Loading [MathJax]/extensions/MathZoom.js
Scippy

SCIP

Solving Constraint Integer Programs

branch_relpscost.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) 2002-2025 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 branch_relpscost.c
26 * @ingroup DEFPLUGINS_BRANCH
27 * @brief reliable pseudo costs branching rule
28 * @author Tobias Achterberg
29 * @author Timo Berthold
30 * @author Gerald Gamrath
31 * @author Marc Pfetsch
32 */
33
34/*---+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8----+----9----+----0----+----1----+----2*/
35
36#include "blockmemshell/memory.h"
37#include "scip/branch_relpscost.h"
38#include "scip/treemodel.h"
39#include "scip/cons_and.h"
40#include "scip/pub_branch.h"
41#include "scip/pub_cons.h"
42#include "scip/pub_message.h"
43#include "scip/pub_misc.h"
44#include "scip/pub_sol.h"
45#include "scip/pub_tree.h"
46#include "scip/pub_var.h"
47#include "scip/scip_branch.h"
48#include "scip/scip_cons.h"
49#include "scip/scip_general.h"
50#include "scip/scip_lp.h"
51#include "scip/scip_mem.h"
52#include "scip/scip_message.h"
53#include "scip/scip_nlp.h"
54#include "scip/scip_numerics.h"
55#include "scip/scip_param.h"
56#include "scip/scip_prob.h"
57#include "scip/scip_randnumgen.h"
58#include "scip/scip_sol.h"
59#include "scip/scip_solvingstats.h"
60#include "scip/scip_tree.h"
61#include "scip/scip_var.h"
62#include "scip/prop_symmetry.h"
63#include "scip/symmetry.h"
64#include <string.h>
65
66#define BRANCHRULE_NAME "relpscost"
67#define BRANCHRULE_DESC "reliability branching on pseudo cost values"
68#define BRANCHRULE_PRIORITY 10000
69#define BRANCHRULE_MAXDEPTH -1
70#define BRANCHRULE_MAXBOUNDDIST 1.0
71
72#define DEFAULT_CONFLICTWEIGHT 0.01 /**< weight in score calculations for conflict score */
73#define DEFAULT_CONFLENGTHWEIGHT 0.0 /**< weight in score calculations for conflict length score*/
74#define DEFAULT_INFERENCEWEIGHT 0.0001 /**< weight in score calculations for inference score */
75#define DEFAULT_CUTOFFWEIGHT 0.0001 /**< weight in score calculations for cutoff score */
76#define DEFAULT_GMIAVGEFFWEIGHT 0.0 /**< weight in score calculations of average GMI cut normed efficacies */
77#define DEFAULT_GMILASTEFFWEIGHT 0.00001 /**< weight in score calculations of last GMI cut normed efficacy */
78#define DEFAULT_PSCOSTWEIGHT 1.0 /**< weight in score calculations for pseudo cost score */
79#define DEFAULT_NLSCOREWEIGHT 0.1 /**< weight in score calculations for nlcount score */
80#define DEFAULT_MINRELIABLE 1.0 /**< minimal value for minimum pseudo cost size to regard pseudo cost value as reliable */
81#define DEFAULT_MAXRELIABLE 5.0 /**< maximal value for minimum pseudo cost size to regard pseudo cost value as reliable */
82#define DEFAULT_SBITERQUOT 0.5 /**< maximal fraction of strong branching LP iterations compared to normal iterations */
83#define DEFAULT_SBITEROFS 100000 /**< additional number of allowed strong branching LP iterations */
84#define DEFAULT_MAXLOOKAHEAD 9 /**< maximal number of further variables evaluated without better score */
85#define DEFAULT_INITCAND 100 /**< maximal number of candidates initialized with strong branching per node */
86#define DEFAULT_INITITER 0 /**< iteration limit for strong branching initialization of pseudo cost entries (0: auto) */
87#define DEFAULT_MAXBDCHGS 5 /**< maximal number of bound tightenings before the node is reevaluated (-1: unlimited) */
88#define DEFAULT_MAXPROPROUNDS -2 /**< maximum number of propagation rounds to be performed during strong branching
89 * before solving the LP (-1: no limit, -2: parameter settings) */
90#define DEFAULT_PROBINGBOUNDS TRUE /**< should valid bounds be identified in a probing-like fashion during strong
91 * branching (only with propagation)? */
92#define DEFAULT_USERELERRORFORRELIABILITY FALSE /**< should reliability be based on relative errors? */
93#define DEFAULT_LOWERRORTOL 0.05 /**< lowest tolerance beneath which relative errors are reliable */
94#define DEFAULT_HIGHERRORTOL 1.0 /**< highest tolerance beneath which relative errors are reliable */
95#define DEFAULT_USEHYPTESTFORRELIABILITY FALSE /**< should the strong branching decision be based on a hypothesis test? */
96#define DEFAULT_USEDYNAMICCONFIDENCE FALSE /**< should the confidence level be adjusted dynamically? */
97#define DEFAULT_STORESEMIINITCOSTS FALSE /**< should strong branching result be considered for pseudo costs if the other direction was infeasible? */
98#define DEFAULT_USESBLOCALINFO FALSE /**< should the scoring function use only local cutoff and inference information obtained for strong branching candidates? */
99#define DEFAULT_CONFIDENCELEVEL 2 /**< The confidence level for statistical methods, between 0 (Min) and 4 (Max). */
100#define DEFAULT_SKIPBADINITCANDS TRUE /**< should branching rule skip candidates that have a low probability to be
101 * better than the best strong-branching or pseudo-candidate? */
102#define DEFAULT_STARTRANDSEED 5 /**< start random seed for random number generation */
103#define DEFAULT_RANDINITORDER FALSE /**< should slight perturbation of scores be used to break ties in the prior scores? */
104#define DEFAULT_USESMALLWEIGHTSITLIM FALSE /**< should smaller weights be used for pseudo cost updates after hitting the LP iteration limit? */
105#define DEFAULT_DYNAMICWEIGHTS TRUE /**< should the weights of the branching rule be adjusted dynamically during solving based
106 * infeasible and objective leaf counters? */
107#define DEFAULT_DEGENERACYAWARE 1 /**< should degeneracy be taken into account to update weights and skip strong branching? (0: off, 1: after root, 2: always)*/
108
109/* symmetry handling */
110#define DEFAULT_FILTERCANDSSYM FALSE /**< Use symmetry to filter branching candidates? */
111#define DEFAULT_TRANSSYMPSCOST FALSE /**< Transfer pscost information to symmetric variables if filtering is performed? */
112
113/** branching rule data */
114struct SCIP_BranchruleData
115{
116 SCIP_Real conflictweight; /**< weight in score calculations for conflict score */
117 SCIP_Real conflengthweight; /**< weight in score calculations for conflict length score */
118 SCIP_Real inferenceweight; /**< weight in score calculations for inference score */
119 SCIP_Real cutoffweight; /**< weight in score calculations for cutoff score */
120 SCIP_Real gmiavgeffweight; /**< weight in score calculations of average GMI normed cut efficacies */
121 SCIP_Real gmilasteffweight; /**< weight in score calculations of last GMI cut normalized efficacy */
122 SCIP_Real pscostweight; /**< weight in score calculations for pseudo cost score */
123 SCIP_Real nlscoreweight; /**< weight in score calculations for nlcount score */
124 SCIP_Real minreliable; /**< minimal value for minimum pseudo cost size to regard pseudo cost value as reliable */
125 SCIP_Real maxreliable; /**< maximal value for minimum pseudo cost size to regard pseudo cost value as reliable */
126 SCIP_Real sbiterquot; /**< maximal fraction of strong branching LP iterations compared to normal iterations */
127 int sbiterofs; /**< additional number of allowed strong branching LP iterations */
128 int maxlookahead; /**< maximal number of further variables evaluated without better score */
129 int initcand; /**< maximal number of candidates initialized with strong branching per node */
130 int inititer; /**< iteration limit for strong branching initialization of pseudo cost entries (0: auto) */
131 int maxbdchgs; /**< maximal number of bound tightenings before the node is reevaluated (-1: unlimited) */
132 int maxproprounds; /**< maximum number of propagation rounds to be performed during strong branching
133 * before solving the LP (-1: no limit, -2: parameter settings) */
134 SCIP_Bool probingbounds; /**< should valid bounds be identified in a probing-like fashion during strong
135 * branching (only with propagation)? */
136 SCIP_Bool userelerrorforreliability; /**< should reliability be based on relative errors? */
137 SCIP_Real lowerrortol; /**< lowest tolerance beneath which relative errors are reliable */
138 SCIP_Real higherrortol; /**< highest tolerance beneath which relative errors are reliable */
139 SCIP_Bool usehyptestforreliability; /**< should the strong branching decision be based on a hypothesis test? */
140 SCIP_Bool usedynamicconfidence; /**< should the confidence level be adjusted dynamically? */
141 SCIP_Bool storesemiinitcosts; /**< should strong branching result be considered for pseudo costs if the
142 * other direction was infeasible? */
143 SCIP_Bool usesblocalinfo; /**< should the scoring function disregard cutoffs for variable if sb-lookahead was feasible ? */
144 SCIP_Bool skipbadinitcands; /**< should branching rule skip candidates that have a low probability to be
145 * better than the best strong-branching or pseudo-candidate? */
146 SCIP_Bool dynamicweights; /**< should the weights of the branching rule be adjusted dynamically during
147 * solving based on objective and infeasible leaf counters? */
148 int degeneracyaware; /**< should degeneracy be taken into account to update weights and skip strong branching? (0: off, 1: after root, 2: always) */
149 int confidencelevel; /**< The confidence level for statistical methods, between 0 (Min) and 4 (Max). */
150 int* nlcount; /**< array to store nonlinear count values */
151 int nlcountsize; /**< length of nlcount array */
152 int nlcountmax; /**< maximum entry in nlcount array or 1 if NULL */
153 SCIP_Bool randinitorder; /**< should slight perturbation of scores be used to break ties in the prior scores? */
154 SCIP_RANDNUMGEN* randnumgen; /**< random number generator */
155 int startrandseed; /**< start random seed for random number generation */
156 SCIP_Bool usesmallweightsitlim; /**< should smaller weights be used for pseudo cost updates after hitting the LP iteration limit? */
157 SCIP_TREEMODEL* treemodel; /**< Parameters for the Treemodel branching rules */
158
159 /* for symmetry */
160 SCIP_Bool filtercandssym; /**< Use symmetry to filter branching candidates? */
161 SCIP_Bool transsympscost; /**< Transfer pscost information to symmetric variables? */
162
163 SCIP_Bool nosymmetry; /**< No symmetry present? */
164 int* orbits; /**< array of non-trivial orbits */
165 int* orbitbegins; /**< array containing begin positions of new orbits in orbits array */
166 int norbits; /**< pointer to number of orbits currently stored in orbits */
167 int* varorbitmap; /**< array for storing indices of the containing orbit for each variable */
168 int* orbitrep; /**< representative variable of each orbit */
169 SCIP_VAR** permvars; /**< variables on which permutations act */
170 int npermvars; /**< number of variables for permutations */
171 SCIP_HASHMAP* permvarmap; /**< map of variables to indices in permvars array */
172};
173
174/*
175 * local methods
176 */
177
178/** initialize orbits */
179static
180SCIP_RETCODE initOrbits(
181 SCIP* scip, /**< SCIP data structure */
182 SCIP_BRANCHRULEDATA* branchruledata /**< branching rule data */
183 )
184{
185 int** permstrans = NULL;
186 int* components = NULL;
187 int* componentbegins = NULL;
188 int* vartocomponent = NULL;
189 int ncomponents = 0;
190 int nperms = -1;
191
192 assert( scip != NULL );
193 assert( branchruledata != NULL );
194 assert( branchruledata->filtercandssym );
195
196 /* exit if no symmetry or orbits already available */
197 if( branchruledata->nosymmetry || branchruledata->orbits != NULL )
198 return SCIP_OKAY;
199
200 assert( branchruledata->orbitbegins == NULL );
201 assert( branchruledata->varorbitmap == NULL );
202 assert( branchruledata->orbitrep == NULL );
203
204 /* obtain symmetry including permutations */
205 SCIP_CALL( SCIPgetSymmetry(scip, &branchruledata->npermvars, &branchruledata->permvars, &branchruledata->permvarmap,
206 &nperms, NULL, &permstrans, NULL, NULL, &components, &componentbegins, &vartocomponent, &ncomponents) );
207
208 /* turn off symmetry handling if there is no symmetry or the number of variables is not equal */
209 if( nperms <= 0 || branchruledata->npermvars != SCIPgetNVars(scip) )
210 {
211 branchruledata->nosymmetry = TRUE;
212 return SCIP_OKAY;
213 }
214 assert( branchruledata->permvars != NULL );
215 assert( branchruledata->permvarmap != NULL );
216 assert( branchruledata->npermvars > 0 );
217 assert( permstrans != NULL );
218 assert( components != NULL );
219 assert( componentbegins != NULL );
220 assert( vartocomponent != NULL );
221 assert( ncomponents > 0 );
222
223 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &branchruledata->orbits, branchruledata->npermvars) );
224 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &branchruledata->orbitbegins, branchruledata->npermvars) );
225 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &branchruledata->varorbitmap, branchruledata->npermvars) );
226 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &branchruledata->orbitrep, branchruledata->npermvars) );
227
228 /* Compute orbits on all variables, since this might help for branching and this computation is only done once. */
229 SCIP_CALL( SCIPcomputeOrbitsComponentsSym(scip, branchruledata->npermvars, permstrans, nperms,
230 components, componentbegins, vartocomponent, ncomponents,
231 branchruledata->orbits, branchruledata->orbitbegins, &branchruledata->norbits, branchruledata->varorbitmap) );
232 assert( branchruledata->norbits < branchruledata->npermvars );
233
234 return SCIP_OKAY;
235}
236
237/** filter out symmetric variables from branching variables */
238static
239SCIP_RETCODE filterSymmetricVariables(
240 SCIP* scip, /**< SCIP data structure */
241 SCIP_BRANCHRULEDATA* branchruledata, /**< branching rule data */
242 SCIP_VAR** origbranchcands, /**< original branching candidates */
243 SCIP_Real* origbranchcandssol, /**< original solution value for the branching candidates */
244 SCIP_Real* origbranchcandsfrac,/**< original fractional part of the branching candidates */
245 int norigbranchcands, /**< original number of branching candidates */
246 SCIP_VAR** branchcands, /**< branching candidates */
247 SCIP_Real* branchcandssol, /**< solution value for the branching candidates */
248 SCIP_Real* branchcandsfrac, /**< fractional part of the branching candidates */
249 int* branchorbitidx, /**< array of indices of orbit of branching candidates */
250 int* nbranchcands /**< pointer to store number of branching candidates */
251 )
252{
253 int i;
254
255 assert( scip != NULL );
256 assert( branchruledata != NULL );
257 assert( origbranchcands != NULL );
258 assert( origbranchcandssol != NULL );
259 assert( origbranchcandsfrac != NULL );
260 assert( branchcands != NULL );
261 assert( branchcandssol != NULL );
262 assert( branchcandsfrac != NULL );
263 assert( nbranchcands != NULL );
264
265 assert( ! branchruledata->nosymmetry );
266 assert( branchruledata->orbitbegins != NULL );
267 assert( branchruledata->orbits != NULL );
268 assert( branchruledata->permvarmap != NULL );
269 assert( branchruledata->varorbitmap != NULL );
270 assert( branchruledata->orbitrep != NULL );
271 assert( branchruledata->norbits < branchruledata->npermvars );
272
273 /* init representatives (used to see whether variable is the first in an orbit) */
274 for( i = 0; i < branchruledata->norbits; ++i )
275 branchruledata->orbitrep[i] = -1;
276
277 /* loop through branching variables, determine orbit and whether they are the first ones */
278 *nbranchcands = 0;
279 for( i = 0; i < norigbranchcands; ++i )
280 {
281 int orbitidx = -1;
282 int varidx;
283
284 varidx = SCIPhashmapGetImageInt(branchruledata->permvarmap, (void*) origbranchcands[i]);
285 if( varidx != INT_MAX )
286 {
287 assert( 0 <= varidx && varidx < branchruledata->npermvars );
288 orbitidx = branchruledata->varorbitmap[varidx];
289 }
290 assert( -1 <= orbitidx && orbitidx < branchruledata->norbits );
291
292 /* Check whether the variable is not present (can happen if variable was added after computing symmetries or is in
293 * a singleton orbit). */
294 if( orbitidx == -1 )
295 {
296 branchcands[*nbranchcands] = origbranchcands[i];
297 branchcandssol[*nbranchcands] = origbranchcandssol[i];
298 branchcandsfrac[*nbranchcands] = origbranchcandsfrac[i];
299 branchorbitidx[*nbranchcands] = -1;
300 ++(*nbranchcands);
301 }
302 else if( branchruledata->orbitrep[orbitidx] == -1 )
303 {
304 /* if variable is the first in a nontrivial orbit */
305 assert( 0 <= varidx && varidx < branchruledata->npermvars );
306 branchruledata->orbitrep[orbitidx] = varidx;
307 branchcands[*nbranchcands] = origbranchcands[i];
308 branchcandssol[*nbranchcands] = origbranchcandssol[i];
309 branchcandsfrac[*nbranchcands] = origbranchcandsfrac[i];
310 branchorbitidx[*nbranchcands] = orbitidx;
311 ++(*nbranchcands);
312 }
313 }
314
315 SCIPdebugMsg(scip, "Filtered out %d variables by symmetry.\n", norigbranchcands - *nbranchcands);
316
317 return SCIP_OKAY;
318}
319
320/** updates the pseudo costs of the given variable and all its symmetric variables */
321static
322SCIP_RETCODE SCIPupdateVarPseudocostSymmetric(
323 SCIP* scip, /**< SCIP data structure */
324 SCIP_BRANCHRULEDATA* branchruledata, /**< branching rule data */
325 SCIP_VAR* branchvar, /**< branching variable candidate */
326 int* branchorbitidx, /**< array of orbit indices */
327 int branchvaridx, /**< index of variable in branchorbitidx */
328 SCIP_Real solvaldelta, /**< difference of variable's new LP value - old LP value */
329 SCIP_Real objdelta, /**< difference of new LP's objective value - old LP's objective value */
330 SCIP_Real weight /**< weight in (0,1] of this update in pseudo cost sum */
331 )
332{
333 int orbitidx;
334 int j;
335
336 assert( scip != NULL );
337 assert( branchruledata != NULL );
338
339 if( branchruledata->nosymmetry || ! branchruledata->transsympscost || branchorbitidx == NULL )
340 {
341 /* use original update function */
342 SCIP_CALL( SCIPupdateVarPseudocost(scip, branchvar, solvaldelta, objdelta, weight) );
343 return SCIP_OKAY;
344 }
345
346 assert( branchruledata->orbitbegins != NULL );
347 assert( branchruledata->orbits != NULL );
348 assert( 0 <= branchvaridx && branchvaridx < branchruledata->npermvars );
349
350 orbitidx = branchorbitidx[branchvaridx];
351 if( orbitidx < 0 )
352 {
353 /* only update given variable */
354 SCIP_CALL( SCIPupdateVarPseudocost(scip, branchvar, solvaldelta, objdelta, weight) );
355 return SCIP_OKAY;
356 }
357 assert( 0 <= orbitidx && orbitidx < branchruledata->norbits );
358
359 /* loop through orbit containing variable and update pseudo costs for all variables */
360 for( j = branchruledata->orbitbegins[orbitidx]; j < branchruledata->orbitbegins[orbitidx+1]; ++j )
361 {
362 SCIP_VAR* var;
363 int idx;
364
365 idx = branchruledata->orbits[j];
366 assert( 0 <= idx && idx < branchruledata->npermvars );
367
368 var = branchruledata->permvars[idx];
369 assert( var != NULL );
370
371 if( SCIPvarIsActive(var) )
372 {
373 SCIP_CALL( SCIPupdateVarPseudocost(scip, var, solvaldelta, objdelta, weight) );
374 }
375 }
376
377 return SCIP_OKAY;
378}
379
380/**! [SnippetCodeStyleDeclaration] */
381
382/** counts number of nonlinear constraints in which each variable appears */
383static
384SCIP_RETCODE countNonlinearities(
385 SCIP* scip, /**< SCIP data structure */
386 int* nlcount, /**< pointer to array for storing count values */
387 int nlcountsize, /**< buffer for storing length of nlcount array */
388 int* nlcountmax /**< buffer for storing maximum value in nlcount array */
389 )
390{
391 SCIP_CONSHDLR* andconshdlr;
392 SCIP_VAR** vars;
393 int nvars;
394 int i;
395
396/**! [SnippetCodeStyleDeclaration] */
397
398 assert(scip != NULL);
399 assert(nlcount != NULL);
400 assert(nlcountmax != NULL);
401
402 SCIP_CALL( SCIPgetVarsData(scip, &vars, &nvars, NULL, NULL, NULL, NULL) );
403 assert(nlcountsize >= nvars);
404
405 /* get nonlinearity for constraints in NLP */
406 if( SCIPisNLPConstructed(scip) )
407 {
408 assert(SCIPgetNNLPVars(scip) == nvars);
409 SCIP_CALL( SCIPgetNLPVarsNonlinearity(scip, nlcount) );
410 }
411 else
412 {
413 BMSclearMemoryArray(nlcount, nvars);
414 }
415
416 /* increase counters for and constraints */
417 andconshdlr = SCIPfindConshdlr(scip, "and");
418 if( andconshdlr != NULL )
419 {
420 int c;
421
422 for( c = 0; c < SCIPconshdlrGetNActiveConss(andconshdlr); ++c )
423 {
424 SCIP_CONS* andcons;
425 SCIP_VAR** andvars;
426 SCIP_VAR* andres;
427 int probindex;
428 int nandvars;
429 int v;
430
431 /* get constraint and variables */
432 andcons = SCIPconshdlrGetConss(andconshdlr)[c];
433 nandvars = SCIPgetNVarsAnd(scip, andcons);
434 andvars = SCIPgetVarsAnd(scip, andcons);
435 andres = SCIPgetResultantAnd(scip, andcons);
436
437 /* get active index of resultant */
438 probindex = SCIPvarGetProbindex(SCIPvarGetProbvar(andres));
439
440 /* the resultant might be deleted */
441 if( probindex >= 0 )
442 ++nlcount[probindex];
443
444 /**! [SnippetCodeStyleIfFor] */
445 for( v = 0; v < nandvars; ++v )
446 {
447 /* get active index of operator */
448 probindex = SCIPvarGetProbindex(SCIPvarGetProbvar(andvars[v]));
449
450 /* the operator might be deleted */
451 if( probindex >= 0 )
452 ++nlcount[probindex];
453 }
454 /**! [SnippetCodeStyleIfFor] */
455 }
456 }
457
458 /* compute maximum count value */
459 *nlcountmax = 1;
460 for( i = 0; i < nvars; ++i )
461 {
462 if( *nlcountmax < nlcount[i] )
463 *nlcountmax = nlcount[i];
464 }
465
466 return SCIP_OKAY;
467}
468
469static
470SCIP_RETCODE branchruledataEnsureNlcount(
471 SCIP* scip, /**< SCIP data structure */
472 SCIP_BRANCHRULEDATA* branchruledata /**< branching rule data */
473 )
474{
475 int nvars;
476
477 assert(scip != NULL);
478 assert(branchruledata != NULL);
479
480 nvars = SCIPgetNVars(scip);
481
482 /**@todo test whether we want to apply this as if problem has only and constraints */
483 /**@todo update changes in and constraints */
484 if( branchruledata->nlscoreweight > 0.0 ) /* && SCIPisNLPConstructed(scip) */
485 {
486 if( branchruledata->nlcount == NULL )
487 {
488 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &branchruledata->nlcount, nvars) );
489 branchruledata->nlcountsize = nvars;
490
491 SCIP_CALL( countNonlinearities(scip, branchruledata->nlcount, branchruledata->nlcountsize, &branchruledata->nlcountmax) );
492 }
493 else if( branchruledata->nlcountsize < nvars )
494 {
495 SCIP_CALL( SCIPreallocBlockMemoryArray(scip, &branchruledata->nlcount, branchruledata->nlcountsize, nvars) );
496 /**@todo should we update nlcounts for new variables? */
497 BMSclearMemoryArray(&(branchruledata->nlcount[branchruledata->nlcountsize]), nvars - branchruledata->nlcountsize); /*lint !e866*/
498 branchruledata->nlcountsize = nvars;
499 }
500 assert(branchruledata->nlcount != NULL);
501 assert(branchruledata->nlcountsize == nvars);
502 assert(branchruledata->nlcountmax >= 1);
503 }
504 else
505 {
506 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->nlcount, branchruledata->nlcountsize);
507 branchruledata->nlcountsize = 0;
508 branchruledata->nlcountmax = 1;
509 }
510
511 return SCIP_OKAY;
512}
513
514
515/** calculates nlscore value between 0 and 1 */
516static
517SCIP_Real calcNlscore(
518 SCIP* scip, /**< SCIP data structure */
519 int* nlcount, /**< array to store count values */
520 int nlcountmax, /**< maximum value in nlcount array */
521 int probindex /**< index of branching candidate */
522 )
523{
524 if( nlcountmax >= 1 && nlcount != NULL )
525 {
526 SCIP_Real nlscore;
527
528 assert(scip != NULL);
529 assert(probindex >= 0);
530 assert(probindex < SCIPgetNVars(scip));
531
532 nlscore = nlcount[probindex] / (SCIP_Real)nlcountmax;
533
534 assert(nlscore >= 0.0);
535 assert(nlscore <= 1.0);
536 return nlscore;
537 }
538 else
539 return 0.0;
540}
541
542/** calculates an overall score value for the given individual score values */
543static
544SCIP_Real calcScore(
545 SCIP* scip, /**< SCIP data structure */
546 SCIP_BRANCHRULEDATA* branchruledata, /**< branching rule data */
547 SCIP_Real conflictscore, /**< conflict score of current variable */
548 SCIP_Real avgconflictscore, /**< average conflict score */
549 SCIP_Real conflengthscore, /**< conflict length score of current variable */
550 SCIP_Real avgconflengthscore, /**< average conflict length score */
551 SCIP_Real inferencescore, /**< inference score of current variable */
552 SCIP_Real avginferencescore, /**< average inference score */
553 SCIP_Real cutoffscore, /**< cutoff score of current variable */
554 SCIP_Real avgcutoffscore, /**< average cutoff score */
555 SCIP_Real gmieffscore, /**< normalized-eff of avg GMI cuts from row when var was frac and basic */
556 SCIP_Real lastgmieffscore, /**< last normalized gmieffscore when var was frac and basic */
557 SCIP_Real pscostscore, /**< pscost score of current variable */
558 SCIP_Real avgpscostscore, /**< average pscost score */
559 SCIP_Real nlscore, /**< nonlinear score of current variable between 0 and 1 */
560 SCIP_Real frac, /**< fractional value of variable in current solution */
561 SCIP_Real degeneracyfactor /**< factor to apply because of degeneracy */
562 )
563{
564 SCIP_Real score;
565 SCIP_Real dynamicfactor;
566
567 assert(branchruledata != NULL);
568 assert(0.0 < frac && frac < 1.0);
569
570 if( branchruledata->dynamicweights )
571 {
572 dynamicfactor = (SCIPgetNInfeasibleLeaves(scip) + 1.0) / (SCIPgetNObjlimLeaves(scip) + 1.0);
573 }
574 else
575 dynamicfactor = 1.0;
576
577 dynamicfactor *= degeneracyfactor;
578
579 score = dynamicfactor * (branchruledata->conflictweight * (1.0 - 1.0/(1.0+conflictscore/avgconflictscore))
580 + branchruledata->conflengthweight * (1.0 - 1.0/(1.0+conflengthscore/avgconflengthscore))
581 + branchruledata->inferenceweight * (1.0 - 1.0/(1.0+inferencescore/avginferencescore))
582 + branchruledata->cutoffweight * (1.0 - 1.0/(1.0+cutoffscore/avgcutoffscore))
583 + branchruledata->gmiavgeffweight * gmieffscore + branchruledata->gmilasteffweight * lastgmieffscore)
584 + branchruledata->pscostweight / dynamicfactor * (1.0 - 1.0/(1.0+pscostscore/avgpscostscore))
585 + branchruledata->nlscoreweight * nlscore;
586
587 /* avoid close to integral variables */
588 if( MIN(frac, 1.0 - frac) < 10.0 * SCIPfeastol(scip) )
589 score *= 1e-6;
590
591 return score;
592}
593
594/** adds given index and direction to bound change arrays */
595static
596SCIP_RETCODE addBdchg(
597 SCIP* scip, /**< SCIP data structure */
598 int** bdchginds, /**< pointer to bound change index array */
599 SCIP_BOUNDTYPE** bdchgtypes, /**< pointer to bound change types array */
600 SCIP_Real** bdchgbounds, /**< pointer to bound change new bounds array */
601 int* nbdchgs, /**< pointer to number of bound changes */
602 int ind, /**< index to store in bound change index array */
603 SCIP_BOUNDTYPE type, /**< type of the bound change to store in bound change type array */
604 SCIP_Real bound /**< new bound to store in bound change new bounds array */
605 )
606{
607 assert(bdchginds != NULL);
608 assert(bdchgtypes != NULL);
609 assert(bdchgbounds != NULL);
610 assert(nbdchgs != NULL);
611
612 SCIP_CALL( SCIPreallocBufferArray(scip, bdchginds, (*nbdchgs) + 1) );
613 SCIP_CALL( SCIPreallocBufferArray(scip, bdchgtypes, (*nbdchgs) + 1) );
614 SCIP_CALL( SCIPreallocBufferArray(scip, bdchgbounds, (*nbdchgs) + 1) );
615 assert(*bdchginds != NULL);
616 assert(*bdchgtypes != NULL);
617 assert(*bdchgbounds != NULL);
618
619 (*bdchginds)[*nbdchgs] = ind;
620 (*bdchgtypes)[*nbdchgs] = type;
621 (*bdchgbounds)[*nbdchgs] = bound;
622 (*nbdchgs)++;
623
624 return SCIP_OKAY;
625}
626
627/**! [SnippetCodeStyleStaticAsserts] */
628/** frees bound change arrays */
629static
630void freeBdchgs(
631 SCIP* scip, /**< SCIP data structure */
632 int** bdchginds, /**< pointer to bound change index array */
633 SCIP_BOUNDTYPE** bdchgtypes, /**< pointer to bound change types array */
634 SCIP_Real** bdchgbounds, /**< pointer to bound change new bounds array */
635 int* nbdchgs /**< pointer to number of bound changes */
636 )
637{
638 assert(scip != NULL);
639 assert(bdchginds != NULL);
640 assert(bdchgtypes != NULL);
641 assert(bdchgbounds != NULL);
642 assert(nbdchgs != NULL);
643/**! [SnippetCodeStyleStaticAsserts] */
644
645 SCIPfreeBufferArrayNull(scip, bdchgbounds);
646 SCIPfreeBufferArrayNull(scip, bdchgtypes);
647 SCIPfreeBufferArrayNull(scip, bdchginds);
648 *nbdchgs = 0;
649}
650
651/** applies bound changes stored in bound change arrays */
652static
653SCIP_RETCODE applyBdchgs(
654 SCIP* scip, /**< SCIP data structure */
655 SCIP_VAR** vars, /**< problem variables */
656 int* bdchginds, /**< bound change index array */
657 SCIP_BOUNDTYPE* bdchgtypes, /**< bound change types array */
658 SCIP_Real* bdchgbounds, /**< bound change new bound array */
659 int nbdchgs, /**< number of bound changes */
660 SCIP_RESULT* result /**< result pointer */
661 )
662{
663#ifndef NDEBUG
664 SCIP_BRANCHRULE* branchrule;
665 SCIP_BRANCHRULEDATA* branchruledata;
666#endif
667 SCIP_Bool infeasible;
668 SCIP_Bool tightened;
669 int i;
670
671 assert(vars != NULL);
672
673#ifndef NDEBUG
674 /* find branching rule */
675 branchrule = SCIPfindBranchrule(scip, BRANCHRULE_NAME);
676 assert(branchrule != NULL);
677
678 /* get branching rule data */
679 branchruledata = SCIPbranchruleGetData(branchrule);
680 assert(branchruledata != NULL);
681#endif
682
683 SCIPdebugMsg(scip, "applying %d bound changes\n", nbdchgs);
684
685 for( i = 0; i < nbdchgs; ++i )
686 {
687 int v;
688
689 v = bdchginds[i];
690
691 SCIPdebugMsg(scip, " -> <%s> [%g,%g]\n",
692 SCIPvarGetName(vars[v]), SCIPvarGetLbLocal(vars[v]), SCIPvarGetUbLocal(vars[v]));
693
694 if( bdchgtypes[i] == SCIP_BOUNDTYPE_LOWER )
695 {
696 /* change lower bound of variable to given bound */
697 SCIP_CALL( SCIPtightenVarLb(scip, vars[v], bdchgbounds[i], TRUE, &infeasible, &tightened) );
698 if( infeasible )
699 {
700 *result = SCIP_CUTOFF;
701 return SCIP_OKAY;
702 }
703
704 /* if we did propagation, the bound change might already have been added */
705 assert(tightened || (branchruledata->maxproprounds != 0));
706 }
707 else
708 {
709 assert(bdchgtypes[i] == SCIP_BOUNDTYPE_UPPER);
710
711 /* change upper bound of variable to given bound */
712 SCIP_CALL( SCIPtightenVarUb(scip, vars[v], bdchgbounds[i], TRUE, &infeasible, &tightened) );
713 if( infeasible )
714 {
715 *result = SCIP_CUTOFF;
716 return SCIP_OKAY;
717 }
718
719 /* if we did propagation, the bound change might already have been added */
720 assert(tightened || (branchruledata->maxproprounds != 0));
721 }
722 SCIPdebugMsg(scip, " -> [%g,%g]\n", SCIPvarGetLbLocal(vars[v]), SCIPvarGetUbLocal(vars[v]));
723 }
724
725 return SCIP_OKAY;
726}
727
728/** execute reliability pseudo cost branching */
729static
730SCIP_RETCODE execRelpscost(
731 SCIP* scip, /**< SCIP data structure */
732 SCIP_BRANCHRULE* branchrule, /**< branching rule */
733 SCIP_VAR** branchcands, /**< branching candidates */
734 SCIP_Real* branchcandssol, /**< solution value for the branching candidates */
735 SCIP_Real* branchcandsfrac, /**< fractional part of the branching candidates */
736 int* branchorbitidx, /**< indices of orbit (or NULL) */
737 int nbranchcands, /**< number of branching candidates */
738 SCIP_Bool executebranch, /**< execute a branching step or run probing only */
739 SCIP_RESULT* result /**< pointer to the result of the execution */
740 )
741{ /*lint --e{715}*/
742 SCIP_BRANCHRULEDATA* branchruledata;
743 SCIP_Real lpobjval;
744 SCIP_Real bestsbdown;
745 SCIP_Real bestsbup;
746 SCIP_Real provedbound;
747 SCIP_Bool bestsbdownvalid;
748 SCIP_Bool bestsbupvalid;
749 SCIP_Bool bestsbdowncutoff;
750 SCIP_Bool bestsbupcutoff;
751 SCIP_Bool bestisstrongbranch;
752 SCIP_Bool allcolsinlp;
753 SCIP_Bool exactsolve;
754 int ninitcands;
755 int bestcand;
756
757 /* remember which variables strong branching is performed on, and the
758 * recorded lp bound changes that are observed */
759 SCIP_Real* sbdown = NULL;
760 SCIP_Real* sbup = NULL;
761 SCIP_Bool* sbdownvalid = NULL;
762 SCIP_Bool* sbupvalid = NULL;
763
764 *result = SCIP_DIDNOTRUN;
765
766 assert(SCIPgetLPSolstat(scip) == SCIP_LPSOLSTAT_OPTIMAL);
767
768 /* get branching rule data */
769 branchruledata = SCIPbranchruleGetData(branchrule);
770 assert(branchruledata != NULL);
771
772 /* get current LP objective bound of the local sub problem and global cutoff bound */
773 lpobjval = SCIPgetLPObjval(scip);
774
775 /* check, if we want to solve the problem exactly, meaning that strong branching information is not useful
776 * for cutting off sub problems and improving lower bounds of children
777 */
778 exactsolve = SCIPisExactSolve(scip);
779
780 /* check, if all existing columns are in LP, and thus the strong branching results give lower bounds */
781 allcolsinlp = SCIPallColsInLP(scip);
782
783 bestcand = -1;
784 bestisstrongbranch = FALSE;
785 bestsbdown = SCIP_INVALID;
786 bestsbup = SCIP_INVALID;
787 bestsbdownvalid = FALSE;
788 bestsbupvalid = FALSE;
789 bestsbdowncutoff = FALSE;
790 bestsbupcutoff = FALSE;
791 provedbound = lpobjval;
792
793 /* Allocate memory to store the lp bounds of the up and down children
794 * for those of the variables that we performed sb on
795 */
796 SCIP_CALL( SCIPallocBufferArray(scip, &sbdown, nbranchcands) );
797 SCIP_CALL( SCIPallocBufferArray(scip, &sbup, nbranchcands) );
798 SCIP_CALL( SCIPallocBufferArray(scip, &sbdownvalid, nbranchcands) );
799 SCIP_CALL( SCIPallocBufferArray(scip, &sbupvalid, nbranchcands) );
800
801 if( nbranchcands == 1 )
802 {
803 /* only one candidate: nothing has to be done */
804 bestcand = 0;
805 SCIPdebug(ninitcands = 0);
806 sbdownvalid[0] = FALSE;
807 sbupvalid[0] = FALSE;
808 }
809 else
810 {
811 SCIP_VAR** vars;
812 int* initcands;
813 SCIP_Real* initcandscores;
814 SCIP_Real* newlbs = NULL;
815 SCIP_Real* newubs = NULL;
816 SCIP_Real* mingains = NULL;
817 SCIP_Real* maxgains = NULL;
818 /* scores computed from pseudocost branching */
819 SCIP_Real* scores = NULL;
820 SCIP_Real* scoresfrompc = NULL;
821 SCIP_Real* scoresfromothers = NULL;
822 int* bdchginds;
823 SCIP_BOUNDTYPE* bdchgtypes;
824 SCIP_Real* bdchgbounds;
825 int maxninitcands;
826 int nuninitcands;
827 int nbdchgs;
828 int nbdconflicts;
829 SCIP_Real avgconflictscore;
830 SCIP_Real avgconflengthscore;
831 SCIP_Real avginferencescore;
832 SCIP_Real avgcutoffscore;
833 SCIP_Real avgpscostscore;
834 SCIP_Real bestpsscore;
835 SCIP_Real bestpsfracscore;
836 SCIP_Real bestpsdomainscore;
837 SCIP_Real bestsbscore;
838 SCIP_Real bestuninitsbscore;
839 SCIP_Real bestsbfracscore;
840 SCIP_Real bestsbdomainscore;
841 SCIP_Real prio;
842 SCIP_Real reliable;
843 SCIP_Real maxlookahead;
844 SCIP_Real lookahead;
845 SCIP_Real relerrorthreshold;
846 SCIP_Bool initstrongbranching;
847 SCIP_Bool propagate;
848 SCIP_Bool probingbounds;
849 SCIP_Longint nodenum;
850 SCIP_Longint nlpiterationsquot;
851 SCIP_Longint nsblpiterations;
852 SCIP_Longint maxnsblpiterations;
853 int bestsolidx;
854 int maxbdchgs;
855 int bestpscand;
856 int bestsbcand;
857 int bestuninitsbcand;
858 int inititer;
859 int nvars;
860 int i;
861 int c;
862 SCIP_CONFIDENCELEVEL clevel;
863 SCIP_Real degeneracyfactor = 1.0;
864
865 /* get LP degeneracy information and compute a factor to change weighting of pseudo cost score vs. other scores */
866 if( branchruledata->degeneracyaware > 0 && (SCIPgetDepth(scip) > 0 || branchruledata->degeneracyaware > 1) )
867 {
868 SCIP_Real degeneracy;
869 SCIP_Real varconsratio;
870
871 /* get LP degeneracy information */
872 SCIP_CALL( SCIPgetLPDualDegeneracy(scip, &degeneracy, &varconsratio) );
873
874 assert(degeneracy >= 0.0);
875 assert(degeneracy <= 1.0);
876 assert(varconsratio >= 1.0);
877
878 /* increase factor for a degeneracy >= 80% */
879 if( degeneracy >= 0.8 )
880 {
881 degeneracy = 10.0 * (degeneracy - 0.7);
882 degeneracyfactor = degeneracyfactor * pow(10.0,degeneracy);
883 }
884 /* increase factor for a variable-constraint ratio >= 2.0 */
885 if( varconsratio >= 2.0 )
886 {
887 degeneracyfactor *= 10.0 * varconsratio;
888 }
889 }
890
891 vars = SCIPgetVars(scip);
892 nvars = SCIPgetNVars(scip);
893
894 bestsolidx = SCIPgetBestSol(scip) == NULL ? -1 : SCIPsolGetIndex(SCIPgetBestSol(scip));
895
896 /* get average conflict, inference, and pseudocost scores */
897 avgconflictscore = SCIPgetAvgConflictScore(scip);
898 avgconflictscore = MAX(avgconflictscore, 0.1);
899 avgconflengthscore = SCIPgetAvgConflictlengthScore(scip);
900 avgconflengthscore = MAX(avgconflengthscore, 0.1);
901 avginferencescore = SCIPgetAvgInferenceScore(scip);
902 avginferencescore = MAX(avginferencescore, 0.1);
903 avgcutoffscore = SCIPgetAvgCutoffScore(scip);
904 avgcutoffscore = MAX(avgcutoffscore, 0.1);
905 avgpscostscore = SCIPgetAvgPseudocostScore(scip);
906 avgpscostscore = MAX(avgpscostscore, 0.1);
907
908 /* get nonlinear counts according to parameters */
909 SCIP_CALL( branchruledataEnsureNlcount(scip, branchruledata) );
910
911 initstrongbranching = FALSE;
912
913 /* check whether propagation should be performed */
914 propagate = (branchruledata->maxproprounds != 0);
915
916 /* check whether valid bounds should be identified in probing-like fashion */
917 probingbounds = propagate && branchruledata->probingbounds;
918
919 /* get maximal number of candidates to initialize with strong branching; if the current solutions is not basic,
920 * we cannot warmstart the simplex algorithm and therefore don't initialize any candidates
921 */
922 maxninitcands = MIN(nbranchcands, branchruledata->initcand);
923 if( !SCIPisLPSolBasic(scip) )
924 maxninitcands = 0;
925
926 /* calculate maximal number of strong branching LP iterations; if we used too many, don't apply strong branching
927 * any more; also, if degeneracy is too high, don't run strong branching at this node
928 */
929 nlpiterationsquot = (SCIP_Longint)(branchruledata->sbiterquot * SCIPgetNNodeLPIterations(scip));
930 maxnsblpiterations = nlpiterationsquot + branchruledata->sbiterofs + SCIPgetNRootStrongbranchLPIterations(scip);
931 nsblpiterations = SCIPgetNStrongbranchLPIterations(scip);
932 if( nsblpiterations > maxnsblpiterations || degeneracyfactor >= 10.0 )
933 maxninitcands = 0;
934
935 /* get buffer for storing the unreliable candidates */
936 SCIP_CALL( SCIPallocBufferArray(scip, &initcands, maxninitcands+1) ); /* allocate one additional slot for convenience */
937 SCIP_CALL( SCIPallocBufferArray(scip, &initcandscores, maxninitcands+1) );
938 ninitcands = 0;
939
940 /* Allocate memory for the down and up gains, and the computed pseudocost scores */
941 SCIP_CALL( SCIPallocBufferArray(scip, &mingains, nbranchcands) );
942 SCIP_CALL( SCIPallocBufferArray(scip, &maxgains, nbranchcands) );
943 SCIP_CALL( SCIPallocBufferArray(scip, &scores, nbranchcands) );
944 SCIP_CALL( SCIPallocBufferArray(scip, &scoresfrompc, nbranchcands) );
945 SCIP_CALL( SCIPallocBufferArray(scip, &scoresfromothers, nbranchcands) );
946
947 /* get current node number */
948 nodenum = SCIPgetNNodes(scip);
949
950 /* initialize bound change arrays */
951 bdchginds = NULL;
952 bdchgtypes = NULL;
953 bdchgbounds = NULL;
954 nbdchgs = 0;
955 nbdconflicts = 0;
956 maxbdchgs = branchruledata->maxbdchgs;
957 if( maxbdchgs == -1 )
958 maxbdchgs = INT_MAX;
959
960 /* calculate value used as reliability */
961 prio = (maxnsblpiterations - nsblpiterations)/(nsblpiterations + 1.0);
962 prio = MIN(prio, 1.0);
963 prio = MAX(prio, (nlpiterationsquot - nsblpiterations)/(nsblpiterations + 1.0));
964 reliable = (1.0-prio) * branchruledata->minreliable + prio * branchruledata->maxreliable;
965
966 /* calculate the threshold for the relative error in the same way; low tolerance is more strict than higher tolerance */
967 relerrorthreshold = (1.0 - prio) * branchruledata->higherrortol + prio * branchruledata->lowerrortol;
968
969 clevel = (SCIP_CONFIDENCELEVEL)branchruledata->confidencelevel;
970 /* determine the confidence level for hypothesis testing based on value of prio */
971 if( branchruledata->usedynamicconfidence )
972 {
973 /* with decreasing priority, use a less strict confidence level */
974 if( prio >= 0.9 )
975 clevel = SCIP_CONFIDENCELEVEL_MAX;
976 else if( prio >= 0.7 )
977 clevel = SCIP_CONFIDENCELEVEL_HIGH;
978 else if( prio >= 0.5 )
979 clevel = SCIP_CONFIDENCELEVEL_MEDIUM;
980 else if( prio >= 0.3 )
981 clevel = SCIP_CONFIDENCELEVEL_LOW;
982 else
983 clevel = SCIP_CONFIDENCELEVEL_MIN;
984 }
985
986 /* search for the best pseudo cost candidate, while remembering unreliable candidates in a sorted buffer */
987 nuninitcands = 0;
988 bestpscand = -1;
989 bestpsscore = -SCIPinfinity(scip);
990 bestpsfracscore = -SCIPinfinity(scip);
991 bestpsdomainscore = -SCIPinfinity(scip);
992
993 /* search for the best candidate first */
994 if( branchruledata->usehyptestforreliability )
995 {
996 for( c = 0; c < nbranchcands; ++c )
997 {
998 SCIP_Real conflictscore;
999 SCIP_Real conflengthscore;
1000 SCIP_Real inferencescore;
1001 SCIP_Real cutoffscore;
1002 SCIP_Real gmieffscore;
1003 SCIP_Real lastgmieffscore;
1004 SCIP_Real pscostscore;
1005 SCIP_Real nlscore;
1006 SCIP_Real score;
1007
1008 conflictscore = SCIPgetVarConflictScore(scip, branchcands[c]);
1009 conflengthscore = SCIPgetVarConflictlengthScore(scip, branchcands[c]);
1010 inferencescore = SCIPgetVarAvgInferenceScore(scip, branchcands[c]);
1011 cutoffscore = SCIPgetVarAvgCutoffScore(scip, branchcands[c]);
1012 gmieffscore = SCIPgetVarAvgGMIScore(scip, branchcands[c]);
1013 lastgmieffscore = SCIPgetVarLastGMIScore(scip, branchcands[c]);
1014 nlscore = calcNlscore(scip, branchruledata->nlcount, branchruledata->nlcountmax, SCIPvarGetProbindex(branchcands[c]));
1015 pscostscore = SCIPgetVarPseudocostScore(scip, branchcands[c], branchcandssol[c]);
1016
1017 /* replace the pseudo cost score with the already calculated one;
1018 * @todo: use old data for strong branching with propagation?
1019 */
1020 if( SCIPgetVarStrongbranchNode(scip, branchcands[c]) == nodenum )
1021 {
1022 SCIP_Real down;
1023 SCIP_Real up;
1024 SCIP_Real lastlpobjval;
1025 SCIP_Real downgain;
1026 SCIP_Real upgain;
1027
1028 /* use the score of the strong branching call at the current node */
1029 SCIP_CALL( SCIPgetVarStrongbranchLast(scip, branchcands[c], &down, &up, NULL, NULL, NULL, &lastlpobjval) );
1030 downgain = MAX(down - lastlpobjval, 0.0);
1031 upgain = MAX(up - lastlpobjval, 0.0);
1032 pscostscore = SCIPgetBranchScore(scip, branchcands[c], downgain, upgain);
1033
1034 SCIPdebugMsg(scip, " -> strong branching on variable <%s> already performed (down=%g (%+g), up=%g (%+g), pscostscore=%g)\n",
1035 SCIPvarGetName(branchcands[c]), down, downgain, up, upgain, pscostscore);
1036 }
1037
1038 score = calcScore(scip, branchruledata, conflictscore, avgconflictscore, conflengthscore, avgconflengthscore,
1039 inferencescore, avginferencescore, cutoffscore, avgcutoffscore, gmieffscore, lastgmieffscore,
1040 pscostscore, avgpscostscore, nlscore, branchcandsfrac[c], degeneracyfactor);
1041
1042 /* check for better score of candidate */
1043 if( SCIPisSumGE(scip, score, bestpsscore) )
1044 {
1045 SCIP_Real fracscore;
1046 SCIP_Real domainscore;
1047
1048 fracscore = MIN(branchcandsfrac[c], 1.0 - branchcandsfrac[c]);
1049 domainscore = -(SCIPvarGetUbLocal(branchcands[c]) - SCIPvarGetLbLocal(branchcands[c]));
1050 if( SCIPisSumGT(scip, score, bestpsscore)
1051 || SCIPisSumGT(scip, fracscore, bestpsfracscore)
1052 || (SCIPisSumGE(scip, fracscore, bestpsfracscore) && domainscore > bestpsdomainscore) )
1053 {
1054 bestpscand = c;
1055 bestpsscore = score;
1056 bestpsfracscore = fracscore;
1057 bestpsdomainscore = domainscore;
1058 }
1059 }
1060 }
1061 }
1062
1063 for( c = 0; c < nbranchcands; ++c )
1064 {
1065 SCIP_Real conflictscore;
1066 SCIP_Real conflengthscore;
1067 SCIP_Real inferencescore;
1068 SCIP_Real cutoffscore;
1069 SCIP_Real gmieffscore;
1070 SCIP_Real lastgmieffscore;
1071 SCIP_Real pscostscore;
1072 SCIP_Real nlscore;
1073 SCIP_Real score;
1074 SCIP_Bool usesb;
1075 SCIP_Real downgain;
1076 SCIP_Real upgain;
1077 SCIP_Real fracpart;
1078
1079 assert(branchcands[c] != NULL);
1080 assert(!SCIPisFeasIntegral(scip, branchcandssol[c]));
1081 assert(!SCIPisFeasIntegral(scip, SCIPvarGetLPSol(branchcands[c])));
1082
1083 /* Record the variables current pseudocosts. These may be overwritten if
1084 * strong branching is performed.
1085 */
1086 sbdownvalid[c] = FALSE;
1087 sbupvalid[c] = FALSE;
1088 fracpart = SCIPfeasFrac(scip, SCIPvarGetLPSol(branchcands[c]));
1089 downgain = SCIPgetVarPseudocostVal(scip, branchcands[c], 0.0 - fracpart);
1090 upgain = SCIPgetVarPseudocostVal(scip, branchcands[c], 1.0 - fracpart);
1091 mingains[c] = MIN(downgain, upgain);
1092 maxgains[c] = MAX(downgain, upgain);
1093
1094 /* get conflict, inference, cutoff, nonlinear, and pseudo cost scores for candidate */
1095 conflictscore = SCIPgetVarConflictScore(scip, branchcands[c]);
1096 conflengthscore = SCIPgetVarConflictlengthScore(scip, branchcands[c]);
1097 inferencescore = SCIPgetVarAvgInferenceScore(scip, branchcands[c]);
1098 cutoffscore = SCIPgetVarAvgCutoffScore(scip, branchcands[c]);
1099 gmieffscore = SCIPgetVarAvgGMIScore(scip, branchcands[c]);
1100 lastgmieffscore = SCIPgetVarLastGMIScore(scip, branchcands[c]);
1101 nlscore = calcNlscore(scip, branchruledata->nlcount, branchruledata->nlcountmax, SCIPvarGetProbindex(branchcands[c]));
1102 pscostscore = SCIPgetVarPseudocostScore(scip, branchcands[c], branchcandssol[c]);
1103 usesb = FALSE;
1104
1105 /* don't use strong branching on variables that have already been initialized at the current node;
1106 * instead replace the pseudo cost score with the already calculated one;
1107 * @todo: use old data for strong branching with propagation?
1108 */
1109 if( SCIPgetVarStrongbranchNode(scip, branchcands[c]) == nodenum )
1110 {
1111 SCIP_Real down;
1112 SCIP_Real up;
1113 SCIP_Real lastlpobjval;
1114
1115 /* use the score of the strong branching call at the current node */
1116 SCIP_CALL( SCIPgetVarStrongbranchLast(scip, branchcands[c], &down, &up, NULL, NULL, NULL, &lastlpobjval) );
1117 downgain = MAX(down - lastlpobjval, 0.0);
1118 upgain = MAX(up - lastlpobjval, 0.0);
1119 pscostscore = SCIPgetBranchScore(scip, branchcands[c], downgain, upgain);
1120
1121 mingains[c] = MIN(downgain, upgain);
1122 maxgains[c] = MAX(downgain, upgain);
1123
1124 SCIPdebugMsg(scip, " -> strong branching on variable <%s> already performed (down=%g (%+g), up=%g (%+g), pscostscore=%g)\n",
1125 SCIPvarGetName(branchcands[c]), down, downgain, up, upgain, pscostscore);
1126 }
1127 else if( maxninitcands > 0 )
1128 {
1129 SCIP_Real downsize;
1130 SCIP_Real upsize;
1131 SCIP_Real size;
1132
1133 /* check, if the pseudo cost score of the variable is reliable */
1134 downsize = SCIPgetVarPseudocostCountCurrentRun(scip, branchcands[c], SCIP_BRANCHDIR_DOWNWARDS);
1135 upsize = SCIPgetVarPseudocostCountCurrentRun(scip, branchcands[c], SCIP_BRANCHDIR_UPWARDS);
1136 size = MIN(downsize, upsize);
1137
1138 /* determine if variable is considered reliable based on the current reliability setting */
1139 /* check fixed number threshold (aka original) reliability first */
1140 assert(!branchruledata->usehyptestforreliability || bestpscand >= 0);
1141 usesb = FALSE;
1142 if( size < reliable )
1143 usesb = TRUE;
1144 else if( branchruledata->userelerrorforreliability && branchruledata->usehyptestforreliability )
1145 {
1146 if( !SCIPisVarPscostRelerrorReliable(scip, branchcands[c], relerrorthreshold, clevel) &&
1147 !SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], branchcandsfrac[bestpscand],
1148 branchcands[c], branchcandsfrac[c], SCIP_BRANCHDIR_DOWNWARDS, clevel, TRUE) &&
1149 !SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], 1 - branchcandsfrac[bestpscand],
1150 branchcands[c], 1 - branchcandsfrac[c], SCIP_BRANCHDIR_UPWARDS, clevel, TRUE) )
1151 usesb = TRUE;
1152 }
1153 /* check if relative error is tolerable */
1154 else if( branchruledata->userelerrorforreliability &&
1155 !SCIPisVarPscostRelerrorReliable(scip, branchcands[c], relerrorthreshold, clevel))
1156 usesb = TRUE;
1157 /* check if best pseudo-candidate is significantly better in both directions, use strong-branching otherwise */
1158 else if( branchruledata->usehyptestforreliability &&
1159 !SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], branchcandsfrac[bestpscand],
1160 branchcands[c], branchcandsfrac[c], SCIP_BRANCHDIR_DOWNWARDS, clevel, TRUE) &&
1161 !SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], 1 - branchcandsfrac[bestpscand],
1162 branchcands[c], 1 - branchcandsfrac[c], SCIP_BRANCHDIR_UPWARDS, clevel, TRUE))
1163 usesb = TRUE;
1164
1165 /* count the number of variables that are completely uninitialized */
1166 if( size < 0.1 )
1167 nuninitcands++;
1168 }
1169
1170 /* combine the five score values */
1171 scoresfrompc[c] = calcScore(scip, branchruledata, 0.0, avgconflictscore, 0.0, avgconflengthscore,
1172 0.0, avginferencescore, 0.0, avgcutoffscore, 0.0, 0.0,
1173 pscostscore, avgpscostscore, 0.0, branchcandsfrac[c], degeneracyfactor);
1174 scoresfromothers[c] = calcScore(scip, branchruledata, conflictscore, avgconflictscore, conflengthscore, avgconflengthscore,
1175 inferencescore, avginferencescore, cutoffscore, avgcutoffscore, gmieffscore, lastgmieffscore,
1176 0.0, avgpscostscore, nlscore, branchcandsfrac[c], degeneracyfactor);
1177 score = scoresfrompc[c] + scoresfromothers[c];
1178 scores[c] = score;
1179 /*score = calcScore(scip, branchruledata, conflictscore, avgconflictscore, conflengthscore, avgconflengthscore,
1180 inferencescore, avginferencescore, cutoffscore, avgcutoffscore, gmieffscore, lastgmieffscore,
1181 pscostscore, avgpscostscore, nlscore, branchcandsfrac[c], degeneracyfactor);*/
1182 if( usesb )
1183 {
1184 int j;
1185
1186 mingains[c] = 0;
1187 maxgains[c] = 0;
1188 scoresfrompc[c] = 0;
1189 scoresfromothers[c] = 0;
1190
1191 /* assign a random score to this uninitialized candidate */
1192 if( branchruledata->randinitorder )
1193 score += SCIPrandomGetReal(branchruledata->randnumgen, 0.0, 1e-4);
1194
1195 /* pseudo cost of variable is not reliable: insert candidate in initcands buffer */
1196 for( j = ninitcands; j > 0 && score > initcandscores[j-1]; --j )
1197 {
1198 initcands[j] = initcands[j-1];
1199 initcandscores[j] = initcandscores[j-1];
1200 }
1201 initcands[j] = c;
1202 initcandscores[j] = score;
1203 ninitcands++;
1204 ninitcands = MIN(ninitcands, maxninitcands);
1205 }
1206 /* in the case of hypothesis reliability, the best pseudo candidate has been determined already */
1207 else if( !branchruledata->usehyptestforreliability )
1208 {
1209 /* variable will keep its pseudo cost value: check for better score of candidate */
1210 if( SCIPisSumGE(scip, score, bestpsscore) )
1211 {
1212 SCIP_Real fracscore;
1213 SCIP_Real domainscore;
1214
1215 fracscore = MIN(branchcandsfrac[c], 1.0 - branchcandsfrac[c]);
1216 domainscore = -(SCIPvarGetUbLocal(branchcands[c]) - SCIPvarGetLbLocal(branchcands[c]));
1217 if( SCIPisSumGT(scip, score, bestpsscore)
1218 || SCIPisSumGT(scip, fracscore, bestpsfracscore)
1219 || (SCIPisSumGE(scip, fracscore, bestpsfracscore) && domainscore > bestpsdomainscore) )
1220 {
1221 bestpscand = c;
1222 bestpsscore = score;
1223 bestpsfracscore = fracscore;
1224 bestpsdomainscore = domainscore;
1225 }
1226 }
1227 }
1228 }
1229
1230 /* in the special case that only the best pseudo candidate was selected for strong branching, skip the strong branching */
1231 if( branchruledata->usehyptestforreliability && ninitcands == 1 )
1232 {
1233 ninitcands = 0;
1234 SCIPdebugMsg(scip, "Only one single candidate for initialization-->Skipping strong branching\n");
1235 }
1236
1237 /* initialize unreliable candidates with strong branching until maxlookahead is reached,
1238 * search best strong branching candidate
1239 */
1240 maxlookahead = (SCIP_Real)branchruledata->maxlookahead * (1.0 + (SCIP_Real)nuninitcands/(SCIP_Real)nbranchcands);
1241 inititer = branchruledata->inititer;
1242 if( inititer == 0 )
1243 {
1244 SCIP_Longint nlpiterations;
1245 SCIP_Longint nlps;
1246
1247 /* iteration limit is set to twice the average number of iterations spent to resolve a dual feasible SCIP_LP;
1248 * at the first few nodes, this average is not very exact, so we better increase the iteration limit on
1249 * these very important nodes
1250 */
1251 nlpiterations = SCIPgetNDualResolveLPIterations(scip);
1252 nlps = SCIPgetNDualResolveLPs(scip);
1253 if( nlps == 0 )
1254 {
1255 nlpiterations = SCIPgetNNodeInitLPIterations(scip);
1256 nlps = SCIPgetNNodeInitLPs(scip);
1257 if( nlps == 0 )
1258 {
1259 nlpiterations = 1000;
1260 nlps = 1;
1261 }
1262 }
1263 assert(nlps >= 1);
1264 inititer = (int)(2*nlpiterations / nlps);
1265 inititer = (int)((SCIP_Real)inititer * (1.0 + 20.0/nodenum));
1266 inititer = MAX(inititer, 10);
1267 inititer = MIN(inititer, 500);
1268 }
1269
1270 SCIPdebugMsg(scip, "strong branching (reliable=%g, %d/%d cands, %d uninit, maxcands=%d, maxlookahead=%g, maxbdchgs=%d, inititer=%d, iterations:%" SCIP_LONGINT_FORMAT "/%" SCIP_LONGINT_FORMAT ", basic:%u)\n",
1271 reliable, ninitcands, nbranchcands, nuninitcands, maxninitcands, maxlookahead, maxbdchgs, inititer,
1272 SCIPgetNStrongbranchLPIterations(scip), maxnsblpiterations, SCIPisLPSolBasic(scip));
1273
1274 bestsbcand = -1;
1275 bestsbscore = -SCIPinfinity(scip);
1276 bestsbfracscore = -SCIPinfinity(scip);
1277 bestsbdomainscore = -SCIPinfinity(scip);
1278 bestuninitsbscore = -SCIPinfinity(scip);
1279 bestuninitsbcand = -1;
1280 lookahead = 0.0;
1281 for( i = 0; i < ninitcands && lookahead < maxlookahead && nbdchgs + nbdconflicts < maxbdchgs
1282 && (i < (int) maxlookahead || SCIPgetNStrongbranchLPIterations(scip) < maxnsblpiterations); ++i )
1283 {
1284 SCIP_Real down;
1285 SCIP_Real up;
1286 SCIP_Real downgain;
1287 SCIP_Real upgain;
1288 SCIP_Bool downvalid;
1289 SCIP_Bool upvalid;
1290 SCIP_Longint ndomredsdown;
1291 SCIP_Longint ndomredsup;
1292 SCIP_Bool lperror;
1293 SCIP_Bool downinf;
1294 SCIP_Bool upinf;
1295 SCIP_Bool downconflict;
1296 SCIP_Bool upconflict;
1297
1298 /* get candidate number to initialize */
1299 c = initcands[i];
1300 assert(!SCIPisFeasIntegral(scip, branchcandssol[c]));
1301
1302 if( branchruledata->skipbadinitcands )
1303 {
1304 SCIP_Bool skipsb = FALSE;
1305 /* if the current best candidate is a candidate found by strong branching, determine if candidate pseudo-costs are
1306 * significantly smaller in at least one direction, in which case we safe the execution of strong-branching for now
1307 */
1308 if( bestsbscore > bestpsscore && bestsbscore > bestuninitsbscore && bestsbupvalid && bestsbdownvalid )
1309 {
1310 assert(bestsbcand != -1);
1311 assert(bestsbup != SCIP_INVALID && bestsbdown != SCIP_INVALID); /*lint !e777 lint doesn't like comparing floats */
1312
1313 /* test if the variable is unlikely to produce a better gain than the currently best one. Skip strong-branching
1314 * in such a case
1315 */
1316 if( SCIPpscostThresholdProbabilityTest(scip, branchcands[c], branchcandsfrac[c], bestsbdown,
1317 SCIP_BRANCHDIR_DOWNWARDS, clevel)
1318 || SCIPpscostThresholdProbabilityTest(scip, branchcands[c], 1.0 - branchcandsfrac[c], bestsbup,
1319 SCIP_BRANCHDIR_UPWARDS, clevel) )
1320 skipsb = TRUE;
1321 }
1322 /* the currently best candidate is also a pseudo-candidate; apply significance test and skip candidate if it
1323 * is significantly worse in at least one direction
1324 */
1325 else if( bestpscand != -1 && bestpsscore > bestuninitsbscore )
1326 {
1327 if( SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], branchcandsfrac[bestpscand],
1328 branchcands[c], branchcandsfrac[c], SCIP_BRANCHDIR_DOWNWARDS, clevel, TRUE)
1329 || SCIPsignificantVarPscostDifference(scip, branchcands[bestpscand], 1.0 - branchcandsfrac[bestpscand],
1330 branchcands[c], 1.0 - branchcandsfrac[c], SCIP_BRANCHDIR_UPWARDS, clevel, TRUE) )
1331 skipsb = TRUE;
1332 }
1333 /* compare against the best init cand that has been skipped already */
1334 else if( bestuninitsbcand != -1 )
1335 {
1336 if( SCIPsignificantVarPscostDifference(scip, branchcands[bestuninitsbcand], branchcandsfrac[bestuninitsbcand],
1337 branchcands[c], branchcandsfrac[c], SCIP_BRANCHDIR_DOWNWARDS, clevel, TRUE)
1338 || SCIPsignificantVarPscostDifference(scip, branchcands[bestuninitsbcand], 1.0 - branchcandsfrac[bestuninitsbcand],
1339 branchcands[c], 1.0 - branchcandsfrac[c], SCIP_BRANCHDIR_UPWARDS, clevel, TRUE) )
1340 skipsb = TRUE;
1341 }
1342
1343 /* skip candidate, update the best score of an unitialized candidate */
1344 if( skipsb )
1345 {
1346 if( bestuninitsbcand == -1 )
1347 {
1348 bestuninitsbcand = c;
1349 bestuninitsbscore = initcandscores[i];
1350 }
1351 continue;
1352 }
1353 }
1354 SCIPdebugMsg(scip, "init pseudo cost (%g/%g) of <%s> at %g (score:%g) with strong branching (%d iterations) -- %" SCIP_LONGINT_FORMAT "/%" SCIP_LONGINT_FORMAT " iterations\n",
1355 SCIPgetVarPseudocostCountCurrentRun(scip, branchcands[c], SCIP_BRANCHDIR_DOWNWARDS),
1356 SCIPgetVarPseudocostCountCurrentRun(scip, branchcands[c], SCIP_BRANCHDIR_UPWARDS),
1357 SCIPvarGetName(branchcands[c]), branchcandssol[c], initcandscores[i],
1358 inititer, SCIPgetNStrongbranchLPIterations(scip), maxnsblpiterations);
1359
1360 /* use strong branching on candidate */
1361 if( !initstrongbranching )
1362 {
1363 initstrongbranching = TRUE;
1364
1365 SCIP_CALL( SCIPstartStrongbranch(scip, propagate) );
1366
1367 /* create arrays for probing-like bound tightening */
1368 if( probingbounds )
1369 {
1370 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &newlbs, nvars) );
1371 SCIP_CALL( SCIPallocBlockMemoryArray(scip, &newubs, nvars) );
1372 }
1373 }
1374
1375 if( propagate )
1376 {
1377 /* apply strong branching */
1378 SCIP_CALL( SCIPgetVarStrongbranchWithPropagation(scip, branchcands[c], branchcandssol[c], lpobjval, inititer,
1379 branchruledata->maxproprounds, &down, &up, &downvalid, &upvalid, &ndomredsdown, &ndomredsup, &downinf, &upinf,
1380 &downconflict, &upconflict, &lperror, newlbs, newubs) );
1381 }
1382 else
1383 {
1384 /* apply strong branching */
1385 SCIP_CALL( SCIPgetVarStrongbranchFrac(scip, branchcands[c], inititer, FALSE,
1386 &down, &up, &downvalid, &upvalid, &downinf, &upinf, &downconflict, &upconflict, &lperror) );
1387
1388 ndomredsdown = ndomredsup = 0;
1389 }
1390
1391 /* check for an error in strong branching */
1392 if( lperror )
1393 {
1394 if( !SCIPisStopped(scip) )
1395 {
1396 SCIPverbMessage(scip, SCIP_VERBLEVEL_HIGH, NULL,
1397 "(node %" SCIP_LONGINT_FORMAT ") error in strong branching call for variable <%s> with solution %g\n",
1398 SCIPgetNNodes(scip), SCIPvarGetName(branchcands[c]), branchcandssol[c]);
1399 }
1400 break;
1401 }
1402
1403 /* Strong branching might have found a new primal solution which updated the cutoff bound. In this case, the
1404 * provedbound computed before can be higher than the cutoffbound and the current node can be cut off.
1405 * Additionally, also if the value for the current best candidate is valid and exceeds the new cutoff bound,
1406 * we want to change the domain of this variable rather than branching on it.
1407 */
1408 if( SCIPgetBestSol(scip) != NULL && SCIPsolGetIndex(SCIPgetBestSol(scip)) != bestsolidx )
1409 {
1410 bestsolidx = SCIPsolGetIndex(SCIPgetBestSol(scip));
1411
1412 SCIPdebugMsg(scip, " -> strong branching on variable <%s> lead to a new incumbent\n",
1413 SCIPvarGetName(branchcands[c]));
1414
1415 /* proved bound for current node is larger than new cutoff bound -> cut off current node */
1416 if( SCIPisGE(scip, provedbound, SCIPgetCutoffbound(scip)) )
1417 {
1418 SCIPdebugMsg(scip, " -> node can be cut off (provedbound=%g, cutoff=%g)\n", provedbound, SCIPgetCutoffbound(scip));
1419
1420 *result = SCIP_CUTOFF;
1421 break; /* terminate initialization loop, because node is infeasible */
1422 }
1423 /* proved bound for down child of best candidate is larger than cutoff bound
1424 * -> increase lower bound of best candidate
1425 * we must only do this if the LP is complete, i.e., we are not doing column generation
1426 */
1427
1428 else if( bestsbcand != -1 && allcolsinlp )
1429 {
1430 if( !bestsbdowncutoff && bestsbdownvalid && SCIPisGE(scip, bestsbdown, SCIPgetCutoffbound(scip)) )
1431 {
1432 bestsbdowncutoff = TRUE;
1433
1434 SCIPdebugMsg(scip, " -> valid dual bound for down child of best candidate <%s> is higher than new cutoff bound (valid=%u, bestsbdown=%g, cutoff=%g)\n",
1435 SCIPvarGetName(branchcands[bestsbcand]), bestsbdownvalid, bestsbdown, SCIPgetCutoffbound(scip));
1436
1437 SCIPdebugMsg(scip, " -> increase lower bound of best candidate <%s> to %g\n",
1438 SCIPvarGetName(branchcands[bestsbcand]), SCIPfeasCeil(scip, branchcandssol[bestsbcand]));
1439
1440 SCIP_CALL( addBdchg(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs, SCIPvarGetProbindex(branchcands[bestsbcand]),
1441 SCIP_BOUNDTYPE_LOWER, SCIPfeasCeil(scip, branchcandssol[bestsbcand])) );
1442 }
1443 /* proved bound for up child of best candidate is larger than cutoff bound -> decrease upper bound of best candidate */
1444 else if( !bestsbupcutoff && bestsbupvalid && SCIPisGE(scip, bestsbup, SCIPgetCutoffbound(scip)) )
1445 {
1446 bestsbupcutoff = TRUE;
1447
1448 SCIPdebugMsg(scip, " -> valid dual bound for up child of best candidate <%s> is higher than new cutoff bound (valid=%u, bestsbup=%g, cutoff=%g)\n",
1449 SCIPvarGetName(branchcands[bestsbcand]), bestsbupvalid, bestsbup, SCIPgetCutoffbound(scip));
1450
1451 SCIPdebugMsg(scip, " -> decrease upper bound of best candidate <%s> to %g\n",
1452 SCIPvarGetName(branchcands[bestsbcand]), SCIPfeasFloor(scip, branchcandssol[bestsbcand]));
1453
1454 SCIP_CALL( addBdchg(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs, SCIPvarGetProbindex(branchcands[bestsbcand]),
1455 SCIP_BOUNDTYPE_UPPER, SCIPfeasFloor(scip, branchcandssol[bestsbcand])) );
1456 }
1457 }
1458 }
1459
1460 /* evaluate strong branching */
1461 down = MAX(down, lpobjval);
1462 up = MAX(up, lpobjval);
1463 downgain = down - lpobjval;
1464 upgain = up - lpobjval;
1465 assert(!allcolsinlp || exactsolve || !downvalid || downinf == SCIPisGE(scip, down, SCIPgetCutoffbound(scip)));
1466 assert(!allcolsinlp || exactsolve || !upvalid || upinf == SCIPisGE(scip, up, SCIPgetCutoffbound(scip)));
1467 assert(downinf || !downconflict);
1468 assert(upinf || !upconflict);
1469
1470 /* @todo: store pseudo cost only for valid bounds?
1471 * depending on the user parameter choice of storesemiinitcosts, pseudo costs are also updated in single directions,
1472 * if the node in the other direction was infeasible or cut off
1473 */
1474 if( !downinf
1475#ifdef WITH_LPSOLSTAT
1476 && SCIPgetLastStrongbranchLPSolStat(scip, SCIP_BRANCHDIR_DOWNWARDS) != SCIP_LPSOLSTAT_ITERLIMIT
1477#endif
1478 && (!upinf || branchruledata->storesemiinitcosts) )
1479 {
1480 SCIP_Real weight;
1481
1482 /* smaller weights are given if the strong branching hit the time limit in the corresponding direction */
1483 if( branchruledata->usesmallweightsitlim )
1484 weight = SCIPgetLastStrongbranchLPSolStat(scip, SCIP_BRANCHDIR_DOWNWARDS) != SCIP_LPSOLSTAT_ITERLIMIT ? 1.0 : 0.5;
1485 else
1486 weight = 1.0;
1487
1488 /* update pseudo cost values */
1489 SCIP_CALL( SCIPupdateVarPseudocostSymmetric(scip, branchruledata, branchcands[c], branchorbitidx, c, 0.0 - branchcandsfrac[c], downgain, weight) );
1490 }
1491 if( !upinf
1492#ifdef WITH_LPSOLSTAT
1493 && SCIPgetLastStrongbranchLPSolStat(scip, SCIP_BRANCHDIR_UPWARDS) != SCIP_LPSOLSTAT_ITERLIMIT
1494#endif
1495 && (!downinf || branchruledata->storesemiinitcosts) )
1496 {
1497 SCIP_Real weight;
1498
1499 /* smaller weights are given if the strong branching hit the time limit in the corresponding direction */
1500 if( branchruledata->usesmallweightsitlim )
1501 weight = SCIPgetLastStrongbranchLPSolStat(scip, SCIP_BRANCHDIR_UPWARDS) != SCIP_LPSOLSTAT_ITERLIMIT ? 1.0 : 0.5;
1502 else
1503 weight = 1.0;
1504
1505 /* update pseudo cost values */
1506 SCIP_CALL( SCIPupdateVarPseudocostSymmetric(scip, branchruledata, branchcands[c], branchorbitidx, c, 1.0 - branchcandsfrac[c], upgain, weight) );
1507 }
1508
1509 /* the minimal lower bound of both children is a proved lower bound of the current subtree */
1510 if( allcolsinlp && !exactsolve && downvalid && upvalid )
1511 {
1512 SCIP_Real minbound;
1513
1514 minbound = MIN(down, up);
1515 provedbound = MAX(provedbound, minbound);
1516 assert((downinf && upinf) || SCIPisLT(scip, provedbound, SCIPgetCutoffbound(scip)));
1517
1518 /* save probing-like bounds detected during strong branching */
1519 if( probingbounds && ( !downinf || !upinf ) )
1520 {
1521 int v;
1522
1523 assert(newlbs != NULL);
1524 assert(newubs != NULL);
1525
1526 for( v = 0; v < nvars; ++v )
1527 {
1528 if( SCIPisGT(scip, newlbs[v], SCIPvarGetLbLocal(vars[v])) )
1529 {
1530 SCIPdebugMsg(scip, "better lower bound for variable <%s>: %.9g -> %.9g (by strongbranching on <%s>)\n",
1531 SCIPvarGetName(vars[v]), SCIPvarGetLbLocal(vars[v]), newlbs[v], SCIPvarGetName(branchcands[c]));
1532
1533 SCIP_CALL( addBdchg(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs, v,
1534 SCIP_BOUNDTYPE_LOWER, newlbs[v]) );
1535 }
1536 if( SCIPisLT(scip, newubs[v], SCIPvarGetUbLocal(vars[v])) )
1537 {
1538 SCIPdebugMsg(scip, "better upper bound for variable <%s>: %.9g -> %.9g (by strongbranching on <%s>)\n",
1539 SCIPvarGetName(vars[v]), SCIPvarGetUbLocal(vars[v]), newubs[v], SCIPvarGetName(branchcands[c]));
1540
1541 SCIP_CALL( addBdchg(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs, v,
1542 SCIP_BOUNDTYPE_UPPER, newubs[v]) );
1543 }
1544 }
1545 }
1546 }
1547
1548 /* check if there are infeasible roundings */
1549 if( downinf || upinf )
1550 {
1551 assert(allcolsinlp || propagate);
1552 assert(!exactsolve);
1553
1554 if( downinf && upinf )
1555 {
1556 /* both roundings are infeasible -> node is infeasible */
1557 SCIPdebugMsg(scip, " -> variable <%s> is infeasible in both directions (conflict: %u/%u)\n",
1558 SCIPvarGetName(branchcands[c]), downconflict, upconflict);
1559 *result = SCIP_CUTOFF;
1560 break; /* terminate initialization loop, because node is infeasible */
1561 }
1562 else
1563 {
1564 /* rounding is infeasible in one direction -> round variable in other direction */
1565 SCIPdebugMsg(scip, " -> variable <%s> is infeasible in %s branch (conflict: %u/%u)\n",
1566 SCIPvarGetName(branchcands[c]), downinf ? "downward" : "upward", downconflict, upconflict);
1567 SCIP_CALL( addBdchg(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs, SCIPvarGetProbindex(branchcands[c]),
1568 (downinf ? SCIP_BOUNDTYPE_LOWER : SCIP_BOUNDTYPE_UPPER),
1569 (downinf ? SCIPfeasCeil(scip, branchcandssol[c]) : SCIPfeasFloor(scip, branchcandssol[c]))) );
1570 if( nbdchgs + nbdconflicts >= maxbdchgs )
1571 break; /* terminate initialization loop, because enough roundings are performed */
1572 }
1573 }
1574 else
1575 {
1576 SCIP_Real conflictscore;
1577 SCIP_Real conflengthscore;
1578 SCIP_Real inferencescore;
1579 SCIP_Real cutoffscore;
1580 SCIP_Real gmieffscore;
1581 SCIP_Real lastgmieffscore;
1582 SCIP_Real pscostscore;
1583 SCIP_Real nlscore;
1584 SCIP_Real score;
1585
1586 mingains[c] = MIN(downgain, upgain);
1587 maxgains[c] = MAX(downgain, upgain);
1588
1589 sbdown[c] = down;
1590 sbup[c] = up;
1591 sbdownvalid[c] = downvalid;
1592 sbupvalid[c] = upvalid;
1593
1594 /* check for a better score */
1595 conflictscore = SCIPgetVarConflictScore(scip, branchcands[c]);
1596 conflengthscore = SCIPgetVarConflictlengthScore(scip, branchcands[c]);
1597 nlscore = calcNlscore(scip, branchruledata->nlcount, branchruledata->nlcountmax, SCIPvarGetProbindex(branchcands[c]));
1598
1599 /* optionally, use only local information obtained via strong branching for this candidate, i.e., local
1600 * domain reductions and no cutoff score
1601 */
1602 inferencescore = branchruledata->usesblocalinfo ? SCIPgetBranchScore(scip, branchcands[c], (SCIP_Real)ndomredsdown, (SCIP_Real)ndomredsup)
1603 : SCIPgetVarAvgInferenceScore(scip, branchcands[c]);
1604 cutoffscore = branchruledata->usesblocalinfo ? 0.0 : SCIPgetVarAvgCutoffScore(scip, branchcands[c]);
1605 gmieffscore = branchruledata->usesblocalinfo ? 0.0 : SCIPgetVarAvgGMIScore(scip, branchcands[c]);
1606 lastgmieffscore = branchruledata->usesblocalinfo ? 0.0 : SCIPgetVarLastGMIScore(scip, branchcands[c]);
1607 pscostscore = SCIPgetBranchScore(scip, branchcands[c], downgain, upgain);
1608
1609 scoresfrompc[c] = calcScore(scip, branchruledata, 0.0, avgconflictscore, 0.0, avgconflengthscore,
1610 0.0, avginferencescore, 0.0, avgcutoffscore, 0.0, 0.0, pscostscore,
1611 avgpscostscore, 0.0, branchcandsfrac[c], degeneracyfactor);
1612 scoresfromothers[c] = calcScore(scip, branchruledata, conflictscore, avgconflictscore, conflengthscore, avgconflengthscore,
1613 inferencescore, avginferencescore, cutoffscore, avgcutoffscore, gmieffscore,
1614 lastgmieffscore, 0.0, avgpscostscore, nlscore, branchcandsfrac[c],
1615 degeneracyfactor);
1616 score = scoresfrompc[c] + scoresfromothers[c];
1617 scores[c] = score;
1618 /*score = calcScore(scip, branchruledata, conflictscore, avgconflictscore, conflengthscore, avgconflengthscore,
1619 inferencescore, avginferencescore, cutoffscore, avgcutoffscore, gmieffscore, lastgmieffscore,
1620 pscostscore, avgpscostscore, nlscore, branchcandsfrac[c], degeneracyfactor);*/
1621
1622 if( SCIPisSumGE(scip, score, bestsbscore) )
1623 {
1624 SCIP_Real fracscore;
1625 SCIP_Real domainscore;
1626
1627 fracscore = MIN(branchcandsfrac[c], 1.0 - branchcandsfrac[c]);
1628 domainscore = -(SCIPvarGetUbLocal(branchcands[c]) - SCIPvarGetLbLocal(branchcands[c]));
1629 if( SCIPisSumGT(scip, score, bestsbscore)
1630 || SCIPisSumGT(scip, fracscore, bestsbfracscore)
1631 || (SCIPisSumGE(scip, fracscore, bestsbfracscore) && domainscore > bestsbdomainscore) )
1632 {
1633 bestsbcand = c;
1634 bestsbscore = score;
1635 bestsbdown = down;
1636 bestsbup = up;
1637 bestsbdownvalid = downvalid;
1638 bestsbupvalid = upvalid;
1639 bestsbdowncutoff = FALSE;
1640 bestsbupcutoff = FALSE;
1641 bestsbfracscore = fracscore;
1642 bestsbdomainscore = domainscore;
1643 lookahead = 0.0;
1644 }
1645 else
1646 lookahead += 0.5;
1647 }
1648 else
1649 lookahead += 1.0;
1650
1651 SCIPdebugMsg(scip, " -> variable <%s> (solval=%g, down=%g (%+g,valid=%u), up=%g (%+g,valid=%u), score=%g/ %g/%g %g/%g %g -> %g)\n",
1652 SCIPvarGetName(branchcands[c]), branchcandssol[c], down, downgain, downvalid, up, upgain, upvalid,
1653 pscostscore, conflictscore, conflengthscore, inferencescore, cutoffscore, gmieffscore, score);
1654 }
1655 }
1656#ifdef SCIP_DEBUG
1657 if( bestsbcand >= 0 )
1658 {
1659 SCIPdebugMsg(scip, " -> best: <%s> (%g / %g / %g), lookahead=%g/%g\n",
1660 SCIPvarGetName(branchcands[bestsbcand]), bestsbscore, bestsbfracscore, bestsbdomainscore,
1661 lookahead, maxlookahead);
1662 }
1663#endif
1664
1665 if( initstrongbranching )
1666 {
1667 if( probingbounds )
1668 {
1669 assert(newlbs != NULL);
1670 assert(newubs != NULL);
1671
1672 SCIPfreeBlockMemoryArray(scip, &newubs, nvars);
1673 SCIPfreeBlockMemoryArray(scip, &newlbs, nvars);
1674 }
1675
1676 SCIP_CALL( SCIPendStrongbranch(scip) );
1677
1678 if( SCIPgetLPSolstat(scip) == SCIP_LPSOLSTAT_OBJLIMIT || SCIPgetLPSolstat(scip) == SCIP_LPSOLSTAT_INFEASIBLE )
1679 {
1680 assert(SCIPhasCurrentNodeLP(scip));
1681 *result = SCIP_CUTOFF;
1682 }
1683 }
1684
1685 if( *result != SCIP_CUTOFF )
1686 {
1687 /* get the score of the best uninitialized strong branching candidate */
1688 if( i < ninitcands && bestuninitsbcand == -1 )
1689 bestuninitsbscore = initcandscores[i];
1690
1691 /* if the best pseudo cost candidate is better than the best uninitialized strong branching candidate,
1692 * compare it to the best initialized strong branching candidate
1693 */
1694 if( bestpsscore > bestuninitsbscore && SCIPisSumGT(scip, bestpsscore, bestsbscore) )
1695 {
1696 bestcand = bestpscand;
1697 bestisstrongbranch = FALSE;
1698 }
1699 else if( bestsbcand >= 0 )
1700 {
1701 bestcand = bestsbcand;
1702 bestisstrongbranch = TRUE;
1703 }
1704 else
1705 {
1706 /* no candidate was initialized, and the best score is the one of the first candidate in the initialization
1707 * queue
1708 */
1709 assert(ninitcands >= 1);
1710 bestcand = initcands[0];
1711 bestisstrongbranch = FALSE;
1712 }
1713
1714 /* update lower bound of current node */
1715 if( allcolsinlp && !exactsolve )
1716 {
1717 assert(SCIPisLT(scip, provedbound, SCIPgetCutoffbound(scip)));
1718 SCIP_CALL( SCIPupdateLocalLowerbound(scip, provedbound) );
1719 assert(SCIPisGE(scip, SCIPgetLocalLowerbound(scip), provedbound));
1720 }
1721 }
1722
1723 /* apply domain reductions */
1724 if( nbdchgs > 0 )
1725 {
1726 if( *result != SCIP_CUTOFF )
1727 {
1728 SCIP_CALL( applyBdchgs(scip, vars, bdchginds, bdchgtypes, bdchgbounds, nbdchgs, result) );
1729 if( *result != SCIP_CUTOFF )
1730 *result = SCIP_REDUCEDDOM;
1731 }
1732 freeBdchgs(scip, &bdchginds, &bdchgtypes, &bdchgbounds, &nbdchgs);
1733 }
1734
1735 /* Apply the Treemodel branching rule to potentially select a better branching candidate than the current one. */
1736 if( *result != SCIP_CUTOFF && *result != SCIP_REDUCEDDOM && *result != SCIP_CONSADDED && SCIPtreemodelIsEnabled(scip, branchruledata->treemodel) )
1737 {
1738 SCIP_Real smallpscost;
1739 SCIP_Bool usetreemodel;
1740
1741 usetreemodel = TRUE;
1742
1743 /* If the pseudocosts are zero, use SCIPs best variable since the Treemodel is not applicable */
1744 if( SCIPisZero(scip, maxgains[bestcand]))
1745 {
1746 usetreemodel = FALSE;
1747 }
1748
1749 /* If SCIPs best candidate was selected due to hybrid branching scores
1750 * rather than because of pseudocosts, then we keep it.
1751 */
1752 SCIP_CALL( SCIPgetRealParam(scip, "branching/treemodel/smallpscost", &smallpscost) );
1753 if( usetreemodel == TRUE && avgpscostscore <= smallpscost )
1754 {
1755 int cand;
1756 for( cand = 0; cand < nbranchcands; ++cand )
1757 {
1758 if( scoresfrompc[cand] > scoresfrompc[bestcand] )
1759 {
1760 usetreemodel = FALSE;
1761 break;
1762 }
1763 }
1764 }
1765
1766 if( usetreemodel == TRUE )
1767 {
1768 SCIP_CALL( SCIPtreemodelSelectCandidate(
1769 scip, /* SCIP data structure */
1770 branchruledata->treemodel, /* branching rule */
1771 branchcands, /* branching candidate storage */
1772 mingains, /* minimum gain of rounding downwards or upwards */
1773 maxgains, /* maximum gain of rounding downwards or upwards */
1774 scoresfromothers, /* scores from other branching methods */
1775 nbranchcands, /* the number of branching candidates */
1776 &bestcand /* the best branching candidate found by SCIP */
1777 ) );
1778 }
1779 }
1780
1781 /* free buffer for the lp gains and pseudocost scores */
1782 SCIPfreeBufferArray(scip, &scoresfromothers);
1783 SCIPfreeBufferArray(scip, &scoresfrompc);
1784 SCIPfreeBufferArray(scip, &scores);
1785 SCIPfreeBufferArray(scip, &maxgains);
1786 SCIPfreeBufferArray(scip, &mingains);
1787
1788 /* free buffer for the unreliable candidates */
1789 SCIPfreeBufferArray(scip, &initcandscores);
1790 SCIPfreeBufferArray(scip, &initcands);
1791 }
1792
1793 /* if no domain could be reduced, create the branching */
1794 if( *result != SCIP_CUTOFF && *result != SCIP_REDUCEDDOM && *result != SCIP_CONSADDED && executebranch )
1795 {
1796 SCIP_NODE* downchild;
1797 SCIP_NODE* upchild;
1798 SCIP_VAR* var;
1799 SCIP_Real val;
1800
1801 assert(*result == SCIP_DIDNOTRUN);
1802 assert(0 <= bestcand && bestcand < nbranchcands);
1803 assert(!SCIPisFeasIntegral(scip, branchcandssol[bestcand]));
1804 assert(!allcolsinlp || SCIPisLT(scip, provedbound, SCIPgetCutoffbound(scip)));
1805 assert(!bestsbdowncutoff && !bestsbupcutoff);
1806
1807 var = branchcands[bestcand];
1808 val = branchcandssol[bestcand];
1809
1810 /* perform the branching */
1811 SCIPdebugMsg(scip, " -> %d (%d) cands, sel cand %d: var <%s> (sol=%g, down=%g (%+g), up=%g (%+g), sb=%u, psc=%g/%g [%g])\n",
1812 nbranchcands, ninitcands, bestcand, SCIPvarGetName(var), branchcandssol[bestcand],
1813 bestsbdown, bestsbdown - lpobjval, bestsbup, bestsbup - lpobjval, bestisstrongbranch,
1814 SCIPgetVarPseudocostCurrentRun(scip, var, SCIP_BRANCHDIR_DOWNWARDS),
1815 SCIPgetVarPseudocostCurrentRun(scip, var, SCIP_BRANCHDIR_UPWARDS),
1816 SCIPgetVarPseudocostScoreCurrentRun(scip, var, branchcandssol[bestcand]));
1817 SCIP_UNUSED(bestisstrongbranch);
1818 SCIP_CALL( SCIPbranchVarVal(scip, var, val, &downchild, NULL, &upchild) );
1819 assert(downchild != NULL);
1820 assert(upchild != NULL);
1821
1822 /* update the lower bounds in the children */
1823 if( allcolsinlp && !exactsolve )
1824 {
1825 if( sbdownvalid[bestcand] )
1826 {
1827 assert(SCIPisLT(scip, sbdown[bestcand], SCIPgetCutoffbound(scip)));
1828 SCIP_CALL( SCIPupdateNodeLowerbound(scip, downchild, sbdown[bestcand]) );
1829 assert(SCIPisGE(scip, SCIPgetNodeLowerbound(scip, downchild), provedbound));
1830 }
1831 if( sbupvalid[bestcand] )
1832 {
1833 assert(SCIPisLT(scip, sbup[bestcand], SCIPgetCutoffbound(scip)));
1834 SCIP_CALL( SCIPupdateNodeLowerbound(scip, upchild, sbup[bestcand]) );
1835 assert(SCIPisGE(scip, SCIPgetNodeLowerbound(scip, upchild), provedbound));
1836 }
1837 }
1838
1839 SCIPdebugMsg(scip, " -> down child's lowerbound: %g\n", SCIPnodeGetLowerbound(downchild));
1840 SCIPdebugMsg(scip, " -> up child's lowerbound : %g\n", SCIPnodeGetLowerbound(upchild));
1841
1842 assert(SCIPgetLPSolstat(scip) != SCIP_LPSOLSTAT_INFEASIBLE && SCIPgetLPSolstat(scip) != SCIP_LPSOLSTAT_OBJLIMIT);
1843
1844 *result = SCIP_BRANCHED;
1845 }
1846
1847 /* free buffer for the strong branching lp gains */
1848 SCIPfreeBufferArray(scip, &sbupvalid);
1849 SCIPfreeBufferArray(scip, &sbdownvalid);
1850 SCIPfreeBufferArray(scip, &sbup);
1851 SCIPfreeBufferArray(scip, &sbdown);
1852
1853 return SCIP_OKAY;
1854}
1855
1856
1857/*
1858 * Callback methods
1859 */
1860
1861/** copy method for branchrule plugins (called when SCIP copies plugins) */
1862static
1863SCIP_DECL_BRANCHCOPY(branchCopyRelpscost)
1864{ /*lint --e{715}*/
1865 assert(scip != NULL);
1866 assert(branchrule != NULL);
1867 assert(strcmp(SCIPbranchruleGetName(branchrule), BRANCHRULE_NAME) == 0);
1868
1869 /* call inclusion method of branchrule */
1870 SCIP_CALL( SCIPincludeBranchruleRelpscost(scip) );
1871
1872 return SCIP_OKAY;
1873}
1874
1875/** destructor of branching rule to free user data (called when SCIP is exiting) */
1876static
1877SCIP_DECL_BRANCHFREE(branchFreeRelpscost)
1878{ /*lint --e{715}*/
1879 SCIP_BRANCHRULEDATA* branchruledata;
1880 branchruledata = SCIPbranchruleGetData(branchrule);
1881
1882 /* free Treemodel parameter data structure */
1883 SCIP_CALL( SCIPtreemodelFree(scip, &branchruledata->treemodel) );
1884
1885 /* free branching rule data */
1886 SCIPfreeBlockMemory(scip, &branchruledata);
1887 SCIPbranchruleSetData(branchrule, NULL);
1888
1889 return SCIP_OKAY;
1890}
1891
1892
1893/** solving process initialization method of branching rule (called when branch and bound process is about to begin) */
1894static
1895SCIP_DECL_BRANCHINITSOL(branchInitsolRelpscost)
1896{ /*lint --e{715}*/
1897 SCIP_BRANCHRULEDATA* branchruledata;
1898
1899 /* initialize branching rule data */
1900 branchruledata = SCIPbranchruleGetData(branchrule);
1901 branchruledata->nlcount = NULL;
1902 branchruledata->nlcountsize = 0;
1903 branchruledata->nlcountmax = 1;
1904 assert(branchruledata->startrandseed >= 0);
1905
1906 /* create a random number generator */
1907 SCIP_CALL( SCIPcreateRandom(scip, &branchruledata->randnumgen,
1908 (unsigned int)branchruledata->startrandseed, TRUE) );
1909
1910 return SCIP_OKAY;
1911}
1912
1913
1914/** solving process deinitialization method of branching rule (called before branch and bound process data is freed) */
1915static
1916SCIP_DECL_BRANCHEXITSOL(branchExitsolRelpscost)
1917{ /*lint --e{715}*/
1918 SCIP_BRANCHRULEDATA* branchruledata;
1919
1920 /* free memory in branching rule data */
1921 branchruledata = SCIPbranchruleGetData(branchrule);
1922 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->nlcount, branchruledata->nlcountsize);
1923
1924 /* free random number generator */
1925 SCIPfreeRandom(scip, &branchruledata->randnumgen);
1926
1927 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->orbitrep, branchruledata->npermvars);
1928 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->varorbitmap, branchruledata->npermvars);
1929 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->orbitbegins, branchruledata->npermvars);
1930 SCIPfreeBlockMemoryArrayNull(scip, &branchruledata->orbits, branchruledata->npermvars);
1931 branchruledata->nosymmetry = FALSE;
1932 branchruledata->norbits = 0;
1933 branchruledata->permvars = NULL;
1934 branchruledata->permvarmap = NULL;
1935 branchruledata->npermvars = 0;
1936
1937 return SCIP_OKAY;
1938}
1939
1940
1941/** branching execution method for fractional LP solutions */
1942static
1943SCIP_DECL_BRANCHEXECLP(branchExeclpRelpscost)
1944{ /*lint --e{715}*/
1945 SCIP_BRANCHRULEDATA* branchruledata;
1946 SCIP_VAR** lpcands;
1947 SCIP_Real* lpcandssol;
1948 SCIP_Real* lpcandsfrac;
1949 SCIP_VAR** filteredlpcands;
1950 SCIP_Real* filteredlpcandssol;
1951 SCIP_Real* filteredlpcandsfrac;
1952 SCIP_Bool runfiltering;
1953 int* filteredlpcandsorbitidx = NULL;
1954 int nfilteredlpcands;
1955 int nlpcands;
1956
1957 assert(branchrule != NULL);
1958 assert(strcmp(SCIPbranchruleGetName(branchrule), BRANCHRULE_NAME) == 0);
1959 assert(scip != NULL);
1960 assert(result != NULL);
1961
1962 SCIPdebugMsg(scip, "Execlp method of relpscost branching in node %" SCIP_LONGINT_FORMAT "\n", SCIPnodeGetNumber(SCIPgetCurrentNode(scip)));
1963
1964 if( SCIPgetLPSolstat(scip) != SCIP_LPSOLSTAT_OPTIMAL )
1965 {
1966 *result = SCIP_DIDNOTRUN;
1967 SCIPdebugMsg(scip, "Could not apply relpscost branching, as the current LP was not solved to optimality.\n");
1968
1969 return SCIP_OKAY;
1970 }
1971
1972 /* get branching candidates */
1973 SCIP_CALL( SCIPgetLPBranchCands(scip, &lpcands, &lpcandssol, &lpcandsfrac, NULL, &nlpcands, NULL) );
1974 assert(nlpcands > 0);
1975
1976 branchruledata = SCIPbranchruleGetData(branchrule);
1977 assert(branchruledata != NULL);
1978
1979 /* determine whether we should run filtering */
1980 runfiltering = ! branchruledata->nosymmetry && branchruledata->filtercandssym && SCIPgetSubscipDepth(scip) == 0 && ! SCIPinDive(scip) && ! SCIPinProbing(scip);
1981
1982 /* init orbits if necessary */
1983 if( runfiltering )
1984 {
1985 SCIP_CALL( initOrbits(scip, branchruledata) );
1986 }
1987
1988 /* determine fractional variables (possibly filter by using symmetries) */
1989 if( runfiltering && branchruledata->norbits != 0 )
1990 {
1991 SCIP_CALL( SCIPallocBufferArray(scip, &filteredlpcands, nlpcands) );
1992 SCIP_CALL( SCIPallocBufferArray(scip, &filteredlpcandssol, nlpcands) );
1993 SCIP_CALL( SCIPallocBufferArray(scip, &filteredlpcandsfrac, nlpcands) );
1994 SCIP_CALL( SCIPallocBufferArray(scip, &filteredlpcandsorbitidx, nlpcands) );
1995
1996 /* determine filtered fractional variables */
1997 SCIP_CALL( filterSymmetricVariables(scip, branchruledata, lpcands, lpcandssol, lpcandsfrac, nlpcands,
1998 filteredlpcands, filteredlpcandssol, filteredlpcandsfrac, filteredlpcandsorbitidx, &nfilteredlpcands) );
1999 }
2000 else
2001 {
2002 /* No orbits available. Copy all (unfiltered) branching candidates, because they will be updated w.r.t. the strong branching LP solution */
2003 SCIP_CALL( SCIPduplicateBufferArray(scip, &filteredlpcands, lpcands, nlpcands) );
2004 SCIP_CALL( SCIPduplicateBufferArray(scip, &filteredlpcandssol, lpcandssol, nlpcands) );
2005 SCIP_CALL( SCIPduplicateBufferArray(scip, &filteredlpcandsfrac, lpcandsfrac, nlpcands) );
2006 nfilteredlpcands = nlpcands;
2007 }
2008
2009 /* execute branching rule */
2010 SCIP_CALL( execRelpscost(scip, branchrule, filteredlpcands, filteredlpcandssol, filteredlpcandsfrac, filteredlpcandsorbitidx, nfilteredlpcands, TRUE, result) );
2011
2012 SCIPfreeBufferArrayNull(scip, &filteredlpcandsorbitidx);
2013 SCIPfreeBufferArray(scip, &filteredlpcandsfrac);
2014 SCIPfreeBufferArray(scip, &filteredlpcandssol);
2015 SCIPfreeBufferArray(scip, &filteredlpcands);
2016
2017 return SCIP_OKAY;
2018}
2019
2020
2021/*
2022 * branching specific interface methods
2023 */
2024
2025/** creates the reliable pseudo cost branching rule and includes it in SCIP */
2026SCIP_RETCODE SCIPincludeBranchruleRelpscost(
2027 SCIP* scip /**< SCIP data structure */
2028 )
2029{
2030 SCIP_BRANCHRULEDATA* branchruledata;
2031 SCIP_BRANCHRULE* branchrule;
2032
2033 /* create relpscost branching rule data */
2034 SCIP_CALL( SCIPallocBlockMemory(scip, &branchruledata) );
2035
2036 branchruledata->nosymmetry = FALSE;
2037 branchruledata->orbits = NULL;
2038 branchruledata->orbitbegins = NULL;
2039 branchruledata->orbitrep = NULL;
2040 branchruledata->varorbitmap = NULL;
2041 branchruledata->norbits = 0;
2042 branchruledata->permvars = NULL;
2043 branchruledata->npermvars = 0;
2044 branchruledata->permvarmap = NULL;
2045
2046 /* include branching rule */
2047 SCIP_CALL( SCIPincludeBranchruleBasic(scip, &branchrule, BRANCHRULE_NAME, BRANCHRULE_DESC, BRANCHRULE_PRIORITY,
2048 BRANCHRULE_MAXDEPTH, BRANCHRULE_MAXBOUNDDIST, branchruledata) );
2049
2050 assert(branchrule != NULL);
2051
2052 /* set non-fundamental callbacks via specific setter functions*/
2053 SCIP_CALL( SCIPsetBranchruleCopy(scip, branchrule, branchCopyRelpscost) );
2054 SCIP_CALL( SCIPsetBranchruleFree(scip, branchrule, branchFreeRelpscost) );
2055 SCIP_CALL( SCIPsetBranchruleInitsol(scip, branchrule, branchInitsolRelpscost) );
2056 SCIP_CALL( SCIPsetBranchruleExitsol(scip, branchrule, branchExitsolRelpscost) );
2057 SCIP_CALL( SCIPsetBranchruleExecLp(scip, branchrule, branchExeclpRelpscost) );
2058
2059 /* relpscost branching rule parameters */
2060 SCIP_CALL( SCIPaddRealParam(scip,
2061 "branching/relpscost/conflictweight",
2062 "weight in score calculations for conflict score",
2063 &branchruledata->conflictweight, TRUE, DEFAULT_CONFLICTWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2064 SCIP_CALL( SCIPaddRealParam(scip,
2065 "branching/relpscost/conflictlengthweight",
2066 "weight in score calculations for conflict length score",
2067 &branchruledata->conflengthweight, TRUE, DEFAULT_CONFLENGTHWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2068 SCIP_CALL( SCIPaddRealParam(scip,
2069 "branching/relpscost/inferenceweight",
2070 "weight in score calculations for inference score",
2071 &branchruledata->inferenceweight, TRUE, DEFAULT_INFERENCEWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2072 SCIP_CALL( SCIPaddRealParam(scip,
2073 "branching/relpscost/cutoffweight",
2074 "weight in score calculations for cutoff score",
2075 &branchruledata->cutoffweight, TRUE, DEFAULT_CUTOFFWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2076 SCIP_CALL( SCIPaddRealParam(scip,
2077 "branching/relpscost/gmiavgeffweight",
2078 "weight in score calculations for average GMI cuts normalized efficacy",
2079 &branchruledata->gmiavgeffweight, TRUE, DEFAULT_GMIAVGEFFWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2080 SCIP_CALL( SCIPaddRealParam(scip,
2081 "branching/relpscost/gmilasteffweight",
2082 "weight in score calculations for last GMI cuts normalized efficacy",
2083 &branchruledata->gmilasteffweight, TRUE, DEFAULT_GMILASTEFFWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2084 SCIP_CALL( SCIPaddRealParam(scip,
2085 "branching/relpscost/pscostweight",
2086 "weight in score calculations for pseudo cost score",
2087 &branchruledata->pscostweight, TRUE, DEFAULT_PSCOSTWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2088 SCIP_CALL( SCIPaddRealParam(scip,
2089 "branching/relpscost/nlscoreweight",
2090 "weight in score calculations for nlcount score",
2091 &branchruledata->nlscoreweight, TRUE, DEFAULT_NLSCOREWEIGHT, SCIP_REAL_MIN, SCIP_REAL_MAX, NULL, NULL) );
2092 SCIP_CALL( SCIPaddRealParam(scip,
2093 "branching/relpscost/minreliable",
2094 "minimal value for minimum pseudo cost size to regard pseudo cost value as reliable",
2095 &branchruledata->minreliable, TRUE, DEFAULT_MINRELIABLE, 0.0, SCIP_REAL_MAX, NULL, NULL) );
2096 SCIP_CALL( SCIPaddRealParam(scip,
2097 "branching/relpscost/maxreliable",
2098 "maximal value for minimum pseudo cost size to regard pseudo cost value as reliable",
2099 &branchruledata->maxreliable, TRUE, DEFAULT_MAXRELIABLE, 0.0, SCIP_REAL_MAX, NULL, NULL) );
2100 SCIP_CALL( SCIPaddRealParam(scip,
2101 "branching/relpscost/sbiterquot",
2102 "maximal fraction of strong branching LP iterations compared to node relaxation LP iterations",
2103 &branchruledata->sbiterquot, FALSE, DEFAULT_SBITERQUOT, 0.0, SCIP_REAL_MAX, NULL, NULL) );
2104 SCIP_CALL( SCIPaddIntParam(scip,
2105 "branching/relpscost/sbiterofs",
2106 "additional number of allowed strong branching LP iterations",
2107 &branchruledata->sbiterofs, FALSE, DEFAULT_SBITEROFS, 0, INT_MAX, NULL, NULL) );
2108 SCIP_CALL( SCIPaddIntParam(scip,
2109 "branching/relpscost/maxlookahead",
2110 "maximal number of further variables evaluated without better score",
2111 &branchruledata->maxlookahead, TRUE, DEFAULT_MAXLOOKAHEAD, 1, INT_MAX, NULL, NULL) );
2112 SCIP_CALL( SCIPaddIntParam(scip,
2113 "branching/relpscost/initcand",
2114 "maximal number of candidates initialized with strong branching per node",
2115 &branchruledata->initcand, FALSE, DEFAULT_INITCAND, 0, INT_MAX, NULL, NULL) );
2116 SCIP_CALL( SCIPaddIntParam(scip,
2117 "branching/relpscost/inititer",
2118 "iteration limit for strong branching initializations of pseudo cost entries (0: auto)",
2119 &branchruledata->inititer, FALSE, DEFAULT_INITITER, 0, INT_MAX, NULL, NULL) );
2120 SCIP_CALL( SCIPaddIntParam(scip,
2121 "branching/relpscost/maxbdchgs",
2122 "maximal number of bound tightenings before the node is reevaluated (-1: unlimited)",
2123 &branchruledata->maxbdchgs, TRUE, DEFAULT_MAXBDCHGS, -1, INT_MAX, NULL, NULL) );
2124 SCIP_CALL( SCIPaddIntParam(scip,
2125 "branching/relpscost/maxproprounds",
2126 "maximum number of propagation rounds to be performed during strong branching before solving the LP (-1: no limit, -2: parameter settings)",
2127 &branchruledata->maxproprounds, TRUE, DEFAULT_MAXPROPROUNDS, -2, INT_MAX, NULL, NULL) );
2128 SCIP_CALL( SCIPaddBoolParam(scip,
2129 "branching/relpscost/probingbounds",
2130 "should valid bounds be identified in a probing-like fashion during strong branching (only with propagation)?",
2131 &branchruledata->probingbounds, TRUE, DEFAULT_PROBINGBOUNDS, NULL, NULL) );
2132
2133 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/userelerrorreliability",
2134 "should reliability be based on relative errors?", &branchruledata->userelerrorforreliability, TRUE, DEFAULT_USERELERRORFORRELIABILITY,
2135 NULL, NULL) );
2136
2137 SCIP_CALL( SCIPaddRealParam(scip, "branching/relpscost/lowerrortol", "low relative error tolerance for reliability",
2138 &branchruledata->lowerrortol, TRUE, DEFAULT_LOWERRORTOL, 0.0, SCIP_REAL_MAX, NULL, NULL) );
2139
2140 SCIP_CALL( SCIPaddRealParam(scip, "branching/relpscost/higherrortol", "high relative error tolerance for reliability",
2141 &branchruledata->higherrortol, TRUE, DEFAULT_HIGHERRORTOL, 0.0, SCIP_REAL_MAX, NULL, NULL) );
2142
2143/**! [SnippetCodeStyleParenIndent] */
2144
2145 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/storesemiinitcosts",
2146 "should strong branching result be considered for pseudo costs if the other direction was infeasible?",
2147 &branchruledata->storesemiinitcosts, TRUE, DEFAULT_STORESEMIINITCOSTS,
2148 NULL, NULL) );
2149
2150/**! [SnippetCodeStyleParenIndent] */
2151
2152 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/usesblocalinfo",
2153 "should the scoring function use only local cutoff and inference information obtained for strong branching candidates?",
2154 &branchruledata->usesblocalinfo, TRUE, DEFAULT_USESBLOCALINFO,
2155 NULL, NULL) );
2156
2157 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/usehyptestforreliability",
2158 "should the strong branching decision be based on a hypothesis test?",
2159 &branchruledata->usehyptestforreliability, TRUE, DEFAULT_USEHYPTESTFORRELIABILITY,
2160 NULL, NULL) );
2161
2162 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/usedynamicconfidence",
2163 "should the confidence level be adjusted dynamically?",
2164 &branchruledata->usedynamicconfidence, TRUE, DEFAULT_USEDYNAMICCONFIDENCE,
2165 NULL, NULL) );
2166 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/skipbadinitcands",
2167 "should branching rule skip candidates that have a low probability to "
2168 "be better than the best strong-branching or pseudo-candidate?",
2169 &branchruledata->skipbadinitcands, TRUE, DEFAULT_SKIPBADINITCANDS,
2170 NULL, NULL) );
2171 SCIP_CALL( SCIPaddIntParam(scip,
2172 "branching/relpscost/confidencelevel",
2173 "the confidence level for statistical methods, between 0 (Min) and 4 (Max).",
2174 &branchruledata->confidencelevel, TRUE, DEFAULT_CONFIDENCELEVEL, 0, 4, NULL, NULL) );
2175
2176 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/randinitorder",
2177 "should candidates be initialized in randomized order?",
2178 &branchruledata->randinitorder, TRUE, DEFAULT_RANDINITORDER,
2179 NULL, NULL) );
2180
2181 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/usesmallweightsitlim",
2182 "should smaller weights be used for pseudo cost updates after hitting the LP iteration limit?",
2183 &branchruledata->usesmallweightsitlim, TRUE, DEFAULT_USESMALLWEIGHTSITLIM,
2184 NULL, NULL) );
2185
2186 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/dynamicweights",
2187 "should the weights of the branching rule be adjusted dynamically during solving based on objective and infeasible leaf counters?",
2188 &branchruledata->dynamicweights, TRUE, DEFAULT_DYNAMICWEIGHTS,
2189 NULL, NULL) );
2190 SCIP_CALL( SCIPaddIntParam(scip, "branching/relpscost/degeneracyaware",
2191 "should degeneracy be taken into account to update weights and skip strong branching? (0: off, 1: after root, 2: always)",
2192 &branchruledata->degeneracyaware, TRUE, DEFAULT_DEGENERACYAWARE, 0, 2,
2193 NULL, NULL) );
2194 SCIP_CALL( SCIPaddIntParam(scip, "branching/relpscost/startrandseed", "start seed for random number generation",
2195 &branchruledata->startrandseed, TRUE, DEFAULT_STARTRANDSEED, 0, INT_MAX, NULL, NULL) );
2196
2197 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/filtercandssym",
2198 "Use symmetry to filter branching candidates?",
2199 &branchruledata->filtercandssym, TRUE, DEFAULT_FILTERCANDSSYM, NULL, NULL) );
2200
2201 SCIP_CALL( SCIPaddBoolParam(scip, "branching/relpscost/transsympscost",
2202 "Transfer pscost information to symmetric variables?",
2203 &branchruledata->transsympscost, TRUE, DEFAULT_TRANSSYMPSCOST, NULL, NULL) );
2204
2205 /* initialise the Treemodel parameters */
2206 SCIP_CALL( SCIPtreemodelInit(scip, &branchruledata->treemodel) );
2207
2208 return SCIP_OKAY;
2209}
2210
2211/** execution reliability pseudo cost branching with the given branching candidates */
2212SCIP_RETCODE SCIPexecRelpscostBranching(
2213 SCIP* scip, /**< SCIP data structure */
2214 SCIP_VAR** branchcands, /**< branching candidates */
2215 SCIP_Real* branchcandssol, /**< solution value for the branching candidates */
2216 SCIP_Real* branchcandsfrac, /**< fractional part of the branching candidates */
2217 int nbranchcands, /**< number of branching candidates */
2218 SCIP_Bool executebranching, /**< perform a branching step after probing */
2219 SCIP_RESULT* result /**< pointer to the result of the execution */
2220 )
2221{
2222 SCIP_BRANCHRULE* branchrule;
2223
2224 assert(scip != NULL);
2225 assert(result != NULL);
2226
2227 /* find branching rule */
2228 branchrule = SCIPfindBranchrule(scip, BRANCHRULE_NAME);
2229 assert(branchrule != NULL);
2230
2231 /* execute branching rule */
2232 SCIP_CALL( execRelpscost(scip, branchrule, branchcands, branchcandssol, branchcandsfrac, NULL, nbranchcands, executebranching, result) );
2233
2234 return SCIP_OKAY;
2235}
bound
static long bound
Definition: GomoryHuTree.cpp:47
BRANCHRULE_DESC
#define BRANCHRULE_DESC
Definition: branch_relpscost.c:67
DEFAULT_DEGENERACYAWARE
#define DEFAULT_DEGENERACYAWARE
Definition: branch_relpscost.c:107
DEFAULT_USESBLOCALINFO
#define DEFAULT_USESBLOCALINFO
Definition: branch_relpscost.c:98
DEFAULT_SKIPBADINITCANDS
#define DEFAULT_SKIPBADINITCANDS
Definition: branch_relpscost.c:100
DEFAULT_SBITERQUOT
#define DEFAULT_SBITERQUOT
Definition: branch_relpscost.c:82
calcNlscore
static SCIP_Real calcNlscore(SCIP *scip, int *nlcount, int nlcountmax, int probindex)
Definition: branch_relpscost.c:517
DEFAULT_HIGHERRORTOL
#define DEFAULT_HIGHERRORTOL
Definition: branch_relpscost.c:94
BRANCHRULE_PRIORITY
#define BRANCHRULE_PRIORITY
Definition: branch_relpscost.c:68
DEFAULT_PROBINGBOUNDS
#define DEFAULT_PROBINGBOUNDS
Definition: branch_relpscost.c:90
DEFAULT_INITITER
#define DEFAULT_INITITER
Definition: branch_relpscost.c:86
DEFAULT_USESMALLWEIGHTSITLIM
#define DEFAULT_USESMALLWEIGHTSITLIM
Definition: branch_relpscost.c:104
applyBdchgs
static SCIP_RETCODE applyBdchgs(SCIP *scip, SCIP_VAR **vars, int *bdchginds, SCIP_BOUNDTYPE *bdchgtypes, SCIP_Real *bdchgbounds, int nbdchgs, SCIP_RESULT *result)
Definition: branch_relpscost.c:653
DEFAULT_STARTRANDSEED
#define DEFAULT_STARTRANDSEED
Definition: branch_relpscost.c:102
DEFAULT_FILTERCANDSSYM
#define DEFAULT_FILTERCANDSSYM
Definition: branch_relpscost.c:110
DEFAULT_USEDYNAMICCONFIDENCE
#define DEFAULT_USEDYNAMICCONFIDENCE
Definition: branch_relpscost.c:96
DEFAULT_DYNAMICWEIGHTS
#define DEFAULT_DYNAMICWEIGHTS
Definition: branch_relpscost.c:105
DEFAULT_MAXBDCHGS
#define DEFAULT_MAXBDCHGS
Definition: branch_relpscost.c:87
DEFAULT_USEHYPTESTFORRELIABILITY
#define DEFAULT_USEHYPTESTFORRELIABILITY
Definition: branch_relpscost.c:95
DEFAULT_GMIAVGEFFWEIGHT
#define DEFAULT_GMIAVGEFFWEIGHT
Definition: branch_relpscost.c:76
SCIPupdateVarPseudocostSymmetric
static SCIP_RETCODE SCIPupdateVarPseudocostSymmetric(SCIP *scip, SCIP_BRANCHRULEDATA *branchruledata, SCIP_VAR *branchvar, int *branchorbitidx, int branchvaridx, SCIP_Real solvaldelta, SCIP_Real objdelta, SCIP_Real weight)
Definition: branch_relpscost.c:322
DEFAULT_INFERENCEWEIGHT
#define DEFAULT_INFERENCEWEIGHT
Definition: branch_relpscost.c:74
BRANCHRULE_NAME
#define BRANCHRULE_NAME
Definition: branch_relpscost.c:66
DEFAULT_RANDINITORDER
#define DEFAULT_RANDINITORDER
Definition: branch_relpscost.c:103
initOrbits
static SCIP_RETCODE initOrbits(SCIP *scip, SCIP_BRANCHRULEDATA *branchruledata)
Definition: branch_relpscost.c:180
DEFAULT_MAXLOOKAHEAD
#define DEFAULT_MAXLOOKAHEAD
Definition: branch_relpscost.c:84
DEFAULT_SBITEROFS
#define DEFAULT_SBITEROFS
Definition: branch_relpscost.c:83
SCIP_DECL_BRANCHFREE
static SCIP_DECL_BRANCHFREE(branchFreeRelpscost)
Definition: branch_relpscost.c:1877
DEFAULT_NLSCOREWEIGHT
#define DEFAULT_NLSCOREWEIGHT
Definition: branch_relpscost.c:79
DEFAULT_CONFLICTWEIGHT
#define DEFAULT_CONFLICTWEIGHT
Definition: branch_relpscost.c:72
filterSymmetricVariables
static SCIP_RETCODE filterSymmetricVariables(SCIP *scip, SCIP_BRANCHRULEDATA *branchruledata, SCIP_VAR **origbranchcands, SCIP_Real *origbranchcandssol, SCIP_Real *origbranchcandsfrac, int norigbranchcands, SCIP_VAR **branchcands, SCIP_Real *branchcandssol, SCIP_Real *branchcandsfrac, int *branchorbitidx, int *nbranchcands)
Definition: branch_relpscost.c:239
DEFAULT_GMILASTEFFWEIGHT
#define DEFAULT_GMILASTEFFWEIGHT
Definition: branch_relpscost.c:77
DEFAULT_USERELERRORFORRELIABILITY
#define DEFAULT_USERELERRORFORRELIABILITY
Definition: branch_relpscost.c:92
DEFAULT_CUTOFFWEIGHT
#define DEFAULT_CUTOFFWEIGHT
Definition: branch_relpscost.c:75
SCIP_DECL_BRANCHCOPY
static SCIP_DECL_BRANCHCOPY(branchCopyRelpscost)
Definition: branch_relpscost.c:1863
countNonlinearities
static SCIP_RETCODE countNonlinearities(SCIP *scip, int *nlcount, int nlcountsize, int *nlcountmax)
Definition: branch_relpscost.c:384
SCIP_DECL_BRANCHINITSOL
static SCIP_DECL_BRANCHINITSOL(branchInitsolRelpscost)
Definition: branch_relpscost.c:1895
DEFAULT_TRANSSYMPSCOST
#define DEFAULT_TRANSSYMPSCOST
Definition: branch_relpscost.c:111
addBdchg
static SCIP_RETCODE addBdchg(SCIP *scip, int **bdchginds, SCIP_BOUNDTYPE **bdchgtypes, SCIP_Real **bdchgbounds, int *nbdchgs, int ind, SCIP_BOUNDTYPE type, SCIP_Real bound)
Definition: branch_relpscost.c:596
DEFAULT_LOWERRORTOL
#define DEFAULT_LOWERRORTOL
Definition: branch_relpscost.c:93
execRelpscost
static SCIP_RETCODE execRelpscost(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_VAR **branchcands, SCIP_Real *branchcandssol, SCIP_Real *branchcandsfrac, int *branchorbitidx, int nbranchcands, SCIP_Bool executebranch, SCIP_RESULT *result)
Definition: branch_relpscost.c:730
branchruledataEnsureNlcount
static SCIP_RETCODE branchruledataEnsureNlcount(SCIP *scip, SCIP_BRANCHRULEDATA *branchruledata)
Definition: branch_relpscost.c:470
DEFAULT_CONFLENGTHWEIGHT
#define DEFAULT_CONFLENGTHWEIGHT
Definition: branch_relpscost.c:73
DEFAULT_CONFIDENCELEVEL
#define DEFAULT_CONFIDENCELEVEL
Definition: branch_relpscost.c:99
DEFAULT_MINRELIABLE
#define DEFAULT_MINRELIABLE
Definition: branch_relpscost.c:80
calcScore
static SCIP_Real calcScore(SCIP *scip, SCIP_BRANCHRULEDATA *branchruledata, SCIP_Real conflictscore, SCIP_Real avgconflictscore, SCIP_Real conflengthscore, SCIP_Real avgconflengthscore, SCIP_Real inferencescore, SCIP_Real avginferencescore, SCIP_Real cutoffscore, SCIP_Real avgcutoffscore, SCIP_Real gmieffscore, SCIP_Real lastgmieffscore, SCIP_Real pscostscore, SCIP_Real avgpscostscore, SCIP_Real nlscore, SCIP_Real frac, SCIP_Real degeneracyfactor)
Definition: branch_relpscost.c:544
DEFAULT_STORESEMIINITCOSTS
#define DEFAULT_STORESEMIINITCOSTS
Definition: branch_relpscost.c:97
DEFAULT_MAXPROPROUNDS
#define DEFAULT_MAXPROPROUNDS
Definition: branch_relpscost.c:88
SCIP_DECL_BRANCHEXITSOL
static SCIP_DECL_BRANCHEXITSOL(branchExitsolRelpscost)
Definition: branch_relpscost.c:1916
DEFAULT_INITCAND
#define DEFAULT_INITCAND
Definition: branch_relpscost.c:85
BRANCHRULE_MAXDEPTH
#define BRANCHRULE_MAXDEPTH
Definition: branch_relpscost.c:69
DEFAULT_MAXRELIABLE
#define DEFAULT_MAXRELIABLE
Definition: branch_relpscost.c:81
DEFAULT_PSCOSTWEIGHT
#define DEFAULT_PSCOSTWEIGHT
Definition: branch_relpscost.c:78
BRANCHRULE_MAXBOUNDDIST
#define BRANCHRULE_MAXBOUNDDIST
Definition: branch_relpscost.c:70
freeBdchgs
static void freeBdchgs(SCIP *scip, int **bdchginds, SCIP_BOUNDTYPE **bdchgtypes, SCIP_Real **bdchgbounds, int *nbdchgs)
Definition: branch_relpscost.c:630
SCIP_DECL_BRANCHEXECLP
static SCIP_DECL_BRANCHEXECLP(branchExeclpRelpscost)
Definition: branch_relpscost.c:1943
branch_relpscost.h
reliable pseudo costs branching rule
cons_and.h
Constraint handler for AND constraints, .
NULL
#define NULL
Definition: def.h:266
SCIP_Longint
#define SCIP_Longint
Definition: def.h:157
SCIP_UNUSED
#define SCIP_UNUSED(x)
Definition: def.h:427
SCIP_REAL_MAX
#define SCIP_REAL_MAX
Definition: def.h:173
SCIP_INVALID
#define SCIP_INVALID
Definition: def.h:192
SCIP_Bool
#define SCIP_Bool
Definition: def.h:91
MIN
#define MIN(x, y)
Definition: def.h:242
SCIP_Real
#define SCIP_Real
Definition: def.h:172
TRUE
#define TRUE
Definition: def.h:93
FALSE
#define FALSE
Definition: def.h:94
MAX
#define MAX(x, y)
Definition: def.h:238
SCIP_LONGINT_FORMAT
#define SCIP_LONGINT_FORMAT
Definition: def.h:164
SCIP_REAL_MIN
#define SCIP_REAL_MIN
Definition: def.h:174
SCIP_CALL
#define SCIP_CALL(x)
Definition: def.h:373
SCIPexecRelpscostBranching
SCIP_RETCODE SCIPexecRelpscostBranching(SCIP *scip, SCIP_VAR **branchcands, SCIP_Real *branchcandssol, SCIP_Real *branchcandsfrac, int nbranchcands, SCIP_Bool executebranching, SCIP_RESULT *result)
Definition: branch_relpscost.c:2212
SCIPincludeBranchruleRelpscost
SCIP_RETCODE SCIPincludeBranchruleRelpscost(SCIP *scip)
Definition: branch_relpscost.c:2026
SCIPgetResultantAnd
SCIP_VAR * SCIPgetResultantAnd(SCIP *scip, SCIP_CONS *cons)
Definition: cons_and.c:5260
SCIPgetNVarsAnd
int SCIPgetNVarsAnd(SCIP *scip, SCIP_CONS *cons)
Definition: cons_and.c:5211
SCIPgetVarsAnd
SCIP_VAR ** SCIPgetVarsAnd(SCIP *scip, SCIP_CONS *cons)
Definition: cons_and.c:5235
SCIPgetSubscipDepth
int SCIPgetSubscipDepth(SCIP *scip)
Definition: scip_copy.c:2605
SCIPisStopped
SCIP_Bool SCIPisStopped(SCIP *scip)
Definition: scip_general.c:734
SCIPisExactSolve
SCIP_Bool SCIPisExactSolve(SCIP *scip)
Definition: scip_general.c:621
SCIPgetVarsData
SCIP_RETCODE SCIPgetVarsData(SCIP *scip, SCIP_VAR ***vars, int *nvars, int *nbinvars, int *nintvars, int *nimplvars, int *ncontvars)
Definition: scip_prob.c:1866
SCIPgetNVars
int SCIPgetNVars(SCIP *scip)
Definition: scip_prob.c:1992
SCIPgetVars
SCIP_VAR ** SCIPgetVars(SCIP *scip)
Definition: scip_prob.c:1947
SCIPhashmapGetImageInt
int SCIPhashmapGetImageInt(SCIP_HASHMAP *hashmap, void *origin)
Definition: misc.c:3284
SCIPupdateNodeLowerbound
SCIP_RETCODE SCIPupdateNodeLowerbound(SCIP *scip, SCIP_NODE *node, SCIP_Real newbound)
Definition: scip_prob.c:3762
SCIPupdateLocalLowerbound
SCIP_RETCODE SCIPupdateLocalLowerbound(SCIP *scip, SCIP_Real newbound)
Definition: scip_prob.c:3697
SCIPgetLocalLowerbound
SCIP_Real SCIPgetLocalLowerbound(SCIP *scip)
Definition: scip_prob.c:3586
SCIPgetNodeLowerbound
SCIP_Real SCIPgetNodeLowerbound(SCIP *scip, SCIP_NODE *node)
Definition: scip_prob.c:3623
SCIPverbMessage
void SCIPverbMessage(SCIP *scip, SCIP_VERBLEVEL msgverblevel, FILE *file, const char *formatstr,...)
Definition: scip_message.c:225
SCIPdebugMsg
#define SCIPdebugMsg
Definition: scip_message.h:78
SCIPaddIntParam
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
SCIPaddRealParam
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
SCIPgetRealParam
SCIP_RETCODE SCIPgetRealParam(SCIP *scip, const char *name, SCIP_Real *value)
Definition: scip_param.c:307
SCIPaddBoolParam
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
SCIPsetBranchruleExecLp
SCIP_RETCODE SCIPsetBranchruleExecLp(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_DECL_BRANCHEXECLP((*branchexeclp)))
Definition: scip_branch.c:256
SCIPfindBranchrule
SCIP_BRANCHRULE * SCIPfindBranchrule(SCIP *scip, const char *name)
Definition: scip_branch.c:304
SCIPsetBranchruleCopy
SCIP_RETCODE SCIPsetBranchruleCopy(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_DECL_BRANCHCOPY((*branchcopy)))
Definition: scip_branch.c:160
SCIPincludeBranchruleBasic
SCIP_RETCODE SCIPincludeBranchruleBasic(SCIP *scip, SCIP_BRANCHRULE **branchruleptr, const char *name, const char *desc, int priority, int maxdepth, SCIP_Real maxbounddist, SCIP_BRANCHRULEDATA *branchruledata)
Definition: scip_branch.c:123
SCIPbranchruleGetName
const char * SCIPbranchruleGetName(SCIP_BRANCHRULE *branchrule)
Definition: branch.c:1971
SCIPbranchruleGetData
SCIP_BRANCHRULEDATA * SCIPbranchruleGetData(SCIP_BRANCHRULE *branchrule)
Definition: branch.c:1849
SCIPsetBranchruleFree
SCIP_RETCODE SCIPsetBranchruleFree(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_DECL_BRANCHFREE((*branchfree)))
Definition: scip_branch.c:176
SCIPsetBranchruleExitsol
SCIP_RETCODE SCIPsetBranchruleExitsol(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_DECL_BRANCHEXITSOL((*branchexitsol)))
Definition: scip_branch.c:240
SCIPbranchruleSetData
void SCIPbranchruleSetData(SCIP_BRANCHRULE *branchrule, SCIP_BRANCHRULEDATA *branchruledata)
Definition: branch.c:1859
SCIPsetBranchruleInitsol
SCIP_RETCODE SCIPsetBranchruleInitsol(SCIP *scip, SCIP_BRANCHRULE *branchrule, SCIP_DECL_BRANCHINITSOL((*branchinitsol)))
Definition: scip_branch.c:224
SCIPbranchVarVal
SCIP_RETCODE SCIPbranchVarVal(SCIP *scip, SCIP_VAR *var, SCIP_Real val, SCIP_NODE **downchild, SCIP_NODE **eqchild, SCIP_NODE **upchild)
Definition: scip_branch.c:1133
SCIPgetLPBranchCands
SCIP_RETCODE SCIPgetLPBranchCands(SCIP *scip, SCIP_VAR ***lpcands, SCIP_Real **lpcandssol, SCIP_Real **lpcandsfrac, int *nlpcands, int *npriolpcands, int *nfracimplvars)
Definition: scip_branch.c:402
SCIPgetBranchScore
SCIP_Real SCIPgetBranchScore(SCIP *scip, SCIP_VAR *var, SCIP_Real downgain, SCIP_Real upgain)
Definition: scip_branch.c:856
SCIPfindConshdlr
SCIP_CONSHDLR * SCIPfindConshdlr(SCIP *scip, const char *name)
Definition: scip_cons.c:941
SCIPconshdlrGetNActiveConss
int SCIPconshdlrGetNActiveConss(SCIP_CONSHDLR *conshdlr)
Definition: cons.c:4670
SCIPconshdlrGetConss
SCIP_CONS ** SCIPconshdlrGetConss(SCIP_CONSHDLR *conshdlr)
Definition: cons.c:4593
SCIPgetLPDualDegeneracy
SCIP_RETCODE SCIPgetLPDualDegeneracy(SCIP *scip, SCIP_Real *degeneracy, SCIP_Real *varconsratio)
Definition: scip_lp.c:2792
SCIPinDive
SCIP_Bool SCIPinDive(SCIP *scip)
Definition: scip_lp.c:2775
SCIPhasCurrentNodeLP
SCIP_Bool SCIPhasCurrentNodeLP(SCIP *scip)
Definition: scip_lp.c:83
SCIPgetLPSolstat
SCIP_LPSOLSTAT SCIPgetLPSolstat(SCIP *scip)
Definition: scip_lp.c:168
SCIPallColsInLP
SCIP_Bool SCIPallColsInLP(SCIP *scip)
Definition: scip_lp.c:649
SCIPgetLPObjval
SCIP_Real SCIPgetLPObjval(SCIP *scip)
Definition: scip_lp.c:247
SCIPisLPSolBasic
SCIP_Bool SCIPisLPSolBasic(SCIP *scip)
Definition: scip_lp.c:667
SCIPfreeBlockMemoryArray
#define SCIPfreeBlockMemoryArray(scip, ptr, num)
Definition: scip_mem.h:110
SCIPallocBufferArray
#define SCIPallocBufferArray(scip, ptr, num)
Definition: scip_mem.h:124
SCIPreallocBufferArray
#define SCIPreallocBufferArray(scip, ptr, num)
Definition: scip_mem.h:128
SCIPfreeBufferArray
#define SCIPfreeBufferArray(scip, ptr)
Definition: scip_mem.h:136
SCIPduplicateBufferArray
#define SCIPduplicateBufferArray(scip, ptr, source, num)
Definition: scip_mem.h:132
SCIPallocBlockMemoryArray
#define SCIPallocBlockMemoryArray(scip, ptr, num)
Definition: scip_mem.h:93
SCIPreallocBlockMemoryArray
#define SCIPreallocBlockMemoryArray(scip, ptr, oldnum, newnum)
Definition: scip_mem.h:99
SCIPfreeBlockMemory
#define SCIPfreeBlockMemory(scip, ptr)
Definition: scip_mem.h:108
SCIPfreeBlockMemoryArrayNull
#define SCIPfreeBlockMemoryArrayNull(scip, ptr, num)
Definition: scip_mem.h:111
SCIPfreeBufferArrayNull
#define SCIPfreeBufferArrayNull(scip, ptr)
Definition: scip_mem.h:137
SCIPallocBlockMemory
#define SCIPallocBlockMemory(scip, ptr)
Definition: scip_mem.h:89
SCIPisNLPConstructed
SCIP_Bool SCIPisNLPConstructed(SCIP *scip)
Definition: scip_nlp.c:110
SCIPgetNNLPVars
int SCIPgetNNLPVars(SCIP *scip)
Definition: scip_nlp.c:201
SCIPgetNLPVarsNonlinearity
SCIP_RETCODE SCIPgetNLPVarsNonlinearity(SCIP *scip, int *nlcount)
Definition: scip_nlp.c:223
SCIPnodeGetLowerbound
SCIP_Real SCIPnodeGetLowerbound(SCIP_NODE *node)
Definition: tree.c:7531
SCIPnodeGetNumber
SCIP_Longint SCIPnodeGetNumber(SCIP_NODE *node)
Definition: tree.c:7511
SCIPinProbing
SCIP_Bool SCIPinProbing(SCIP *scip)
Definition: scip_probing.c:97
SCIPgetBestSol
SCIP_SOL * SCIPgetBestSol(SCIP *scip)
Definition: scip_sol.c:2165
SCIPsolGetIndex
int SCIPsolGetIndex(SCIP_SOL *sol)
Definition: sol.c:2835
SCIPgetAvgInferenceScore
SCIP_Real SCIPgetAvgInferenceScore(SCIP *scip)
Definition: scip_solvingstats.c:2118
SCIPgetAvgConflictlengthScore
SCIP_Real SCIPgetAvgConflictlengthScore(SCIP *scip)
Definition: scip_solvingstats.c:2032
SCIPgetNInfeasibleLeaves
SCIP_Longint SCIPgetNInfeasibleLeaves(SCIP *scip)
Definition: scip_solvingstats.c:292
SCIPgetNNodes
SCIP_Longint SCIPgetNNodes(SCIP *scip)
Definition: scip_solvingstats.c:211
SCIPgetNNodeInitLPs
SCIP_Longint SCIPgetNNodeInitLPs(SCIP *scip)
Definition: scip_solvingstats.c:803
SCIPgetNStrongbranchLPIterations
SCIP_Longint SCIPgetNStrongbranchLPIterations(SCIP *scip)
Definition: scip_solvingstats.c:893
SCIPgetNNodeLPIterations
SCIP_Longint SCIPgetNNodeLPIterations(SCIP *scip)
Definition: scip_solvingstats.c:785
SCIPgetAvgConflictScore
SCIP_Real SCIPgetAvgConflictScore(SCIP *scip)
Definition: scip_solvingstats.c:1982
SCIPgetNDualResolveLPIterations
SCIP_Longint SCIPgetNDualResolveLPIterations(SCIP *scip)
Definition: scip_solvingstats.c:731
SCIPgetAvgPseudocostScore
SCIP_Real SCIPgetAvgPseudocostScore(SCIP *scip)
Definition: scip_solvingstats.c:1885
SCIPgetCutoffbound
SCIP_Real SCIPgetCutoffbound(SCIP *scip)
Definition: scip_solvingstats.c:1581
SCIPgetNObjlimLeaves
SCIP_Longint SCIPgetNObjlimLeaves(SCIP *scip)
Definition: scip_solvingstats.c:319
SCIPgetNNodeInitLPIterations
SCIP_Longint SCIPgetNNodeInitLPIterations(SCIP *scip)
Definition: scip_solvingstats.c:821
SCIPgetNDualResolveLPs
SCIP_Longint SCIPgetNDualResolveLPs(SCIP *scip)
Definition: scip_solvingstats.c:711
SCIPgetNRootStrongbranchLPIterations
SCIP_Longint SCIPgetNRootStrongbranchLPIterations(SCIP *scip)
Definition: scip_solvingstats.c:929
SCIPgetAvgCutoffScore
SCIP_Real SCIPgetAvgCutoffScore(SCIP *scip)
Definition: scip_solvingstats.c:2204
SCIPcomputeOrbitsComponentsSym
SCIP_RETCODE SCIPcomputeOrbitsComponentsSym(SCIP *scip, int npermvars, int **permstrans, int nperms, int *components, int *componentbegins, int *vartocomponent, int ncomponents, int *orbits, int *orbitbegins, int *norbits, int *varorbitmap)
Definition: symmetry.c:420
SCIPinfinity
SCIP_Real SCIPinfinity(SCIP *scip)
Definition: scip_numerics.c:510
SCIPisGE
SCIP_Bool SCIPisGE(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
Definition: scip_numerics.c:497
SCIPfeasFrac
SCIP_Real SCIPfeasFrac(SCIP *scip, SCIP_Real val)
Definition: scip_numerics.c:941
SCIPfeasCeil
SCIP_Real SCIPfeasCeil(SCIP *scip, SCIP_Real val)
Definition: scip_numerics.c:917
SCIPfeasFloor
SCIP_Real SCIPfeasFloor(SCIP *scip, SCIP_Real val)
Definition: scip_numerics.c:905
SCIPisFeasIntegral
SCIP_Bool SCIPisFeasIntegral(SCIP *scip, SCIP_Real val)
Definition: scip_numerics.c:881
SCIPfeastol
SCIP_Real SCIPfeastol(SCIP *scip)
Definition: scip_numerics.c:165
SCIPisSumGE
SCIP_Bool SCIPisSumGE(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
Definition: scip_numerics.c:731
SCIPisGT
SCIP_Bool SCIPisGT(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
Definition: scip_numerics.c:484
SCIPisZero
SCIP_Bool SCIPisZero(SCIP *scip, SCIP_Real val)
Definition: scip_numerics.c:558
SCIPisLT
SCIP_Bool SCIPisLT(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
Definition: scip_numerics.c:458
SCIPisSumGT
SCIP_Bool SCIPisSumGT(SCIP *scip, SCIP_Real val1, SCIP_Real val2)
Definition: scip_numerics.c:718
SCIPgetDepth
int SCIPgetDepth(SCIP *scip)
Definition: scip_tree.c:672
SCIPgetCurrentNode
SCIP_NODE * SCIPgetCurrentNode(SCIP *scip)
Definition: scip_tree.c:91
SCIPgetVarStrongbranchFrac
SCIP_RETCODE SCIPgetVarStrongbranchFrac(SCIP *scip, SCIP_VAR *var, int itlim, SCIP_Bool idempotent, SCIP_Real *down, SCIP_Real *up, SCIP_Bool *downvalid, SCIP_Bool *upvalid, SCIP_Bool *downinf, SCIP_Bool *upinf, SCIP_Bool *downconflict, SCIP_Bool *upconflict, SCIP_Bool *lperror)
Definition: scip_var.c:2919
SCIPtightenVarLb
SCIP_RETCODE SCIPtightenVarLb(SCIP *scip, SCIP_VAR *var, SCIP_Real newbound, SCIP_Bool force, SCIP_Bool *infeasible, SCIP_Bool *tightened)
Definition: scip_var.c:5326
SCIPpscostThresholdProbabilityTest
SCIP_Bool SCIPpscostThresholdProbabilityTest(SCIP *scip, SCIP_VAR *var, SCIP_Real frac, SCIP_Real threshold, SCIP_BRANCHDIR dir, SCIP_CONFIDENCELEVEL clevel)
Definition: scip_var.c:9170
SCIPvarIsActive
SCIP_Bool SCIPvarIsActive(SCIP_VAR *var)
Definition: var.c:17775
SCIPgetVarPseudocostCountCurrentRun
SCIP_Real SCIPgetVarPseudocostCountCurrentRun(SCIP *scip, SCIP_VAR *var, SCIP_BRANCHDIR dir)
Definition: scip_var.c:9061
SCIPgetVarAvgInferenceScore
SCIP_Real SCIPgetVarAvgInferenceScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:9584
SCIPendStrongbranch
SCIP_RETCODE SCIPendStrongbranch(SCIP *scip)
Definition: scip_var.c:2744
SCIPgetVarPseudocostCurrentRun
SCIP_Real SCIPgetVarPseudocostCurrentRun(SCIP *scip, SCIP_VAR *var, SCIP_BRANCHDIR dir)
Definition: scip_var.c:9007
SCIPvarGetUbLocal
SCIP_Real SCIPvarGetUbLocal(SCIP_VAR *var)
Definition: var.c:18171
SCIPisVarPscostRelerrorReliable
SCIP_Bool SCIPisVarPscostRelerrorReliable(SCIP *scip, SCIP_VAR *var, SCIP_Real threshold, SCIP_CONFIDENCELEVEL clevel)
Definition: scip_var.c:9189
SCIPvarGetProbvar
SCIP_VAR * SCIPvarGetProbvar(SCIP_VAR *var)
Definition: var.c:12245
SCIPtightenVarUb
SCIP_RETCODE SCIPtightenVarUb(SCIP *scip, SCIP_VAR *var, SCIP_Real newbound, SCIP_Bool force, SCIP_Bool *infeasible, SCIP_Bool *tightened)
Definition: scip_var.c:5443
SCIPvarGetProbindex
int SCIPvarGetProbindex(SCIP_VAR *var)
Definition: var.c:17795
SCIPvarGetName
const char * SCIPvarGetName(SCIP_VAR *var)
Definition: var.c:17446
SCIPgetVarStrongbranchNode
SCIP_Longint SCIPgetVarStrongbranchNode(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:4283
SCIPgetLastStrongbranchLPSolStat
SCIP_LPSOLSTAT SCIPgetLastStrongbranchLPSolStat(SCIP *scip, SCIP_BRANCHDIR branchdir)
Definition: scip_var.c:4111
SCIPgetVarPseudocostVal
SCIP_Real SCIPgetVarPseudocostVal(SCIP *scip, SCIP_VAR *var, SCIP_Real solvaldelta)
Definition: scip_var.c:8925
SCIPgetVarConflictlengthScore
SCIP_Real SCIPgetVarConflictlengthScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:9414
SCIPsignificantVarPscostDifference
SCIP_Bool SCIPsignificantVarPscostDifference(SCIP *scip, SCIP_VAR *varx, SCIP_Real fracx, SCIP_VAR *vary, SCIP_Real fracy, SCIP_BRANCHDIR dir, SCIP_CONFIDENCELEVEL clevel, SCIP_Bool onesided)
Definition: scip_var.c:9140
SCIPgetVarStrongbranchWithPropagation
SCIP_RETCODE SCIPgetVarStrongbranchWithPropagation(SCIP *scip, SCIP_VAR *var, SCIP_Real solval, SCIP_Real lpobjval, int itlim, int maxproprounds, SCIP_Real *down, SCIP_Real *up, SCIP_Bool *downvalid, SCIP_Bool *upvalid, SCIP_Longint *ndomredsdown, SCIP_Longint *ndomredsup, SCIP_Bool *downinf, SCIP_Bool *upinf, SCIP_Bool *downconflict, SCIP_Bool *upconflict, SCIP_Bool *lperror, SCIP_Real *newlbs, SCIP_Real *newubs)
Definition: scip_var.c:3396
SCIPvarGetLPSol
SCIP_Real SCIPvarGetLPSol(SCIP_VAR *var)
Definition: var.c:18479
SCIPgetVarPseudocostScoreCurrentRun
SCIP_Real SCIPgetVarPseudocostScoreCurrentRun(SCIP *scip, SCIP_VAR *var, SCIP_Real solval)
Definition: scip_var.c:9252
SCIPvarGetLbLocal
SCIP_Real SCIPvarGetLbLocal(SCIP_VAR *var)
Definition: var.c:18161
SCIPupdateVarPseudocost
SCIP_RETCODE SCIPupdateVarPseudocost(SCIP *scip, SCIP_VAR *var, SCIP_Real solvaldelta, SCIP_Real objdelta, SCIP_Real weight)
Definition: scip_var.c:8891
SCIPgetVarAvgGMIScore
SCIP_Real SCIPgetVarAvgGMIScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:9988
SCIPgetVarLastGMIScore
SCIP_Real SCIPgetVarLastGMIScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:10043
SCIPgetVarConflictScore
SCIP_Real SCIPgetVarConflictScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:9352
SCIPgetVarAvgCutoffScore
SCIP_Real SCIPgetVarAvgCutoffScore(SCIP *scip, SCIP_VAR *var)
Definition: scip_var.c:9838
SCIPgetVarStrongbranchLast
SCIP_RETCODE SCIPgetVarStrongbranchLast(SCIP *scip, SCIP_VAR *var, SCIP_Real *down, SCIP_Real *up, SCIP_Bool *downvalid, SCIP_Bool *upvalid, SCIP_Real *solval, SCIP_Real *lpobjval)
Definition: scip_var.c:4133
SCIPgetVarPseudocostScore
SCIP_Real SCIPgetVarPseudocostScore(SCIP *scip, SCIP_VAR *var, SCIP_Real solval)
Definition: scip_var.c:9214
SCIPstartStrongbranch
SCIP_RETCODE SCIPstartStrongbranch(SCIP *scip, SCIP_Bool enablepropagation)
Definition: scip_var.c:2686
SCIPfreeRandom
void SCIPfreeRandom(SCIP *scip, SCIP_RANDNUMGEN **randnumgen)
Definition: scip_randnumgen.c:79
SCIPrandomGetReal
SCIP_Real SCIPrandomGetReal(SCIP_RANDNUMGEN *randnumgen, SCIP_Real minrandval, SCIP_Real maxrandval)
Definition: misc.c:10133
SCIPcreateRandom
SCIP_RETCODE SCIPcreateRandom(SCIP *scip, SCIP_RANDNUMGEN **randnumgen, unsigned int initialseed, SCIP_Bool useglobalseed)
Definition: scip_randnumgen.c:56
memory.h
memory allocation routines
BMSclearMemoryArray
#define BMSclearMemoryArray(ptr, num)
Definition: memory.h:130
SCIPgetSymmetry
SCIP_RETCODE SCIPgetSymmetry(SCIP *scip, int *npermvars, SCIP_VAR ***permvars, SCIP_HASHMAP **permvarmap, int *nperms, int ***perms, int ***permstrans, SCIP_Real *log10groupsize, SCIP_Bool *binvaraffected, int **components, int **componentbegins, int **vartocomponent, int *ncomponents)
Definition: prop_symmetry.c:7015
prop_symmetry.h
propagator for symmetry handling
pub_branch.h
public methods for branching rules
pub_cons.h
public methods for managing constraints
pub_message.h
public methods for message output
SCIPdebug
#define SCIPdebug(x)
Definition: pub_message.h:93
pub_misc.h
public data structures and miscellaneous methods
pub_sol.h
public methods for primal CIP solutions
pub_tree.h
public methods for branch and bound tree
pub_var.h
public methods for problem variables
scip_branch.h
public methods for branching rule plugins and branching
scip_cons.h
public methods for constraint handler plugins and constraints
scip_general.h
general public methods
scip_lp.h
public methods for the LP relaxation, rows and columns
scip_mem.h
public methods for memory management
scip_message.h
public methods for message handling
scip_nlp.h
public methods for nonlinear relaxation
scip_numerics.h
public methods for numerical tolerances
scip_param.h
public methods for SCIP parameter handling
scip_prob.h
public methods for global and local (sub)problems
scip_randnumgen.h
public methods for random numbers
scip_sol.h
public methods for solutions
scip_solvingstats.h
public methods for querying solving statistics
scip_tree.h
public methods for the branch-and-bound tree
scip_var.h
public methods for SCIP variables
SCIP_Branchrule
Definition: struct_branch.h:79
SCIP_Cons
Definition: struct_cons.h:47
SCIP_Conshdlr
Definition: struct_cons.h:127
SCIP_HashMap
Definition: struct_misc.h:138
SCIP_Node
Definition: struct_tree.h:142
SCIP_RandNumGen
Definition: struct_misc.h:269
SCIP_Treemodel
Definition: treemodel.c:94
SCIP_Var
Definition: struct_var.h:208
symmetry.h
methods for handling symmetries
SCIPtreemodelSelectCandidate
SCIP_RETCODE SCIPtreemodelSelectCandidate(SCIP *scip, SCIP_TREEMODEL *treemodel, SCIP_VAR **branchcands, SCIP_Real *mingains, SCIP_Real *maxgains, SCIP_Real *tiebreakerscore, int nbranchcands, int *bestcand)
Definition: treemodel.c:912
SCIPtreemodelInit
SCIP_RETCODE SCIPtreemodelInit(SCIP *scip, SCIP_TREEMODEL **treemodel)
Definition: treemodel.c:826
SCIPtreemodelFree
SCIP_RETCODE SCIPtreemodelFree(SCIP *scip, SCIP_TREEMODEL **treemodel)
Definition: treemodel.c:884
SCIPtreemodelIsEnabled
SCIP_Bool SCIPtreemodelIsEnabled(SCIP *scip, SCIP_TREEMODEL *treemodel)
Definition: treemodel.c:900
treemodel.h
Branching rules based on the Single-Variable-Branching (SVB) model.
SCIP_BRANCHRULEDATA
struct SCIP_BranchruleData SCIP_BRANCHRULEDATA
Definition: type_branch.h:57
SCIP_BRANCHDIR_DOWNWARDS
@ SCIP_BRANCHDIR_DOWNWARDS
Definition: type_history.h:43
SCIP_BRANCHDIR_UPWARDS
@ SCIP_BRANCHDIR_UPWARDS
Definition: type_history.h:44
SCIP_BOUNDTYPE_UPPER
@ SCIP_BOUNDTYPE_UPPER
Definition: type_lp.h:57
SCIP_BOUNDTYPE_LOWER
@ SCIP_BOUNDTYPE_LOWER
Definition: type_lp.h:56
SCIP_BOUNDTYPE
enum SCIP_BoundType SCIP_BOUNDTYPE
Definition: type_lp.h:59
SCIP_LPSOLSTAT_OPTIMAL
@ SCIP_LPSOLSTAT_OPTIMAL
Definition: type_lp.h:43
SCIP_LPSOLSTAT_INFEASIBLE
@ SCIP_LPSOLSTAT_INFEASIBLE
Definition: type_lp.h:44
SCIP_LPSOLSTAT_OBJLIMIT
@ SCIP_LPSOLSTAT_OBJLIMIT
Definition: type_lp.h:46
SCIP_LPSOLSTAT_ITERLIMIT
@ SCIP_LPSOLSTAT_ITERLIMIT
Definition: type_lp.h:47
SCIP_VERBLEVEL_HIGH
@ SCIP_VERBLEVEL_HIGH
Definition: type_message.h:56
SCIP_CONFIDENCELEVEL_MAX
@ SCIP_CONFIDENCELEVEL_MAX
Definition: type_misc.h:51
SCIP_CONFIDENCELEVEL_MEDIUM
@ SCIP_CONFIDENCELEVEL_MEDIUM
Definition: type_misc.h:49
SCIP_CONFIDENCELEVEL_HIGH
@ SCIP_CONFIDENCELEVEL_HIGH
Definition: type_misc.h:50
SCIP_CONFIDENCELEVEL_MIN
@ SCIP_CONFIDENCELEVEL_MIN
Definition: type_misc.h:47
SCIP_CONFIDENCELEVEL_LOW
@ SCIP_CONFIDENCELEVEL_LOW
Definition: type_misc.h:48
SCIP_CONFIDENCELEVEL
enum SCIP_Confidencelevel SCIP_CONFIDENCELEVEL
Definition: type_misc.h:53
SCIP_DIDNOTRUN
@ SCIP_DIDNOTRUN
Definition: type_result.h:42
SCIP_CUTOFF
@ SCIP_CUTOFF
Definition: type_result.h:48
SCIP_REDUCEDDOM
@ SCIP_REDUCEDDOM
Definition: type_result.h:51
SCIP_CONSADDED
@ SCIP_CONSADDED
Definition: type_result.h:52
SCIP_BRANCHED
@ SCIP_BRANCHED
Definition: type_result.h:54
SCIP_RESULT
enum SCIP_Result SCIP_RESULT
Definition: type_result.h:61
SCIP_OKAY
@ SCIP_OKAY
Definition: type_retcode.h:42
SCIP_RETCODE
enum SCIP_Retcode SCIP_RETCODE
Definition: type_retcode.h:63