nsols;
3016  int lastdraw;
3017  assert(scip != NULL);
3018  assert(varbuf != NULL);
3019  assert(nfixings != NULL);
3020  assert(valbuf != NULL);
3021 
3022  data = neighborhood->data.crossover;
3023 
3024  assert(data != NULL);
3025  nsols = data->nsols;
3026  data->selsol = NULL;
3027 
3028  *result = SCIP_DIDNOTRUN;
3029 
3030  /* return if the pool has not enough solutions */
3031  if( nsols > SCIPgetNSols(scip) )
3032  return SCIP_OKAY;
3033 
3034  /* return if no binary or integer variables are present */
3035  if( SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip) == 0 )
3036  return SCIP_OKAY;
3037 
3038  rng = data->rng;
3039  lastdraw = SCIPgetNSols(scip);
3040  SCIP_CALL( SCIPallocBufferArray(scip, &sols, nsols) );
3041  scipsols = SCIPgetSols(scip);
3042 
3043  /* draw as many solutions from the pool as required by crossover, biased towards
3044  * better solutions; therefore, the sorting of the solutions by objective is implicitly used
3045  */
3046  while( nsols > 0 )
3047  {
3048  /* no need for randomization anymore, exactly nsols many solutions remain for the selection */
3049  if( lastdraw == nsols )
3050  {
3051  int s;
3052 
3053  /* fill the remaining slots 0,...,nsols - 1 by the solutions at the same places */
3054  for( s = 0; s < nsols; ++s )
3055  sols[s] = scipsols[s];
3056 
3057  nsols = 0;
3058  }
3059  else
3060  {
3061  int nextdraw;
3062 
3063  assert(nsols < lastdraw);
3064 
3065  /* draw from the lastdraw - nsols many solutions nsols - 1, ... lastdraw - 1 such that nsols many solution */
3066  nextdraw = SCIPrandomGetInt(rng, nsols - 1, lastdraw - 1);
3067  assert(nextdraw >= 0);
3068 
3069  sols[nsols - 1] = scipsols[nextdraw];
3070  nsols--;
3071  lastdraw = nextdraw;
3072  }
3073  }
3074 
3075  SCIP_CALL( fixMatchingSolutionValues(scip, sols, data->nsols, NULL, -1, varbuf, valbuf, nfixings) );
3076 
3077  /* store best selected solution as reference solution */
3078  data->selsol = sols[0];
3079  assert(data->selsol != NULL);
3080 
3081  *result = SCIP_SUCCESS;
3082 
3083  SCIPfreeBufferArray(scip, &sols);
3084 
3085  return SCIP_OKAY;
3086 }
3087 
3088 /** callback for crossover reference solution */
3089 static
3090 DECL_NHREFSOL(nhRefsolCrossover)
3091 { /*lint --e{715}*/
3092  DATA_CROSSOVER* data;
3093 
3094  data = neighborhood->data.crossover;
3095 
3096  if( data->selsol != NULL )
3097  {
3098  *solptr = data->selsol;
3099  *result = SCIP_SUCCESS;
3100  }
3101  else
3102  {
3103  *result = SCIP_DIDNOTFIND;
3104  }
3105 
3106  return SCIP_OKAY;
3107 }
3108 
3109 /** initialization callback for mutation when a new problem is read */
3110 static
3111 DECL_NHINIT(nhInitMutation)
3112 { /*lint --e{715}*/
3113  DATA_MUTATION* data;
3114  assert(scip != NULL);
3115  assert(neighborhood != NULL);
3116 
3117  SCIP_CALL( SCIPallocBlockMemory(scip, &neighborhood->data.mutation) );
3118 
3119  data = neighborhood->data.mutation;
3120  assert(data != NULL);
3121 
3122  SCIP_CALL( SCIPcreateRandom(scip, &data->rng, MUTATIONSEED + (unsigned int)SCIPgetNVars(scip), TRUE) );
3123 
3124  return SCIP_OKAY;
3125 }
3126 
3127 /** deinitialization callback for mutation when exiting a problem */
3128 static
3129 DECL_NHEXIT(nhExitMutation)
3130 { /*lint --e{715}*/
3131  DATA_MUTATION* data;
3132  assert(scip != NULL);
3133  assert(neighborhood != NULL);
3134  data = neighborhood->data.mutation;
3135  assert(data != NULL);
3136 
3137  SCIPfreeRandom(scip, &data->rng);
3138 
3139  SCIPfreeBlockMemory(scip, &neighborhood->data.mutation);
3140 
3141  return SCIP_OKAY;
3142 }
3143 
3144 /** callback to collect variable fixings of mutation */
3145 static
3146 DECL_VARFIXINGS(varFixingsMutation)
3147 { /*lint --e{715}*/
3148  SCIP_RANDNUMGEN* rng;
3149 
3150  SCIP_VAR** vars;
3151  SCIP_VAR** varscpy;
3152  int i;
3153  int nvars;
3154  int nbinvars;
3155  int nintvars;
3156  int nbinintvars;
3157  int ntargetfixings;
3158  SCIP_SOL* incumbentsol;
3159  SCIP_Real targetfixingrate;
3160 
3161  assert(scip != NULL);
3162  assert(neighborhood != NULL);
3163  assert(neighborhood->data.mutation != NULL);
3164  assert(neighborhood->data.mutation->rng != NULL);
3165  rng = neighborhood->data.mutation->rng;
3166 
3167  *result = SCIP_DIDNOTRUN;
3168 
3169  /* get the problem variables */
3170  SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, &nbinvars, &nintvars, NULL, NULL) );
3171 
3172  nbinintvars = nbinvars + nintvars;
3173  if( nbinintvars == 0 )
3174  return SCIP_OKAY;
3175 
3176  incumbentsol = SCIPgetBestSol(scip);
3177  if( incumbentsol == NULL )
3178  return SCIP_OKAY;
3179 
3180  targetfixingrate = neighborhood->fixingrate.targetfixingrate;
3181  ntargetfixings = (int)(targetfixingrate * nbinintvars) + 1;
3182 
3183  /* don't continue if number of discrete variables is too small to reach target fixing rate */
3184  if( nbinintvars <= ntargetfixings )
3185  return SCIP_OKAY;
3186 
3187  *result = SCIP_DIDNOTFIND;
3188 
3189  /* copy variables into a buffer array */
3190  SCIP_CALL( SCIPduplicateBufferArray(scip, &varscpy, vars, nbinintvars) );
3191 
3192  /* partially perturb the array until the number of target fixings is reached */
3193  for( i = 0; *nfixings < ntargetfixings && i < nbinintvars; ++i )
3194  {
3195  int randint = SCIPrandomGetInt(rng, i, nbinintvars - 1);
3196  assert(randint < nbinintvars);
3197 
3198  if( randint > i )
3199  {
3200  SCIPswapPointers((void**)&varscpy[i], (void**)&varscpy[randint]);
3201  }
3202  /* copy the selected variables and their solution values into the buffer */
3203  tryAdd2variableBuffer(scip, varscpy[i], SCIPgetSolVal(scip, incumbentsol, varscpy[i]), varbuf, valbuf, nfixings, TRUE);
3204  }
3205 
3206  assert(i == nbinintvars || *nfixings == ntargetfixings);
3207 
3208  /* Not reaching the number of target fixings means that there is a significant fraction (at least 1 - targetfixingrate)
3209  * of variables for which the incumbent solution value does not lie within the global bounds anymore. This is a nonsuccess
3210  * for the neighborhood (additional fixings are not possible), which is okay because the incumbent solution is
3211  * significantly outdated
3212  */
3213  if( *nfixings == ntargetfixings )
3214  *result = SCIP_SUCCESS;
3215 
3216  /* free the buffer array */
3217  SCIPfreeBufferArray(scip, &varscpy);
3218 
3219  return SCIP_OKAY;
3220 }
3221 
3222 /** add local branching constraint */
3223 static
3225  SCIP* sourcescip, /**< source SCIP data structure */
3226  SCIP* targetscip, /**< target SCIP data structure */
3227  SCIP_VAR** subvars, /**< array of sub SCIP variables in same order as source SCIP variables */
3228  int distance, /**< right hand side of the local branching constraint */
3229  SCIP_Bool* success, /**< pointer to store of a local branching constraint has been successfully added */
3230  int* naddedconss /**< pointer to increase the number of added constraints */
3231  )
3232 {
3233  int nbinvars;
3234  int i;
3235  SCIP_SOL* referencesol;
3236  SCIP_CONS* localbranchcons;
3237  SCIP_VAR** vars;
3238  SCIP_Real* consvals;
3239  SCIP_Real rhs;
3240 
3241  assert(sourcescip != NULL);
3242  assert(*success == FALSE);
3243 
3244  nbinvars = SCIPgetNBinVars(sourcescip);
3245  vars = SCIPgetVars(sourcescip);
3246 
3247  if( nbinvars <= 3 )
3248  return SCIP_OKAY;
3249 
3250  referencesol = SCIPgetBestSol(sourcescip);
3251  if( referencesol == NULL )
3252  return SCIP_OKAY;
3253 
3254  rhs = (SCIP_Real)distance;
3255  rhs = MAX(rhs, 2.0);
3256 
3257  SCIP_CALL( SCIPallocBufferArray(sourcescip, &consvals, nbinvars) );
3258 
3259  /* loop over binary variables and fill the local branching constraint */
3260  for( i = 0; i < nbinvars; ++i )
3261  {
3262  /* skip variables that are not present in sub-SCIP */
3263  if( subvars[i] == NULL )
3264  continue;
3265 
3266  if( SCIPisEQ(sourcescip, SCIPgetSolVal(sourcescip, referencesol, vars[i]), 0.0) )
3267  consvals[i] = 1.0;
3268  else
3269  {
3270  consvals[i] = -1.0;
3271  rhs -= 1.0;
3272  }
3273  }
3274 
3275  /* create the local branching constraint in the target scip */
3276  SCIP_CALL( SCIPcreateConsBasicLinear(targetscip, &localbranchcons, "localbranch", nbinvars, subvars, consvals, -SCIPinfinity(sourcescip), rhs) );
3277  SCIP_CALL( SCIPaddCons(targetscip, localbranchcons) );
3278  SCIP_CALL( SCIPreleaseCons(targetscip, &localbranchcons) );
3279 
3280  *naddedconss = 1;
3281  *success = TRUE;
3282 
3283  SCIPfreeBufferArray(sourcescip, &consvals);
3284 
3285  return SCIP_OKAY;
3286 }
3287 
3288 /** callback for local branching subproblem changes */
3289 static
3290 DECL_CHANGESUBSCIP(changeSubscipLocalbranching)
3291 { /*lint --e{715}*/
3292 
3293  SCIP_CALL( addLocalBranchingConstraint(sourcescip, targetscip, subvars, (int)(0.2 * SCIPgetNBinVars(sourcescip)), success, naddedconss) );
3294 
3295  return SCIP_OKAY;
3296 }
3297 
3298 /** callback for proximity subproblem changes */
3299 static
3300 DECL_CHANGESUBSCIP(changeSubscipProximity)
3301 { /*lint --e{715}*/
3302  SCIP_SOL* referencesol;
3303  SCIP_VAR** vars;
3304  int nbinvars;
3305  int nintvars;
3306  int nvars;
3307  int i;
3308 
3309  SCIP_CALL( SCIPgetVarsData(sourcescip, &vars, &nvars, &nbinvars, &nintvars, NULL, NULL) );
3310 
3311  if( nbinvars == 0 )
3312  return SCIP_OKAY;
3313 
3314  referencesol = SCIPgetBestSol(sourcescip);
3315  if( referencesol == NULL )
3316  return SCIP_OKAY;
3317 
3318  /* loop over binary variables, set objective coefficients based on reference solution in a local branching fashion */
3319  for( i = 0; i < nbinvars; ++i )
3320  {
3321  SCIP_Real newobj;
3322 
3323  /* skip variables not present in sub-SCIP */
3324  if( subvars[i] == NULL )
3325  continue;
3326 
3327  if( SCIPgetSolVal(sourcescip, referencesol, vars[i]) < 0.5 )
3328  newobj = -1.0;
3329  else
3330  newobj = 1.0;
3331  SCIP_CALL( SCIPchgVarObj(targetscip, subvars[i], newobj) );
3332  }
3333 
3334  /* loop over the remaining variables and change their objective coefficients to 0 */
3335  for( ; i < nvars; ++i )
3336  {
3337  /* skip variables not present in sub-SCIP */
3338  if( subvars[i] == NULL )
3339  continue;
3340 
3341  SCIP_CALL( SCIPchgVarObj(targetscip, subvars[i], 0.0) );
3342  }
3343 
3344  *nchgobjs = nvars;
3345  *success = TRUE;
3346 
3347  return SCIP_OKAY;
3348 }
3349 
3350 /** callback for zeroobjective subproblem changes */
3351 static
3352 DECL_CHANGESUBSCIP(changeSubscipZeroobjective)
3353 { /*lint --e{715}*/
3354  SCIP_CONSHDLR* conshdlrnl;
3355  SCIP_VAR** vars;
3356  int nvars;
3357  int i;
3358 
3359  assert(*success == FALSE);
3360 
3361  SCIP_CALL( SCIPgetVarsData(sourcescip, &vars, &nvars, NULL, NULL, NULL, NULL) );
3362 
3363  /* do not run if no objective variables are present */
3364  if( SCIPgetNObjVars(sourcescip) == 0 )
3365  return SCIP_OKAY;
3366 
3367  /* zeroobj may trigger fixing objvar in nonlinear constraint to infinity,
3368  * which expr_var.c:simplify cannot handle at the moment; also #3273
3369  */
3370  conshdlrnl = SCIPfindConshdlr(sourcescip, "nonlinear");
3371  if( conshdlrnl != NULL && SCIPconshdlrGetNActiveConss(conshdlrnl) > 0 )
3372  return SCIP_OKAY;
3373 
3374  /* loop over the variables and change their objective coefficients to 0 */
3375  for( i = 0; i < nvars; ++i )
3376  {
3377  /* skip variables not present in sub-SCIP */
3378  if( subvars[i] == NULL )
3379  continue;
3380 
3381  SCIP_CALL( SCIPchgVarObj(targetscip, subvars[i], 0.0) );
3382  }
3383 
3384  *nchgobjs = nvars;
3385  *success = TRUE;
3386 
3387  return SCIP_OKAY;
3388 }
3389 
3390 /** compute tightened bounds for integer variables depending on how much the LP and the incumbent solution values differ */
3391 static
3393  SCIP* scip, /**< SCIP data structure of the original problem */
3394  SCIP_VAR* var, /**< the variable for which bounds should be computed */
3395  SCIP_Real* lbptr, /**< pointer to store the lower bound in the DINS sub-SCIP */
3396  SCIP_Real* ubptr /**< pointer to store the upper bound in the DINS sub-SCIP */
3397  )
3398 {
3399  SCIP_Real mipsol;
3400  SCIP_Real lpsol;
3401 
3402  SCIP_Real lbglobal;
3403  SCIP_Real ubglobal;
3404  SCIP_SOL* bestsol;
3405 
3406  /* get the bounds for each variable */
3407  lbglobal = SCIPvarGetLbGlobal(var);
3408  ubglobal = SCIPvarGetUbGlobal(var);
3409 
3410  assert(SCIPvarGetType(var) == SCIP_VARTYPE_INTEGER);
3411  /* get the current LP solution for each variable */
3412  lpsol = SCIPvarGetLPSol(var);
3413 
3414  /* get the current MIP solution for each variable */
3415  bestsol = SCIPgetBestSol(scip);
3416  mipsol = SCIPgetSolVal(scip, bestsol, var);
3417 
3418  /* if the solution values differ by 0.5 or more, the variable is rebounded, otherwise it is just copied */
3419  if( REALABS(lpsol - mipsol) >= 0.5 )
3420  {
3421  SCIP_Real range;
3422 
3423  *lbptr = lbglobal;
3424  *ubptr = ubglobal;
3425 
3426  /* create an equally sized range around lpsol for general integers: bounds are lpsol +- (mipsol-lpsol) */
3427  range = 2 * lpsol - mipsol;
3428 
3429  if( mipsol >= lpsol )
3430  {
3431  range = SCIPfeasCeil(scip, range);
3432  *lbptr = MAX(*lbptr, range);
3433 
3434  /* when the bound new upper bound is equal to the current MIP solution, we set both bounds to the integral bound (without eps) */
3435  if( SCIPisFeasEQ(scip, mipsol, *lbptr) )
3436  *ubptr = *lbptr;
3437  else
3438  *ubptr = mipsol;
3439  }
3440  else
3441  {
3442  range = SCIPfeasFloor(scip, range);
3443  *ubptr = MIN(*ubptr, range);
3444 
3445  /* when the bound new upper bound is equal to the current MIP solution, we set both bounds to the integral bound (without eps) */
3446  if( SCIPisFeasEQ(scip, mipsol, *ubptr) )
3447  *lbptr = *ubptr;
3448  else
3449  *lbptr = mipsol;
3450  }
3451 
3452  /* the global domain of variables might have been reduced since incumbent was found: adjust lb and ub accordingly */
3453  *lbptr = MAX(*lbptr, lbglobal);
3454  *ubptr = MIN(*ubptr, ubglobal);
3455  }
3456  else
3457  {
3458  /* the global domain of variables might have been reduced since incumbent was found: adjust it accordingly */
3459  *lbptr = MAX(mipsol, lbglobal);
3460  *ubptr = MIN(mipsol, ubglobal);
3461  }
3462 }
3463 
3464 /** callback to collect variable fixings of DINS */
3465 static
3466 DECL_VARFIXINGS(varFixingsDins)
3468  DATA_DINS* data;
3469  SCIP_SOL* rootlpsol;
3470  SCIP_SOL** sols;
3471  int nsols;
3472  int nmipsols;
3473  int nbinvars;
3474  int nintvars;
3475  SCIP_VAR** vars;
3476  int v;
3477 
3478  data = neighborhood->data.dins;
3479  assert(data != NULL);
3480  nmipsols = SCIPgetNSols(scip);
3481  nmipsols = MIN(nmipsols, data->npoolsols);
3482 
3483  *result = SCIP_DELAYED;
3484 
3486  return SCIP_OKAY;
3487 
3488  *result = SCIP_DIDNOTRUN;
3489 
3490  if( nmipsols == 0 )
3491  return SCIP_OKAY;
3492 
3493  SCIP_CALL( SCIPgetVarsData(scip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
3494 
3495  if( nbinvars + nintvars == 0 )
3496  return SCIP_OKAY;
3497 
3498  SCIP_CALL( SCIPcreateSol(scip, &rootlpsol, NULL) );
3499 
3500  /* save root solution LP values in solution */
3501  for( v = 0; v < nbinvars + nintvars; ++v )
3502  {
3503  SCIP_CALL( SCIPsetSolVal(scip, rootlpsol, vars[v], SCIPvarGetRootSol(vars[v])) );
3504  }
3505 
3506  /* add the node and the root LP solution */
3507  nsols = nmipsols + 2;
3508 
3509  SCIP_CALL( SCIPallocBufferArray(scip, &sols, nsols) );
3510  sols[0] = NULL; /* node LP solution */
3511  sols[1] = rootlpsol;
3512 
3513  /* copy the remaining MIP solutions after the LP solutions */
3514  BMScopyMemoryArray(&sols[2], SCIPgetSols(scip), nmipsols); /*lint !e866*/
3515 
3516  /* 1. Binary variables are fixed if their values agree in all the solutions */
3517  if( nbinvars > 0 )
3518  {
3519  SCIP_CALL( fixMatchingSolutionValues(scip, sols, nsols, vars, nbinvars, varbuf, valbuf, nfixings) );
3520  }
3521 
3522  /* 2. Integer variables are fixed if they have a very low distance between the incumbent and the root LP solution */
3523  for( v = nbinvars; v < nintvars; ++v )
3524  {
3525  SCIP_Real lb;
3526  SCIP_Real ub;
3527  computeIntegerVariableBoundsDins(scip, vars[v], &lb, &ub);
3528 
3529  if( ub - lb < 0.5 )
3530  {
3531  assert(SCIPisFeasIntegral(scip, lb));
3532  tryAdd2variableBuffer(scip, vars[v], lb, varbuf, valbuf, nfixings, TRUE);
3533  }
3534  }
3535 
3536  *result = SCIP_SUCCESS;
3537 
3538  SCIPfreeBufferArray(scip, &sols);
3539 
3540  SCIP_CALL( SCIPfreeSol(scip, &rootlpsol) );
3541 
3542  return SCIP_OKAY;
3543 }
3544 
3545 /** callback for DINS subproblem changes */
3546 static
3547 DECL_CHANGESUBSCIP(changeSubscipDins)
3548 { /*lint --e{715}*/
3549  SCIP_VAR** vars;
3550  int nintvars;
3551  int nbinvars;
3552  int v;
3553 
3554  SCIP_CALL( SCIPgetVarsData(sourcescip, &vars, NULL, &nbinvars, &nintvars, NULL, NULL) );
3555 
3556  /* 1. loop over integer variables and tighten the bounds */
3557  for( v = nbinvars; v < nintvars; ++v )
3558  {
3559  SCIP_Real lb;
3560  SCIP_Real ub;
3561 
3562  /* skip variables not present in sub-SCIP */
3563  if( subvars[v] == NULL )
3564  continue;
3565 
3566  computeIntegerVariableBoundsDins(sourcescip, vars[v], &lb, &ub);
3567 
3568  SCIP_CALL( SCIPchgVarLbGlobal(targetscip, subvars[v], lb) );
3569  SCIP_CALL( SCIPchgVarUbGlobal(targetscip, subvars[v], ub) );
3570  ++(*ndomchgs);
3571  }
3572 
3573  /* 2. add local branching constraint for binary variables */
3574  SCIP_CALL( addLocalBranchingConstraint(sourcescip, targetscip, subvars, (int)(0.1 * SCIPgetNBinVars(sourcescip)), success, naddedconss) );
3575 
3576  *success = TRUE;
3577 
3578  return SCIP_OKAY;
3579 }
3580 
3581 /** deinitialization callback for DINS before SCIP is freed */
3582 static
3583 DECL_NHFREE(nhFreeDins)
3585  assert(neighborhood->data.dins != NULL);
3586 
3587  SCIPfreeBlockMemory(scip, &neighborhood->data.dins);
3588 
3589  return SCIP_OKAY;
3590 }
3591 
3592 /** deinitialization callback for trustregion before SCIP is freed */
3593 static
3594 DECL_NHFREE(nhFreeTrustregion)
3596  assert(neighborhood->data.trustregion != NULL);
3597 
3598  SCIPfreeBlockMemory(scip, &neighborhood->data.trustregion);
3599 
3600  return SCIP_OKAY;
3601 }
3602 
3603 /** add trust region neighborhood constraint and auxiliary objective variable */
3604 static
3605 DECL_CHANGESUBSCIP(changeSubscipTrustregion)
3606 { /*lint --e{715}*/
3607  DATA_TRUSTREGION* data;
3608 
3609  assert(success != NULL);
3610 
3611  if( !SCIPgetBestSol(sourcescip) )
3612  {
3613  SCIPdebugMsg(sourcescip, "changeSubscipTrustregion unsuccessful, because it was called without incumbent being present\n");
3614  *success = FALSE;
3615 
3616  return SCIP_OKAY;
3617  }
3618 
3619  data = neighborhood->data.trustregion;
3620 
3621  /* adding the neighborhood constraint for the trust region heuristic */
3622  SCIP_CALL( SCIPaddTrustregionNeighborhoodConstraint(sourcescip, targetscip, subvars, data->violpenalty) );
3623 
3624  /* incrementing the change in objective since an additional variable is added to the objective to penalize the
3625  * violation of the trust region.
3626  */
3627  ++(*nchgobjs);
3628 
3629  return SCIP_OKAY;
3630 }
3631 
3632 /** callback that returns the incumbent solution as a reference point */
3633 static
3634 DECL_NHREFSOL(nhRefsolIncumbent)
3635 { /*lint --e{715}*/
3636  assert(scip != NULL);
3637 
3638  if( SCIPgetBestSol(scip) != NULL )
3639  {
3640  *result = SCIP_SUCCESS;
3641  *solptr = SCIPgetBestSol(scip);
3642  }
3643  else
3644  {
3645  *result = SCIP_DIDNOTFIND;
3646  }
3647 
3648  return SCIP_OKAY;
3649 }
3650 
3651 
3652 /** callback function that deactivates a neighborhood on problems with no discrete variables */
3653 static
3654 DECL_NHDEACTIVATE(nhDeactivateDiscreteVars)
3655 { /*lint --e{715}*/
3656  assert(scip != NULL);
3657  assert(deactivate != NULL);
3658 
3659  /* deactivate if no discrete variables are present */
3660  *deactivate = (SCIPgetNBinVars(scip) + SCIPgetNIntVars(scip) == 0);
3661 
3662  return SCIP_OKAY;
3663 }
3664 
3665 /** callback function that deactivates a neighborhood on problems with no binary variables */
3666 static
3667 DECL_NHDEACTIVATE(nhDeactivateBinVars)
3668 { /*lint --e{715}*/
3669  assert(scip != NULL);
3670  assert(deactivate != NULL);
3671 
3672  /* deactivate if no discrete variables are present */
3673  *deactivate = (SCIPgetNBinVars(scip) == 0);
3674 
3675  return SCIP_OKAY;
3676 }
3677 
3678 /** callback function that deactivates a neighborhood on problems with no objective variables */
3679 static
3680 DECL_NHDEACTIVATE(nhDeactivateObjVars)
3681 { /*lint --e{715}*/
3682  assert(scip != NULL);
3683  assert(deactivate != NULL);
3684 
3685  /* deactivate if no discrete variables are present */
3686  *deactivate = (SCIPgetNObjVars(scip) == 0);
3687 
3688  return SCIP_OKAY;
3689 }
3690 
3691 
3692 /** include all neighborhoods */
3693 static
3695  SCIP* scip, /**< SCIP data structure */
3696  SCIP_HEURDATA* heurdata /**< heuristic data of the ALNS heuristic */
3697  )
3698 {
3699  NH* rens;
3700  NH* rins;
3701  NH* mutation;
3702  NH* localbranching;
3703  NH* crossover;
3704  NH* proximity;
3705  NH* zeroobjective;
3706  NH* dins;
3707  NH* trustregion;
3708 
3709  heurdata->nneighborhoods = 0;
3710 
3711  /* include the RENS neighborhood */
3712  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &rens, "rens",
3714  varFixingsRens, changeSubscipRens, NULL, NULL, NULL, NULL, nhDeactivateDiscreteVars) );
3715 
3716  /* include the RINS neighborhood */
3717  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &rins, "rins",
3719  varFixingsRins, NULL, NULL, NULL, NULL, nhRefsolIncumbent, nhDeactivateDiscreteVars) );
3720 
3721  /* include the mutation neighborhood */
3722  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &mutation, "mutation",
3724  varFixingsMutation, NULL, nhInitMutation, nhExitMutation, NULL, nhRefsolIncumbent, nhDeactivateDiscreteVars) );
3725 
3726  /* include the local branching neighborhood */
3727  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &localbranching, "localbranching",
3729  NULL, changeSubscipLocalbranching, NULL, NULL, NULL, nhRefsolIncumbent, nhDeactivateBinVars) );
3730 
3731  /* include the crossover neighborhood */
3732  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &crossover, "crossover",
3734  varFixingsCrossover, NULL,
3735  nhInitCrossover, nhExitCrossover, nhFreeCrossover, nhRefsolCrossover, nhDeactivateDiscreteVars) );
3736 
3737  /* allocate data for crossover to include the parameter */
3738  SCIP_CALL( SCIPallocBlockMemory(scip, &crossover->data.crossover) );
3739  crossover->data.crossover->rng = NULL;
3740 
3741  /* add crossover neighborhood parameters */
3742  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/alns/crossover/nsols", "the number of solutions that crossover should combine",
3743  &crossover->data.crossover->nsols, TRUE, DEFAULT_NSOLS_CROSSOVER, 2, 10, NULL, NULL) );
3744 
3745  /* include the Proximity neighborhood */
3746  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &proximity, "proximity",
3748  NULL, changeSubscipProximity, NULL, NULL, NULL, nhRefsolIncumbent, nhDeactivateBinVars) );
3749 
3750  /* include the Zeroobjective neighborhood */
3751  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &zeroobjective, "zeroobjective",
3753  NULL, changeSubscipZeroobjective, NULL, NULL, NULL, nhRefsolIncumbent, nhDeactivateObjVars) );
3754 
3755  /* include the DINS neighborhood */
3756  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &dins, "dins",
3758  varFixingsDins, changeSubscipDins, NULL, NULL, nhFreeDins, nhRefsolIncumbent, nhDeactivateBinVars) );
3759 
3760  /* allocate data for DINS to include the parameter */
3761  SCIP_CALL( SCIPallocBlockMemory(scip, &dins->data.dins) );
3762 
3763  /* add DINS neighborhood parameters */
3764  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/alns/dins/npoolsols",
3765  "number of pool solutions where binary solution values must agree",
3766  &dins->data.dins->npoolsols, TRUE, DEFAULT_NPOOLSOLS_DINS, 1, 100, NULL, NULL) );
3767 
3768  /* include the trustregion neighborhood */
3769  SCIP_CALL( alnsIncludeNeighborhood(scip, heurdata, &trustregion, "trustregion",
3771  NULL, changeSubscipTrustregion, NULL, NULL, nhFreeTrustregion, nhRefsolIncumbent, nhDeactivateBinVars) );
3772 
3773  /* allocate data for trustregion to include the parameter */
3774  SCIP_CALL( SCIPallocBlockMemory(scip, &trustregion->data.trustregion) );
3775 
3776  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/trustregion/violpenalty",
3777  "the penalty for each change in the binary variables from the candidate solution",
3779 
3780  return SCIP_OKAY;
3781 }
3782 
3783 /** initialization method of primal heuristic (called after problem was transformed) */
3784 static
3785 SCIP_DECL_HEURINIT(heurInitAlns)
3786 { /*lint --e{715}*/
3787  SCIP_HEURDATA* heurdata;
3788  int i;
3789 
3790  assert(scip != NULL);
3791  assert(heur != NULL);
3792 
3793  heurdata = SCIPheurGetData(heur);
3794  assert(heurdata != NULL);
3795 
3796  /* reactivate all neighborhoods if a new problem is read in */
3797  heurdata->nactiveneighborhoods = heurdata->nneighborhoods;
3798 
3799  /* initialize neighborhoods for new problem */
3800  for( i = 0; i < heurdata->nneighborhoods; ++i )
3801  {
3802  NH* neighborhood = heurdata->neighborhoods[i];
3803 
3804  SCIP_CALL( neighborhoodInit(scip, neighborhood) );
3805 
3806  SCIP_CALL( resetFixingRate(scip, &neighborhood->fixingrate) );
3807 
3808  SCIP_CALL( neighborhoodStatsReset(scip, &neighborhood->stats) );
3809  }
3810 
3811  /* open reward file for reading */
3812  if( strncasecmp(heurdata->rewardfilename, DEFAULT_REWARDFILENAME, strlen(DEFAULT_REWARDFILENAME)) != 0 )
3813  {
3814  heurdata->rewardfile = fopen(heurdata->rewardfilename, "w");
3815 
3816  if( heurdata->rewardfile == NULL )
3817  {
3818  SCIPerrorMessage("Error: Could not open reward file <%s>\n", heurdata->rewardfilename);
3819  return SCIP_FILECREATEERROR;
3820  }
3821 
3822  SCIPdebugMsg(scip, "Writing reward information to <%s>\n", heurdata->rewardfilename);
3823  }
3824  else
3825  heurdata->rewardfile = NULL;
3826 
3827  return SCIP_OKAY;
3828 }
3829 
3830 
3831 /** solving process initialization method of primal heuristic (called when branch and bound process is about to begin) */
3832 static
3833 SCIP_DECL_HEURINITSOL(heurInitsolAlns)
3834 { /*lint --e{715}*/
3835  SCIP_HEURDATA* heurdata;
3836  int i;
3837  SCIP_Real* priorities;
3838  unsigned int initseed;
3839 
3840  assert(scip != NULL);
3841  assert(heur != NULL);
3842 
3843  heurdata = SCIPheurGetData(heur);
3844  assert(heurdata != NULL);
3845  heurdata->nactiveneighborhoods = heurdata->nneighborhoods;
3846 
3847  SCIP_CALL( SCIPallocBufferArray(scip, &priorities, heurdata->nactiveneighborhoods) );
3848 
3849  /* init neighborhoods for new problem by resetting their statistics and fixing rate */
3850  for( i = heurdata->nneighborhoods - 1; i >= 0; --i )
3851  {
3852  NH* neighborhood = heurdata->neighborhoods[i];
3853  SCIP_Bool deactivate;
3854 
3855  SCIP_CALL( neighborhood->nhdeactivate(scip, &deactivate) );
3856 
3857  /* disable inactive neighborhoods */
3858  if( deactivate || ! neighborhood->active )
3859  {
3860  if( heurdata->nactiveneighborhoods - 1 > i )
3861  {
3862  assert(heurdata->neighborhoods[heurdata->nactiveneighborhoods - 1]->active);
3863  SCIPswapPointers((void **)&heurdata->neighborhoods[i], (void **)&heurdata->neighborhoods[heurdata->nactiveneighborhoods - 1]);
3864  }
3865  heurdata->nactiveneighborhoods--;
3866  }
3867  }
3868 
3869  /* collect neighborhood priorities */
3870  for( i = 0; i < heurdata->nactiveneighborhoods; ++i )
3871  priorities[i] = heurdata->neighborhoods[i]->priority;
3872 
3873  initseed = (unsigned int)(heurdata->seed + SCIPgetNVars(scip));
3874 
3875  /* active neighborhoods might change between init calls, reset functionality must take this into account */
3876  if( heurdata->bandit != NULL && SCIPbanditGetNActions(heurdata->bandit) != heurdata->nactiveneighborhoods )
3877  {
3878  SCIP_CALL( SCIPfreeBandit(scip, &heurdata->bandit) );
3879 
3880  heurdata->bandit = NULL;
3881  }
3882 
3883  if( heurdata->nactiveneighborhoods > 0 )
3884  { /* create or reset bandit algorithm */
3885  if( heurdata->bandit == NULL )
3886  {
3887  SCIP_CALL( createBandit(scip, heurdata, priorities, initseed) );
3888 
3889  resetMinimumImprovement(heurdata);
3890  resetTargetNodeLimit(heurdata);
3891  }
3892  else if( heurdata->resetweights )
3893  {
3894  SCIP_CALL( SCIPresetBandit(scip, heurdata->bandit, priorities, initseed) );
3895 
3896  resetMinimumImprovement(heurdata);
3897  resetTargetNodeLimit(heurdata);
3898  }
3899  }
3900 
3901  heurdata->usednodes = 0;
3902  heurdata->ninitneighborhoods = heurdata->nactiveneighborhoods;
3903 
3904  heurdata->lastcallsol = NULL;
3905  heurdata->firstcallthissol = 0;
3906 
3907  resetCurrentNeighborhood(heurdata);
3908 
3909  SCIPfreeBufferArray(scip, &priorities);
3910 
3911  return SCIP_OKAY;
3912 }
3913 
3914 
3915 /** deinitialization method of primal heuristic (called before transformed problem is freed) */
3916 static
3917 SCIP_DECL_HEUREXIT(heurExitAlns)
3918 { /*lint --e{715}*/
3919  SCIP_HEURDATA* heurdata;
3920  int i;
3921 
3922  assert(scip != NULL);
3923  assert(heur != NULL);
3924 
3925  heurdata = SCIPheurGetData(heur);
3926  assert(heurdata != NULL);
3927 
3928  /* free neighborhood specific data */
3929  for( i = 0; i < heurdata->nneighborhoods; ++i )
3930  {
3931  NH* neighborhood = heurdata->neighborhoods[i];
3932 
3933  SCIP_CALL( neighborhoodExit(scip, neighborhood) );
3934  }
3935 
3936  if( heurdata->rewardfile != NULL )
3937  {
3938  fclose(heurdata->rewardfile);
3939  heurdata->rewardfile = NULL;
3940  }
3941 
3942  return SCIP_OKAY;
3943 }
3944 
3945 /** destructor of primal heuristic to free user data (called when SCIP is exiting) */
3946 static
3947 SCIP_DECL_HEURFREE(heurFreeAlns)
3948 { /*lint --e{715}*/
3949  SCIP_HEURDATA* heurdata;
3950  int i;
3951 
3952  assert(scip != NULL);
3953  assert(heur != NULL);
3954 
3955  heurdata = SCIPheurGetData(heur);
3956  assert(heurdata != NULL);
3957 
3958  /* bandits are only initialized if a problem has been read */
3959  if( heurdata->bandit != NULL )
3960  {
3961  SCIP_CALL( SCIPfreeBandit(scip, &heurdata->bandit) );
3962  }
3963 
3964  /* free neighborhoods */
3965  for( i = 0; i < heurdata->nneighborhoods; ++i )
3966  {
3967  SCIP_CALL( alnsFreeNeighborhood(scip, &(heurdata->neighborhoods[i])) );
3968  }
3969 
3970  SCIPfreeBlockMemoryArray(scip, &heurdata->neighborhoods, NNEIGHBORHOODS);
3971 
3972  SCIPfreeBlockMemory(scip, &heurdata);
3973 
3974  return SCIP_OKAY;
3975 }
3976 
3977 /** output method of statistics table to output file stream 'file' */
3978 static
3979 SCIP_DECL_TABLEOUTPUT(tableOutputNeighborhood)
3980 { /*lint --e{715}*/
3981  SCIP_HEURDATA* heurdata;
3982 
3983  assert(SCIPfindHeur(scip, HEUR_NAME) != NULL);
3984  heurdata = SCIPheurGetData(SCIPfindHeur(scip, HEUR_NAME));
3985  assert(heurdata != NULL);
3986 
3987  printNeighborhoodStatistics(scip, heurdata, file);
3988 
3989  return SCIP_OKAY;
3990 }
3991 
3992 /*
3993  * primal heuristic specific interface methods
3994  */
3995 
3996 /** creates the alns primal heuristic and includes it in SCIP */
3998  SCIP* scip /**< SCIP data structure */
3999  )
4000 {
4001  SCIP_HEURDATA* heurdata;
4002  SCIP_HEUR* heur;
4003 
4004  /* create alns primal heuristic data */
4005  heurdata = NULL;
4006  heur = NULL;
4007 
4008  SCIP_CALL( SCIPallocBlockMemory(scip, &heurdata) );
4009  BMSclearMemory(heurdata);
4010 
4011  /* TODO make this a user parameter? */
4012  heurdata->lplimfac = LPLIMFAC;
4013 
4014  SCIP_CALL( SCIPallocBlockMemoryArray(scip, &heurdata->neighborhoods, NNEIGHBORHOODS) );
4015 
4016  /* include primal heuristic */
4017  SCIP_CALL( SCIPincludeHeurBasic(scip, &heur,
4019  HEUR_MAXDEPTH, HEUR_TIMING, HEUR_USESSUBSCIP, heurExecAlns, heurdata) );
4020 
4021  assert(heur != NULL);
4022 
4023  /* include all neighborhoods */
4024  SCIP_CALL( includeNeighborhoods(scip, heurdata) );
4025 
4026  /* set non fundamental callbacks via setter functions */
4027  SCIP_CALL( SCIPsetHeurCopy(scip, heur, heurCopyAlns) );
4028  SCIP_CALL( SCIPsetHeurFree(scip, heur, heurFreeAlns) );
4029  SCIP_CALL( SCIPsetHeurInit(scip, heur, heurInitAlns) );
4030  SCIP_CALL( SCIPsetHeurInitsol(scip, heur, heurInitsolAlns) );
4031  SCIP_CALL( SCIPsetHeurExit(scip, heur, heurExitAlns) );
4032 
4033  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/shownbstats",
4034  "show statistics on neighborhoods?",
4035  &heurdata->shownbstats, TRUE, DEFAULT_SHOWNBSTATS, NULL, NULL) );
4036 
4037  /* add alns primal heuristic parameters */
4038  SCIP_CALL( SCIPaddLongintParam(scip, "heuristics/" HEUR_NAME "/maxnodes",
4039  "maximum number of nodes to regard in the subproblem",
4040  &heurdata->maxnodes, TRUE,DEFAULT_MAXNODES, 0LL, SCIP_LONGINT_MAX, NULL, NULL) );
4041 
4042  SCIP_CALL( SCIPaddLongintParam(scip, "heuristics/" HEUR_NAME "/nodesofs",
4043  "offset added to the nodes budget",
4044  &heurdata->nodesoffset, FALSE, DEFAULT_NODESOFFSET, 0LL, SCIP_LONGINT_MAX, NULL, NULL) );
4045 
4046  SCIP_CALL( SCIPaddLongintParam(scip, "heuristics/" HEUR_NAME "/minnodes",
4047  "minimum number of nodes required to start a sub-SCIP",
4048  &heurdata->minnodes, TRUE, DEFAULT_MINNODES, 0LL, SCIP_LONGINT_MAX, NULL, NULL) );
4049 
4050  SCIP_CALL( SCIPaddLongintParam(scip, "heuristics/" HEUR_NAME "/waitingnodes",
4051  "number of nodes since last incumbent solution that the heuristic should wait",
4052  &heurdata->waitingnodes, TRUE, DEFAULT_WAITINGNODES, 0LL, SCIP_LONGINT_MAX, NULL, NULL) );
4053 
4054  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/nodesquot",
4055  "fraction of nodes compared to the main SCIP for budget computation",
4056  &heurdata->nodesquot, FALSE, DEFAULT_NODESQUOT, 0.0, 1.0, NULL, NULL) );
4057  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/nodesquotmin",
4058  "lower bound fraction of nodes compared to the main SCIP for budget computation",
4059  &heurdata->nodesquotmin, FALSE, DEFAULT_NODESQUOTMIN, 0.0, 1.0, NULL, NULL) );
4060 
4061  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/startminimprove",
4062  "initial factor by which ALNS should at least improve the incumbent",
4063  &heurdata->startminimprove, TRUE, DEFAULT_STARTMINIMPROVE, 0.0, 1.0, NULL, NULL) );
4064 
4065  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/minimprovelow",
4066  "lower threshold for the minimal improvement over the incumbent",
4067  &heurdata->minimprovelow, TRUE, DEFAULT_MINIMPROVELOW, 0.0, 1.0, NULL, NULL) );
4068 
4069  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/minimprovehigh",
4070  "upper bound for the minimal improvement over the incumbent",
4071  &heurdata->minimprovehigh, TRUE, DEFAULT_MINIMPROVEHIGH, 0.0, 1.0, NULL, NULL) );
4072 
4073  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/" HEUR_NAME "/nsolslim",
4074  "limit on the number of improving solutions in a sub-SCIP call",
4075  &heurdata->nsolslim, FALSE, DEFAULT_NSOLSLIM, -1, INT_MAX, NULL, NULL) );
4076 
4077  SCIP_CALL( SCIPaddCharParam(scip, "heuristics/" HEUR_NAME "/banditalgo",
4078  "the bandit algorithm: (u)pper confidence bounds, (e)xp.3, epsilon (g)reedy",
4079  &heurdata->banditalgo, TRUE, DEFAULT_BANDITALGO, "ueg", NULL, NULL) );
4080 
4081  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/gamma",
4082  "weight between uniform (gamma ~ 1) and weight driven (gamma ~ 0) probability distribution for exp3",
4083  &heurdata->exp3_gamma, TRUE, DEFAULT_GAMMA, 0.0, 1.0, NULL, NULL) );
4084 
4085  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/beta",
4086  "reward offset between 0 and 1 at every observation for Exp.3",
4087  &heurdata->exp3_beta, TRUE, DEFAULT_BETA, 0.0, 1.0, NULL, NULL) );
4088 
4089  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/alpha",
4090  "parameter to increase the confidence width in UCB",
4091  &heurdata->ucb_alpha, TRUE, DEFAULT_ALPHA, 0.0, 100.0, NULL, NULL) );
4092 
4093  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/usedistances",
4094  "distances from fixed variables be used for variable prioritization",
4095  &heurdata->usedistances, TRUE, DEFAULT_USEDISTANCES, NULL, NULL) );
4096 
4097  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/useredcost",
4098  "should reduced cost scores be used for variable prioritization?",
4099  &heurdata->useredcost, TRUE, DEFAULT_USEREDCOST, NULL, NULL) );
4100 
4101  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/domorefixings",
4102  "should the ALNS heuristic do more fixings by itself based on variable prioritization "
4103  "until the target fixing rate is reached?",
4104  &heurdata->domorefixings, TRUE, DEFAULT_DOMOREFIXINGS, NULL, NULL) );
4105 
4106  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/adjustfixingrate",
4107  "should the heuristic adjust the target fixing rate based on the success?",
4108  &heurdata->adjustfixingrate, TRUE, DEFAULT_ADJUSTFIXINGRATE, NULL, NULL) );
4109 
4110  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/usesubscipheurs",
4111  "should the heuristic activate other sub-SCIP heuristics during its search?",
4112  &heurdata->usesubscipheurs, TRUE, DEFAULT_USESUBSCIPHEURS, NULL, NULL) );
4113 
4114  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/rewardcontrol",
4115  "reward control to increase the weight of the simple solution indicator and decrease the weight of the closed gap reward",
4116  &heurdata->rewardcontrol, TRUE, DEFAULT_REWARDCONTROL, 0.0, 1.0, NULL, NULL) );
4117 
4118  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/targetnodefactor",
4119  "factor by which target node number is eventually increased",
4120  &heurdata->targetnodefactor, TRUE, DEFAULT_TARGETNODEFACTOR, 1.0, 1e+5, NULL, NULL) );
4121 
4122  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/" HEUR_NAME "/seed",
4123  "initial random seed for bandit algorithms and random decisions by neighborhoods",
4124  &heurdata->seed, FALSE, DEFAULT_SEED, 0, INT_MAX, NULL, NULL) );
4125  SCIP_CALL( SCIPaddIntParam(scip, "heuristics/" HEUR_NAME "/maxcallssamesol",
4126  "number of allowed executions of the heuristic on the same incumbent solution (-1: no limit, 0: number of active neighborhoods)",
4127  &heurdata->maxcallssamesol, TRUE, DEFAULT_MAXCALLSSAMESOL, -1, 100, NULL, NULL) );
4128 
4129  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/adjustminimprove",
4130  "should the factor by which the minimum improvement is bound be dynamically updated?",
4131  &heurdata->adjustminimprove, TRUE, DEFAULT_ADJUSTMINIMPROVE, NULL, NULL) );
4132 
4133  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/adjusttargetnodes",
4134  "should the target nodes be dynamically adjusted?",
4135  &heurdata->adjusttargetnodes, TRUE, DEFAULT_ADJUSTTARGETNODES, NULL, NULL) );
4136 
4137  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/eps",
4138  "increase exploration in epsilon-greedy bandit algorithm",
4139  &heurdata->epsgreedy_eps, TRUE, DEFAULT_EPS, 0.0, 1.0, NULL, NULL) );
4140 
4141  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/rewardbaseline",
4142  "the reward baseline to separate successful and failed calls",
4143  &heurdata->rewardbaseline, TRUE, DEFAULT_REWARDBASELINE, 0.0, 0.99, NULL, NULL) );
4144 
4145  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/resetweights",
4146  "should the bandit algorithms be reset when a new problem is read?",
4147  &heurdata->resetweights, TRUE, DEFAULT_RESETWEIGHTS, NULL, NULL) );
4148 
4149  SCIP_CALL( SCIPaddStringParam(scip, "heuristics/" HEUR_NAME "/rewardfilename", "file name to store all rewards and the selection of the bandit",
4150  &heurdata->rewardfilename, TRUE, DEFAULT_REWARDFILENAME, NULL, NULL) );
4151 
4152  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/subsciprandseeds",
4153  "should random seeds of sub-SCIPs be altered to increase diversification?",
4154  &heurdata->subsciprandseeds, TRUE, DEFAULT_SUBSCIPRANDSEEDS, NULL, NULL) );
4155 
4156  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/scalebyeffort",
4157  "should the reward be scaled by the effort?",
4158  &heurdata->scalebyeffort, TRUE, DEFAULT_SCALEBYEFFORT, NULL, NULL) );
4159 
4160  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/copycuts",
4161  "should cutting planes be copied to the sub-SCIP?",
4162  &heurdata->copycuts, TRUE, DEFAULT_COPYCUTS, NULL, NULL) );
4163 
4164  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/fixtol",
4165  "tolerance by which the fixing rate may be missed without generic fixing",
4166  &heurdata->fixtol, TRUE, DEFAULT_FIXTOL, 0.0, 1.0, NULL, NULL) );
4167 
4168  SCIP_CALL( SCIPaddRealParam(scip, "heuristics/" HEUR_NAME "/unfixtol",
4169  "tolerance by which the fixing rate may be exceeded without generic unfixing",
4170  &heurdata->unfixtol, TRUE, DEFAULT_UNFIXTOL, 0.0, 1.0, NULL, NULL) );
4171 
4172  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/uselocalredcost",
4173  "should local reduced costs be used for generic (un)fixing?",
4174  &heurdata->uselocalredcost, TRUE, DEFAULT_USELOCALREDCOST, NULL, NULL) );
4175 
4176  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/usepscost",
4177  "should pseudo cost scores be used for variable priorization?",
4178  &heurdata->usepscost, TRUE, DEFAULT_USEPSCOST, NULL, NULL) );
4179  SCIP_CALL( SCIPaddBoolParam(scip, "heuristics/" HEUR_NAME "/initduringroot",
4180  "should the heuristic be executed multiple times during the root node?",
4181  &heurdata->initduringroot, TRUE, DEFAULT_INITDURINGROOT, NULL, NULL) );
4182 
4183  assert(SCIPfindTable(scip, TABLE_NAME_NEIGHBORHOOD) == NULL);
4185  NULL, NULL, NULL, NULL, NULL, NULL, tableOutputNeighborhood,
4187 
4188  return SCIP_OKAY;
4189 }
enum SCIP_Result SCIP_RESULT
Definition: type_result.h:61
#define DEFAULT_BANDITALGO
Definition: heur_alns.c:121
SCIP_Real targetfixingrate
Definition: heur_alns.c:365
void SCIPfreeRandom(SCIP *scip, SCIP_RANDNUMGEN **randnumgen)
#define SCIPfreeBlockMemoryArray(scip, ptr, num)
Definition: scip_mem.h:110
int SCIPgetNIntVars(SCIP *scip)
Definition: scip_prob.c:2082
#define DEFAULT_GAMMA
Definition: heur_alns.c:138
SCIP_Bool SCIPisFeasZero(SCIP *scip, SCIP_Real val)
#define SCIP_HEURTIMING_DURINGLPLOOP
Definition: type_timing.h:80
NH_STATS stats
Definition: heur_alns.c:375
public methods for the epsilon greedy bandit selector
#define DEFAULT_USESUBSCIPHEURS
Definition: heur_alns.c:151
#define DEFAULT_RESETWEIGHTS
Definition: heur_alns.c:124
static SCIP_DECL_TABLEOUTPUT(tableOutputNeighborhood)
Definition: heur_alns.c:3980
static SCIP_RETCODE addLocalBranchingConstraint(SCIP *sourcescip, SCIP *targetscip, SCIP_VAR **subvars, int distance, SCIP_Bool *success, int *naddedconss)
Definition: heur_alns.c:3225
SCIP_RETCODE SCIPchgVarLbGlobal(SCIP *scip, SCIP_VAR *var, SCIP_Real newbound)
Definition: scip_var.c:4943
int nfixings
Definition: heur_alns.c:356
SCIP_Real SCIPgetSolvingTime(SCIP *scip)
Definition: scip_timing.c:378
#define SCIP_EVENTTYPE_LPSOLVED
Definition: type_event.h:101
SCIP_RETCODE SCIPsetSeparating(SCIP *scip, SCIP_PARAMSETTING paramsetting, SCIP_Bool quiet)
Definition: scip_param.c:958
SCIP_RETCODE SCIPcreateBanditEpsgreedy(SCIP *scip, SCIP_BANDIT **epsgreedy, SCIP_Real *priorities, SCIP_Real eps, SCIP_Bool preferrecent, SCIP_Real decayfactor, int avglim, int nactions, unsigned int initseed)
SCIP_Real oldupperbound
Definition: heur_alns.c:350
int nruns
Definition: heur_alns.c:352
#define DEFAULT_STARTMINIMPROVE
Definition: heur_alns.c:113
#define SCIPallocBlockMemoryArray(scip, ptr, num)
Definition: scip_mem.h:93
static SCIP_RETCODE createBandit(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_Real *priorities, unsigned int initseed)
Definition: heur_alns.c:1598
SCIP_RETCODE SCIPincludeTable(SCIP *scip, const char *name, const char *desc, SCIP_Bool active, SCIP_DECL_TABLECOPY((*tablecopy)), SCIP_DECL_TABLEFREE((*tablefree)), SCIP_DECL_TABLEINIT((*tableinit)), SCIP_DECL_TABLEEXIT((*tableexit)), SCIP_DECL_TABLEINITSOL((*tableinitsol)), SCIP_DECL_TABLEEXITSOL((*tableexitsol)), SCIP_DECL_TABLEOUTPUT((*tableoutput)), SCIP_TABLEDATA *tabledata, int position, SCIP_STAGE earlieststage)
Definition: scip_table.c:56
SCIP_Bool SCIPisFeasEQ(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
public methods for SCIP parameter handling
HistIndex
Definition: heur_alns.c:330
SCIP_RETCODE SCIPcreateBanditExp3(SCIP *scip, SCIP_BANDIT **exp3, SCIP_Real *priorities, SCIP_Real gammaparam, SCIP_Real beta, int nactions, unsigned int initseed)
Definition: bandit_exp3.c:311
#define HEUR_DISPCHAR
Definition: heur_alns.c:79
#define DEFAULT_REWARDBASELINE
Definition: heur_alns.c:126
SCIP_Real * pscostscores
Definition: heur_alns.c:509
SCIP_RETCODE SCIPcreateBanditUcb(SCIP *scip, SCIP_BANDIT **ucb, SCIP_Real *priorities, SCIP_Real alpha, int nactions, unsigned int initseed)
Definition: bandit_ucb.c:339
SCIP_RETCODE SCIPcreateConsBasicLinear(SCIP *scip, SCIP_CONS **cons, const char *name, int nvars, SCIP_VAR **vars, SCIP_Real *vals, SCIP_Real lhs, SCIP_Real rhs)
public methods for node selector plugins
SCIP_RETCODE SCIPincludeHeurAlns(SCIP *scip)
Definition: heur_alns.c:3998
SCIP_SOL * selsol
Definition: heur_alns.c:405
#define DEFAULT_MAXFIXINGRATE_LOCALBRANCHING
Definition: heur_alns.c:177
static void updateMinimumImprovement(SCIP_HEURDATA *heurdata, SCIP_STATUS subscipstatus, NH_STATS *runstats)
Definition: heur_alns.c:725
public methods for memory management
enum HistIndex HISTINDEX
Definition: heur_alns.c:340
static SCIP_RETCODE alnsUnfixVariables(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_VAR **varbuf, SCIP_Real *valbuf, int *nfixings, int ntargetfixings, SCIP_Bool *success)
Definition: heur_alns.c:1654
static void decreaseMinimumImprovement(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:712
SCIP_CONSHDLR * SCIPfindConshdlr(SCIP *scip, const char *name)
Definition: scip_cons.c:886
SCIP_Real SCIPgetCutoffbound(SCIP *scip)
DATA_DINS * dins
Definition: heur_alns.c:389
static void resetMinimumImprovement(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:690
SCIP_Real SCIPvarGetLbGlobal(SCIP_VAR *var)
Definition: var.c:17901
#define DECL_NHINIT(x)
Definition: heur_alns.c:291
#define HEUR_FREQ
Definition: heur_alns.c:81
SCIP_RETCODE SCIPgetRealParam(SCIP *scip, const char *name, SCIP_Real *value)
Definition: scip_param.c:307
#define SCIP_MAXSTRLEN
Definition: def.h:302
SCIP_Real SCIPgetVarPseudocostVal(SCIP *scip, SCIP_VAR *var, SCIP_Real solvaldelta)
Definition: scip_var.c:8814
#define DEFAULT_NSOLSLIM
Definition: heur_alns.c:97
static SCIP_DECL_HEURINIT(heurInitAlns)
Definition: heur_alns.c:3786
SCIP_Longint SCIPheurGetNBestSolsFound(SCIP_HEUR *heur)
Definition: heur.c:1596
#define DEFAULT_ACTIVE_CROSSOVER
Definition: heur_alns.c:188
#define DEFAULT_PRIORITY_PROXIMITY
Definition: heur_alns.c:184
#define HEUR_TIMING
Definition: heur_alns.c:84
SCIP_RETCODE SCIPbanditSelect(SCIP_BANDIT *bandit, int *action)
Definition: bandit.c:153
SCIP_RETCODE SCIPsetHeurExit(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEUREXIT((*heurexit)))
Definition: scip_heur.c:210
SCIP_Real SCIPgetVarRedcost(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:1864
static SCIP_RETCODE alnsIncludeNeighborhood(SCIP *scip, SCIP_HEURDATA *heurdata, NH **neighborhood, const char *name, SCIP_Real minfixingrate, SCIP_Real maxfixingrate, SCIP_Bool active, SCIP_Real priority, DECL_VARFIXINGS((*varfixings)), DECL_CHANGESUBSCIP((*changesubscip)), DECL_NHINIT((*nhinit)), DECL_NHEXIT((*nhexit)), DECL_NHFREE((*nhfree)), DECL_NHREFSOL((*nhrefsol)), DECL_NHDEACTIVATE((*nhdeactivate)))
Definition: heur_alns.c:799
#define DEFAULT_PRIORITY_DINS
Definition: heur_alns.c:199
int SCIPgetNOrigVars(SCIP *scip)
Definition: scip_prob.c:2432
#define DEFAULT_USEREDCOST
Definition: heur_alns.c:142
public solving methods
static void updateNeighborhoodStats(NH_STATS *runstats, NH *neighborhood, SCIP_STATUS subscipstatus)
Definition: heur_alns.c:1167
SCIP * scip
Definition: heur_alns.c:505
SCIP_RETCODE SCIPincludeEventhdlrBasic(SCIP *scip, SCIP_EVENTHDLR **eventhdlrptr, const char *name, const char *desc, SCIP_DECL_EVENTEXEC((*eventexec)), SCIP_EVENTHDLRDATA *eventhdlrdata)
Definition: scip_event.c:104
public methods for timing
static SCIP_RETCODE updateBanditAlgorithm(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_Real reward, int neighborhoodidx)
Definition: heur_alns.c:2141
SCIP_Bool SCIPvarIsBinary(SCIP_VAR *var)
Definition: var.c:17422
#define DECL_VARFIXINGS(x)
Definition: heur_alns.c:262
SCIP_TABLE * SCIPfindTable(SCIP *scip, const char *name)
Definition: scip_table.c:94
SCIP_RETCODE SCIPstopClock(SCIP *scip, SCIP_CLOCK *clck)
Definition: scip_timing.c:178
SCIP_Bool SCIPisFeasNegative(SCIP *scip, SCIP_Real val)
#define DEFAULT_ACTIVE_RINS
Definition: heur_alns.c:168
#define NNEIGHBORHOODS
Definition: heur_alns.c:87
DATA_CROSSOVER * crossover
Definition: heur_alns.c:388
SCIP_Real SCIPvarGetRootSol(SCIP_VAR *var)
Definition: var.c:13340
static SCIP_Real getVariablePscostScore(SCIP *scip, SCIP_VAR *var, SCIP_Real refsolval, SCIP_Bool uselocallpsol)
Definition: heur_alns.c:1329
#define TABLE_DESC_NEIGHBORHOOD
Definition: heur_alns.c:218
#define DECL_NHREFSOL(x)
Definition: heur_alns.c:316
void SCIPswapPointers(void **pointer1, void **pointer2)
Definition: misc.c:10307
SCIP_RETCODE SCIPgetVarsData(SCIP *scip, SCIP_VAR ***vars, int *nvars, int *nbinvars, int *nintvars, int *nimplvars, int *ncontvars)
Definition: scip_prob.c:1866
#define LPLIMFAC
Definition: heur_alns.c:103
SCIP_SOL ** SCIPgetSols(SCIP *scip)
Definition: scip_sol.c:2263
#define FALSE
Definition: def.h:96
SCIP_RETCODE SCIPhashmapCreate(SCIP_HASHMAP **hashmap, BMS_BLKMEM *blkmem, int mapsize)
Definition: misc.c:3024
#define DEFAULT_ACTIVE_LOCALBRANCHING
Definition: heur_alns.c:178
const char * SCIPeventhdlrGetName(SCIP_EVENTHDLR *eventhdlr)
Definition: event.c:324
SCIP_RETCODE SCIPaddLongintParam(SCIP *scip, const char *name, const char *desc, SCIP_Longint *valueptr, SCIP_Bool isadvanced, SCIP_Longint defaultvalue, SCIP_Longint minvalue, SCIP_Longint maxvalue, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:111
#define DEFAULT_NODESOFFSET
Definition: heur_alns.c:96
SCIP_Real SCIPinfinity(SCIP *scip)
int SCIPsnprintf(char *t, int len, const char *s,...)
Definition: misc.c:10788
#define TRUE
Definition: def.h:95
SCIP_Longint stallnodes
Definition: heur_alns.c:497
#define DEFAULT_PRIORITY_RINS
Definition: heur_alns.c:169
enum SCIP_Retcode SCIP_RETCODE
Definition: type_retcode.h:63
methods commonly used by primal heuristics
#define DEFAULT_ACTIVE_PROXIMITY
Definition: heur_alns.c:183
#define DEFAULT_MINFIXINGRATE_PROXIMITY
Definition: heur_alns.c:181
SCIP_RETCODE SCIPsetPresolving(SCIP *scip, SCIP_PARAMSETTING paramsetting, SCIP_Bool quiet)
Definition: scip_param.c:932
#define DEFAULT_WAITINGNODES
Definition: heur_alns.c:100
SCIP_BRANCHRULE * SCIPfindBranchrule(SCIP *scip, const char *name)
Definition: scip_branch.c:297
SCIP_RETCODE SCIPtranslateSubSol(SCIP *scip, SCIP *subscip, SCIP_SOL *subsol, SCIP_HEUR *heur, SCIP_VAR **subvars, SCIP_SOL **newsol)
Definition: scip_copy.c:1408
NH_FIXINGRATE fixingrate
Definition: heur_alns.c:374
int SCIPvarGetProbindex(SCIP_VAR *var)
Definition: var.c:17591
#define DEFAULT_NODESQUOT
Definition: heur_alns.c:94
#define DEFAULT_VIOLPENALTY_TRUSTREGION
Definition: heur_alns.c:209
SCIP_Bool SCIPisDualfeasZero(SCIP *scip, SCIP_Real val)
#define DEFAULT_MINIMPROVELOW
Definition: heur_alns.c:110
#define DEFAULT_SHOWNBSTATS
Definition: heur_alns.c:89
#define DECL_NHFREE(x)
Definition: heur_alns.c:303
struct SCIP_HeurData SCIP_HEURDATA
Definition: type_heur.h:76
public methods for problem variables
static GRAPHNODE ** active
void SCIPselectInd(int *indarray, SCIP_DECL_SORTINDCOMP((*indcomp)), void *dataptr, int k, int len)
static SCIP_RETCODE neighborhoodFixVariables(SCIP *scip, SCIP_HEURDATA *heurdata, NH *neighborhood, SCIP_VAR **varbuf, SCIP_Real *valbuf, int *nfixings, SCIP_RESULT *result)
Definition: heur_alns.c:1801
static SCIP_DECL_HEUREXIT(heurExitAlns)
Definition: heur_alns.c:3918
#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 SCIPaddStringParam(SCIP *scip, const char *name, const char *desc, char **valueptr, SCIP_Bool isadvanced, const char *defaultvalue, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:194
#define DEFAULT_REWARDFILENAME
Definition: heur_alns.c:153
#define DEFAULT_MAXFIXINGRATE_DINS
Definition: heur_alns.c:197
#define SCIPdebugMessage
Definition: pub_message.h:96
SCIP_Real timelimit
Definition: heur_alns.c:496
int SCIPrandomGetInt(SCIP_RANDNUMGEN *randnumgen, int minrandval, int maxrandval)
Definition: misc.c:10019
#define HEUR_NAME
Definition: heur_alns.c:77
#define DEFAULT_MINFIXINGRATE_RENS
Definition: heur_alns.c:161
static SCIP_RETCODE alnsFixMoreVariables(SCIP *scip, SCIP_HEURDATA *heurdata, SCIP_SOL *refsol, SCIP_VAR **varbuf, SCIP_Real *valbuf, int *nfixings, int ntargetfixings, SCIP_Bool *success)
Definition: heur_alns.c:1428
#define SCIPduplicateBufferArray(scip, ptr, source, num)
Definition: scip_mem.h:132
SCIP_RETCODE SCIPfreeClock(SCIP *scip, SCIP_CLOCK **clck)
Definition: scip_timing.c:127
void SCIPsortDownRealInt(SCIP_Real *realarray, int *intarray, int len)
void * SCIPhashmapGetImage(SCIP_HASHMAP *hashmap, void *origin)
Definition: misc.c:3211
SCIP_Bool SCIPisEQ(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
DATA_TRUSTREGION * trustregion
Definition: heur_alns.c:390
#define SCIP_LONGINT_MAX
Definition: def.h:172
#define SCIPfreeBufferArray(scip, ptr)
Definition: scip_mem.h:136
#define DEFAULT_UNFIXTOL
Definition: heur_alns.c:128
SCIP_RETCODE SCIPcreate(SCIP **scip)
Definition: scip_general.c:292
#define SCIPallocBlockMemory(scip, ptr)
Definition: scip_mem.h:89
SCIP_Bool SCIPisTransformed(SCIP *scip)
Definition: scip_general.c:575
public methods for SCIP variables
SCIP_RETCODE SCIPsetRealParam(SCIP *scip, const char *name, SCIP_Real value)
Definition: scip_param.c:603
Definition: heur_alns.c:371
#define SCIP_EVENTTYPE_ALNS
Definition: heur_alns.c:214
SCIP_Real violpenalty
Definition: heur_alns.c:416
void SCIPwarningMessage(SCIP *scip, const char *formatstr,...)
Definition: scip_message.c:120
SCIP_RETCODE SCIPchgVarUbGlobal(SCIP *scip, SCIP_VAR *var, SCIP_Real newbound)
Definition: scip_var.c:5032
#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 SCIPprintStatistics(SCIP *scip, FILE *file)
SCIP_RETCODE SCIPaddCoefLinear(SCIP *scip, SCIP_CONS *cons, SCIP_VAR *var, SCIP_Real val)
void SCIPinfoMessage(SCIP *scip, FILE *file, const char *formatstr,...)
Definition: scip_message.c:208
#define DEFAULT_PRIORITY_ZEROOBJECTIVE
Definition: heur_alns.c:194
#define LRATEMIN
Definition: heur_alns.c:102
#define DEFAULT_ADJUSTFIXINGRATE
Definition: heur_alns.c:148
SCIP_Real SCIPfeasCeil(SCIP *scip, SCIP_Real val)
static SCIP_RETCODE alnsFreeNeighborhood(SCIP *scip, NH **neighborhood)
Definition: heur_alns.c:862
public methods for numerical tolerances
SCIP_Real SCIPfeasFloor(SCIP *scip, SCIP_Real val)
static void updateTargetNodeLimit(SCIP_HEURDATA *heurdata, NH_STATS *runstats, SCIP_STATUS subscipstatus)
Definition: heur_alns.c:653
#define HEUR_USESSUBSCIP
Definition: heur_alns.c:85
#define SCIP_REAL_FORMAT
Definition: def.h:189
public methods for querying solving statistics
union Nh::@5 data
#define DEFAULT_ACTIVE_RENS
Definition: heur_alns.c:163
public methods for the branch-and-bound tree
static void increaseTargetNodeLimit(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:631
#define DEFAULT_NODESQUOTMIN
Definition: heur_alns.c:95
#define DEFAULT_USEDISTANCES
Definition: heur_alns.c:144
#define DEFAULT_MINIMPROVEHIGH
Definition: heur_alns.c:111
static void updateRunStats(NH_STATS *stats, SCIP *subscip)
Definition: heur_alns.c:1055
SCIP_Real increment
Definition: heur_alns.c:366
SCIP_Real SCIPvarGetUbGlobal(SCIP_VAR *var)
Definition: var.c:17911
#define DEFAULT_PRIORITY_LOCALBRANCHING
Definition: heur_alns.c:179
#define TABLE_POSITION_NEIGHBORHOOD
Definition: heur_alns.c:219
#define DEFAULT_ADJUSTTARGETNODES
Definition: heur_alns.c:115
#define DEFAULT_MINFIXINGRATE_RINS
Definition: heur_alns.c:166
static void resetCurrentNeighborhood(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:538
public methods for managing constraints
#define SCIP_EVENTTYPE_SOLFOUND
Definition: type_event.h:144
SCIP_Real memorylimit
Definition: heur_alns.c:495
static SCIP_RETCODE neighborhoodStatsReset(SCIP *scip, NH_STATS *stats)
Definition: heur_alns.c:774
static void increaseFixingRate(NH_FIXINGRATE *fx)
Definition: heur_alns.c:563
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
SCIP_RETCODE SCIPcreateClock(SCIP *scip, SCIP_CLOCK **clck)
Definition: scip_timing.c:76
SCIP_HEUR * SCIPfindHeur(SCIP *scip, const char *name)
Definition: scip_heur.c:258
#define BMSfreeMemoryArray(ptr)
Definition: memory.h:149
#define SCIPerrorMessage
Definition: pub_message.h:64
DATA_MUTATION * mutation
Definition: heur_alns.c:387
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
static SCIP_DECL_HEURCOPY(heurCopyAlns)
Definition: heur_alns.c:1636
#define DECL_NHDEACTIVATE(x)
Definition: heur_alns.c:324
SCIP_RETCODE SCIPaddTrustregionNeighborhoodConstraint(SCIP *sourcescip, SCIP *targetscip, SCIP_VAR **subvars, SCIP_Real violpenalty)
Definition: heuristics.c:999
SCIP_RETCODE SCIPsetHeurFree(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURFREE((*heurfree)))
Definition: scip_heur.c:178
#define DEFAULT_MINFIXINGRATE_DINS
Definition: heur_alns.c:196
SCIP_Real * redcostscores
Definition: heur_alns.c:508
public methods for event handler plugins and event handlers
SCIP_RETCODE SCIPgetSolVals(SCIP *scip, SCIP_SOL *sol, int nvars, SCIP_VAR **vars, SCIP_Real *vals)
Definition: scip_sol.c:1398
SCIP_RETCODE SCIPfreeBandit(SCIP *scip, SCIP_BANDIT **bandit)
Definition: scip_bandit.c:107
char * name
Definition: heur_alns.c:373
static SCIP_DECL_HEURFREE(heurFreeAlns)
Definition: heur_alns.c:3948
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 SCIPpresolve(SCIP *scip)
Definition: scip_solve.c:2461
BMS_BLKMEM * SCIPblkmem(SCIP *scip)
Definition: scip_mem.c:57
#define DEFAULT_MINNODES
Definition: heur_alns.c:98
SCIP_CLOCK * submipclock
Definition: heur_alns.c:348
void SCIPselectDownInd(int *indarray, SCIP_DECL_SORTINDCOMP((*indcomp)), void *dataptr, int k, int len)
#define FIXINGRATE_DECAY
Definition: heur_alns.c:149
#define DEFAULT_ACTIVE_ZEROOBJECTIVE
Definition: heur_alns.c:193
struct SCIP_EventData SCIP_EVENTDATA
Definition: type_event.h:173
static SCIP_DECL_SORTINDCOMP(sortIndCompAlns)
Definition: heur_alns.c:1207
const char * SCIPvarGetName(SCIP_VAR *var)
Definition: var.c:17242
void SCIPhashmapFree(SCIP_HASHMAP **hashmap)
Definition: misc.c:3058
#define DEFAULT_ALPHA
Definition: heur_alns.c:137
SCIP_RETCODE SCIPgetBoolParam(SCIP *scip, const char *name, SCIP_Bool *value)
Definition: scip_param.c:250
static SCIP_DECL_HEURINITSOL(heurInitsolAlns)
Definition: heur_alns.c:3834
#define NULL
Definition: lpi_spx1.cpp:164
SCIP_Real SCIPgetSolTransObj(SCIP *scip, SCIP_SOL *sol)
Definition: scip_sol.c:1491
#define DEFAULT_MAXFIXINGRATE_RENS
Definition: heur_alns.c:162
SCIP_RANDNUMGEN * rng
Definition: heur_alns.c:404
#define EVENTHDLR_DESC
Definition: heur_alns.c:213
#define REALABS(x)
Definition: def.h:210
int SCIPheurGetFreq(SCIP_HEUR *heur)
Definition: heur.c:1535
SCIP_Real SCIPvarGetLPSol(SCIP_VAR *var)
Definition: var.c:18275
#define HEUR_FREQOFS
Definition: heur_alns.c:82
public methods for problem copies
public methods for primal CIP solutions
Adaptive large neighborhood search heuristic that orchestrates popular LNS heuristics.
static SCIP_RETCODE neighborhoodInit(SCIP *scip, NH *neighborhood)
Definition: heur_alns.c:893
#define SCIP_CALL(x)
Definition: def.h:394
SCIP_Real SCIPgetLowerbound(SCIP *scip)
#define DEFAULT_SUBSCIPRANDSEEDS
Definition: heur_alns.c:125
static void updateFixingRate(NH *neighborhood, SCIP_STATUS subscipstatus, NH_STATS *runstats)
Definition: heur_alns.c:586
SCIP_Longint nodelimit
Definition: heur_alns.c:494
unsigned int useredcost
Definition: heur_alns.c:510
SCIP_Longint nbestsolsfound
Definition: heur_alns.c:355
#define DEFAULT_PRIORITY_MUTATION
Definition: heur_alns.c:174
static SCIP_Real getVariableRedcostScore(SCIP *scip, SCIP_VAR *var, SCIP_Real refsolval, SCIP_Bool uselocalredcost)
Definition: heur_alns.c:1276
SCIP_Longint SCIPheurGetNCalls(SCIP_HEUR *heur)
Definition: heur.c:1576
#define DEFAULT_MAXNODES
Definition: heur_alns.c:99
static void updateFixingRateIncrement(NH_FIXINGRATE *fx)
Definition: heur_alns.c:549
SCIP_CLOCK * setupclock
Definition: heur_alns.c:347
SCIP_Bool SCIPisDualfeasNegative(SCIP *scip, SCIP_Real val)
#define BMSduplicateMemoryArray(ptr, source, num)
Definition: memory.h:145
SCIP_Bool SCIPhasCurrentNodeLP(SCIP *scip)
Definition: scip_lp.c:83
public methods for primal heuristic plugins and divesets
public methods for constraint handler plugins and constraints
#define DEFAULT_MINFIXINGRATE_TRUSTREGION
Definition: heur_alns.c:201
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)
SCIP_Real newupperbound
Definition: heur_alns.c:351
#define SCIPallocBufferArray(scip, ptr, num)
Definition: scip_mem.h:124
#define DEFAULT_USEPSCOST
Definition: heur_alns.c:143
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_BESTSOLWEIGHT
Definition: heur_alns.c:120
#define DECL_NHEXIT(x)
Definition: heur_alns.c:297
SCIP_RETCODE SCIPcheckSol(SCIP *scip, SCIP_SOL *sol, SCIP_Bool printreason, SCIP_Bool completely, SCIP_Bool checkbounds, SCIP_Bool checkintegrality, SCIP_Bool checklprows, SCIP_Bool *feasible)
Definition: scip_sol.c:3391
SCIP_Real SCIPgetProbabilityExp3(SCIP_BANDIT *exp3, int action)
Definition: bandit_exp3.c:363
#define DEFAULT_MAXCALLSSAMESOL
Definition: heur_alns.c:105
#define SCIP_Bool
Definition: def.h:93
SCIP_RETCODE SCIPcatchEvent(SCIP *scip, SCIP_EVENTTYPE eventtype, SCIP_EVENTHDLR *eventhdlr, SCIP_EVENTDATA *eventdata, int *filterpos)
Definition: scip_event.c:286
SCIP_LPSOLSTAT SCIPgetLPSolstat(SCIP *scip)
Definition: scip_lp.c:168
SCIP_EVENTTYPE SCIPeventGetType(SCIP_EVENT *event)
Definition: event.c:1030
SCIP_Longint usednodes
Definition: heur_alns.c:349
static SCIP_RETCODE fixMatchingSolutionValues(SCIP *scip, SCIP_SOL **sols, int nsols, SCIP_VAR **vars, int nvars, SCIP_VAR **varbuf, SCIP_Real *valbuf, int *nfixings)
Definition: heur_alns.c:2852
SCIP_Longint SCIPsolGetNodenum(SCIP_SOL *sol)
Definition: sol.c:2618
SCIP_RETCODE SCIPvariablegraphBreadthFirst(SCIP *scip, SCIP_VGRAPH *vargraph, SCIP_VAR **startvars, int nstartvars, int *distances, int maxdistance, int maxvars, int maxbinintvars)
Definition: heur.c:1687
enum SCIP_Status SCIP_STATUS
Definition: type_stat.h:67
static const char * paramname[]
Definition: lpi_msk.c:5096
#define DEFAULT_MAXFIXINGRATE_TRUSTREGION
Definition: heur_alns.c:202
#define DEFAULT_MAXFIXINGRATE_ZEROOBJECTIVE
Definition: heur_alns.c:192
SCIP_RETCODE SCIPsetObjlimit(SCIP *scip, SCIP_Real objlimit)
Definition: scip_prob.c:1422
SCIP_Real SCIPgetClockTime(SCIP *scip, SCIP_CLOCK *clck)
Definition: scip_timing.c:319
int SCIPgetDepth(SCIP *scip)
Definition: scip_tree.c:670
public methods for statistics table plugins
#define DEFAULT_ACTIVE_TRUSTREGION
Definition: heur_alns.c:203
static SCIP_RETCODE transferSolution(SCIP *subscip, SCIP_EVENTDATA *eventdata)
Definition: heur_alns.c:930
#define MAX(x, y)
Definition: tclique_def.h:92
SCIP_RETCODE SCIPtrySolFree(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:3193
static void initRunStats(SCIP *scip, NH_STATS *stats)
Definition: heur_alns.c:1040
SCIP_Bool active
Definition: heur_alns.c:383
#define DECL_CHANGESUBSCIP(x)
Definition: heur_alns.c:279
SCIP_RETCODE SCIPsetIntParam(SCIP *scip, const char *name, int value)
Definition: scip_param.c:487
#define DEFAULT_MAXFIXINGRATE_PROXIMITY
Definition: heur_alns.c:182
static void tryAdd2variableBuffer(SCIP *scip, SCIP_VAR *var, SCIP_Real val, SCIP_VAR **varbuf, SCIP_Real *valbuf, int *nfixings, SCIP_Bool integer)
Definition: heur_alns.c:1358
#define DEFAULT_INITDURINGROOT
Definition: heur_alns.c:104
SCIP_RETCODE SCIPfreeSol(SCIP *scip, SCIP_SOL **sol)
Definition: scip_sol.c:985
SCIP_Real SCIPvarGetObj(SCIP_VAR *var)
Definition: var.c:17749
public methods for bandit algorithms
SCIP_Real SCIPvarGetBestRootRedcost(SCIP_VAR *var)
Definition: var.c:13772
#define DEFAULT_MINFIXINGRATE_CROSSOVER
Definition: heur_alns.c:186
int SCIPgetNSols(SCIP *scip)
Definition: scip_sol.c:2214
#define DEFAULT_ADJUSTMINIMPROVE
Definition: heur_alns.c:114
#define BMScopyMemoryArray(ptr, source, num)
Definition: memory.h:136
static void decreaseFixingRate(NH_FIXINGRATE *fx)
Definition: heur_alns.c:576
static SCIP_RETCODE neighborhoodExit(SCIP *scip, NH *neighborhood)
Definition: heur_alns.c:912
#define DEFAULT_SCALEBYEFFORT
Definition: heur_alns.c:123
static SCIP_DECL_HEUREXEC(heurExecAlns)
Definition: heur_alns.c:2315
#define DEFAULT_EPS
Definition: heur_alns.c:136
#define DEFAULT_MINFIXINGRATE_ZEROOBJECTIVE
Definition: heur_alns.c:191
#define HEUR_DESC
Definition: heur_alns.c:78
Constraint handler for linear constraints in their most general form, .
#define DEFAULT_MINFIXINGRATE_LOCALBRANCHING
Definition: heur_alns.c:176
int SCIPgetNObjVars(SCIP *scip)
Definition: scip_prob.c:2220
static SCIP_RETCODE neighborhoodGetRefsol(SCIP *scip, NH *neighborhood, SCIP_SOL **solptr)
Definition: heur_alns.c:1394
SCIP_Bool SCIPisInfinity(SCIP *scip, SCIP_Real val)
int npoolsols
Definition: heur_alns.c:411
#define BMSclearMemory(ptr)
Definition: memory.h:131
SCIP_Longint nsolsfound
Definition: heur_alns.c:354
static SCIP_RETCODE determineLimits(SCIP *scip, SCIP_HEUR *heur, SOLVELIMITS *solvelimits, SCIP_Bool *runagain)
Definition: heur_alns.c:1959
int SCIPgetNBinVars(SCIP *scip)
Definition: scip_prob.c:2037
#define DEFAULT_DOMOREFIXINGS
Definition: heur_alns.c:145
int SCIPconshdlrGetNActiveConss(SCIP_CONSHDLR *conshdlr)
Definition: cons.c:4631
unsigned int usedistances
Definition: heur_alns.c:511
SCIP_Real SCIPrandomGetReal(SCIP_RANDNUMGEN *randnumgen, SCIP_Real minrandval, SCIP_Real maxrandval)
Definition: misc.c:10041
static SCIP_BANDIT * getBandit(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:2027
public methods for the LP relaxation, rows and columns
int nrunsbestsol
Definition: heur_alns.c:353
SCIP_RETCODE SCIPsetCharParam(SCIP *scip, const char *name, char value)
Definition: scip_param.c:661
public methods for bandit algorithms
int SCIPgetNVars(SCIP *scip)
Definition: scip_prob.c:1992
SCIP_RETCODE SCIPbanditUpdate(SCIP_BANDIT *bandit, int action, SCIP_Real score)
Definition: bandit.c:174
#define SCIP_REAL_MAX
Definition: def.h:187
#define EVENTHDLR_NAME
Definition: heur_alns.c:212
#define HEUR_MAXDEPTH
Definition: heur_alns.c:83
static SCIP_RETCODE resetFixingRate(SCIP *scip, NH_FIXINGRATE *fixingrate)
Definition: heur_alns.c:521
static SCIP_DECL_EVENTEXEC(eventExecAlns)
Definition: heur_alns.c:1006
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)
public methods for branching rule plugins and branching
SCIP_Bool SCIPisDualfeasPositive(SCIP *scip, SCIP_Real val)
SCIP_VAR ** b
Definition: circlepacking.c:65
SCIP_Bool SCIPisObjIntegral(SCIP *scip)
Definition: scip_prob.c:1562
#define DEFAULT_PRIORITY_RENS
Definition: heur_alns.c:164
#define DEFAULT_MINFIXINGRATE_MUTATION
Definition: heur_alns.c:171
public methods for managing events
#define SCIP_EVENTTYPE_BESTSOLFOUND
Definition: type_event.h:105
general public methods
#define DEFAULT_MAXFIXINGRATE_MUTATION
Definition: heur_alns.c:172
SCIP_SOL * SCIPgetBestSol(SCIP *scip)
Definition: scip_sol.c:2313
SCIP_RETCODE SCIPaddCharParam(SCIP *scip, const char *name, const char *desc, char *valueptr, SCIP_Bool isadvanced, char defaultvalue, const char *allowedvalues, SCIP_DECL_PARAMCHGD((*paramchgd)), SCIP_PARAMDATA *paramdata)
Definition: scip_param.c:167
public methods for solutions
static void printNeighborhoodStatistics(SCIP *scip, SCIP_HEURDATA *heurdata, FILE *file)
Definition: heur_alns.c:1095
static SCIP_RETCODE selectNeighborhood(SCIP *scip, SCIP_HEURDATA *heurdata, int *neighborhoodidx)
Definition: heur_alns.c:2037
SCIP_Longint SCIPgetMemUsed(SCIP *scip)
Definition: scip_mem.c:100
public methods for random numbers
SCIP_Real maxfixingrate
Definition: heur_alns.c:367
#define DEFAULT_NPOOLSOLS_DINS
Definition: heur_alns.c:208
SCIP_RETCODE SCIPresetClock(SCIP *scip, SCIP_CLOCK *clck)
Definition: scip_timing.c:144
#define DEFAULT_PRIORITY_TRUSTREGION
Definition: heur_alns.c:204
#define DEFAULT_COPYCUTS
Definition: heur_alns.c:152
SCIP_RETCODE SCIPreleaseCons(SCIP *scip, SCIP_CONS **cons)
Definition: scip_cons.c:1119
#define DEFAULT_FIXTOL
Definition: heur_alns.c:127
SCIP_Real SCIPvarGetBestRootSol(SCIP_VAR *var)
Definition: var.c:13705
static SCIP_RETCODE getReward(SCIP *scip, SCIP_HEURDATA *heurdata, NH_STATS *runstats, SCIP_Real *rewardptr)
Definition: heur_alns.c:2060
#define CROSSOVERSEED
Definition: heur_alns.c:158
static SCIP_RETCODE setLimits(SCIP *subscip, SOLVELIMITS *solvelimits)
Definition: heur_alns.c:1939
public methods for message output
SCIP_SOLORIGIN SCIPsolGetOrigin(SCIP_SOL *sol)
Definition: sol.c:2545
#define DEFAULT_NSOLS_CROSSOVER
Definition: heur_alns.c:207
SCIP_RETCODE SCIPsetHeurInit(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURINIT((*heurinit)))
Definition: scip_heur.c:194
SCIP_Real SCIPretransformObj(SCIP *scip, SCIP_Real obj)
Definition: scip_sol.c:1576
SCIP_Bool SCIPisFeasPositive(SCIP *scip, SCIP_Real val)
SCIP_Longint SCIPgetMemExternEstim(SCIP *scip)
Definition: scip_mem.c:126
SCIP_NODESEL * SCIPfindNodesel(SCIP *scip, const char *name)
Definition: scip_nodesel.c:234
SCIP_RETCODE SCIPcopyLargeNeighborhoodSearch(SCIP *sourcescip, SCIP *subscip, SCIP_HASHMAP *varmap, const char *suffix, SCIP_VAR **fixedvars, SCIP_Real *fixedvals, int nfixedvars, SCIP_Bool uselprows, SCIP_Bool copycuts, SCIP_Bool *success, SCIP_Bool *valid)
Definition: heuristics.c:925
SCIP_VAR ** SCIPgetVars(SCIP *scip)
Definition: scip_prob.c:1947
SCIP_VARSTATUS SCIPvarGetStatus(SCIP_VAR *var)
Definition: var.c:17361
int statushist[NHISTENTRIES]
Definition: heur_alns.c:357
#define SCIP_Real
Definition: def.h:186
SCIP_Real * SCIPgetWeightsEpsgreedy(SCIP_BANDIT *epsgreedy)
SCIP_RETCODE SCIPresetBandit(SCIP *scip, SCIP_BANDIT *bandit, SCIP_Real *priorities, unsigned int seed)
Definition: scip_bandit.c:91
SCIP_Bool SCIPisStopped(SCIP *scip)
Definition: scip_general.c:703
#define DEFAULT_SEED
Definition: heur_alns.c:156
static int getHistIndex(SCIP_STATUS subscipstatus)
Definition: heur_alns.c:1069
int SCIPbanditGetNActions(SCIP_BANDIT *bandit)
Definition: bandit.c:303
public methods for message handling
public methods for Exp.3
#define SCIP_Longint
Definition: def.h:171
SCIP_Real minfixingrate
Definition: heur_alns.c:364
SCIP_RANDNUMGEN * rng
Definition: heur_alns.c:397
SCIP_Real SCIPfrac(SCIP *scip, SCIP_Real val)
SCIP_VARTYPE SCIPvarGetType(SCIP_VAR *var)
Definition: var.c:17407
SCIP_RANDNUMGEN * SCIPbanditGetRandnumgen(SCIP_BANDIT *bandit)
Definition: bandit.c:293
SCIP_RETCODE SCIPtransformProb(SCIP *scip)
Definition: scip_solve.c:367
#define DEFAULT_PRIORITY_CROSSOVER
Definition: heur_alns.c:189
#define DEFAULT_MAXFIXINGRATE_RINS
Definition: heur_alns.c:167
RewardType
Definition: heur_alns.c:224
#define DEFAULT_ACTIVE_DINS
Definition: heur_alns.c:198
SCIP_RETCODE SCIPsetHeurCopy(SCIP *scip, SCIP_HEUR *heur, SCIP_DECL_HEURCOPY((*heurcopy)))
Definition: scip_heur.c:162
SCIP_Real * randscores
Definition: heur_alns.c:506
SCIP_Bool SCIPisFeasIntegral(SCIP *scip, SCIP_Real val)
#define DEFAULT_MAXFIXINGRATE_CROSSOVER
Definition: heur_alns.c:187
SCIP_Real SCIPsumepsilon(SCIP *scip)
#define NHISTENTRIES
Definition: heur_alns.c:341
SCIP_RETCODE SCIPinterruptSolve(SCIP *scip)
Definition: scip_solve.c:3560
SCIP_Real SCIPgetUpperbound(SCIP *scip)
#define BMSclearMemoryArray(ptr, num)
Definition: memory.h:132
public methods for primal heuristics
#define DEFAULT_USELOCALREDCOST
Definition: heur_alns.c:129
#define MINIMPROVEFAC
Definition: heur_alns.c:112
#define DEFAULT_TARGETNODEFACTOR
Definition: heur_alns.c:101
unsigned int usepscost
Definition: heur_alns.c:512
#define SCIP_CALL_ABORT(x)
Definition: def.h:373
SCIP_HEURDATA * SCIPheurGetData(SCIP_HEUR *heur)
Definition: heur.c:1361
static void computeIntegerVariableBoundsDins(SCIP *scip, SCIP_VAR *var, SCIP_Real *lbptr, SCIP_Real *ubptr)
Definition: heur_alns.c:3393
#define MUTATIONSEED
Definition: heur_alns.c:157
#define DEFAULT_BETA
Definition: heur_alns.c:130
#define TABLE_NAME_NEIGHBORHOOD
Definition: heur_alns.c:217
#define SCIP_ALLOC(x)
Definition: def.h:405
SCIP_Longint SCIPgetNNodes(SCIP *scip)
SCIP_Longint SCIPgetNLPs(SCIP *scip)
#define SCIPABORT()
Definition: def.h:366
public methods for global and local (sub)problems
SCIP_Real SCIPround(SCIP *scip, SCIP_Real val)
SCIP_Real SCIPgetConfidenceBoundUcb(SCIP_BANDIT *ucb, int action)
Definition: bandit_ucb.c:264
SCIP_Bool SCIPvarIsIntegral(SCIP_VAR *var)
Definition: var.c:17433
static SCIP_RETCODE neighborhoodChangeSubscip(SCIP *sourcescip, SCIP *targetscip, NH *neighborhood, SCIP_VAR **targetvars, int *ndomchgs, int *nchgobjs, int *naddedconss, SCIP_Bool *success)
Definition: heur_alns.c:1899
#define TABLE_EARLIEST_STAGE_NEIGHBORHOOD
Definition: heur_alns.c:220
SCIP_RETCODE SCIPstartClock(SCIP *scip, SCIP_CLOCK *clck)
Definition: scip_timing.c:161
static SCIP_RETCODE setupSubScip(SCIP *scip, SCIP *subscip, SCIP_VAR **subvars, SOLVELIMITS *solvelimits, SCIP_HEUR *heur, SCIP_Bool objchgd)
Definition: heur_alns.c:2164
#define DEFAULT_REWARDCONTROL
Definition: heur_alns.c:122
SCIP_Real SCIPgetSolVal(SCIP *scip, SCIP_SOL *sol, SCIP_VAR *var)
Definition: scip_sol.c:1361
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
SCIP_RETCODE SCIPsetSubscipsOff(SCIP *scip, SCIP_Bool quiet)
Definition: scip_param.c:883
SCIP_Real SCIPfloor(SCIP *scip, SCIP_Real val)
public methods for UCB bandit selection
static void increaseMinimumImprovement(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:700
#define FIXINGRATE_STARTINC
Definition: heur_alns.c:150
SCIP_RETCODE SCIPsetLongintParam(SCIP *scip, const char *name, SCIP_Longint value)
Definition: scip_param.c:545
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
int * distances
Definition: heur_alns.c:507
static void resetTargetNodeLimit(SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:644
SCIP_RETCODE SCIPfree(SCIP **scip)
Definition: scip_general.c:324
methods for selecting (weighted) k-medians
#define DEFAULT_ACTIVE_MUTATION
Definition: heur_alns.c:173
SCIP_RETCODE SCIPcreateSol(SCIP *scip, SCIP_SOL **sol, SCIP_HEUR *heur)
Definition: scip_sol.c:328
#define HEUR_PRIORITY
Definition: heur_alns.c:80
memory allocation routines
static SCIP_RETCODE includeNeighborhoods(SCIP *scip, SCIP_HEURDATA *heurdata)
Definition: heur_alns.c:3695