1 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ 2 /* */ 3 /* This file is part of the program and library */ 4 /* SCIP --- Solving Constraint Integer Programs */ 5 /* */ 6 /* Copyright (C) 2002-2021 Konrad-Zuse-Zentrum */ 7 /* fuer Informationstechnik Berlin */ 8 /* */ 9 /* SCIP is distributed under the terms of the ZIB Academic License. */ 10 /* */ 11 /* You should have received a copy of the ZIB Academic License */ 12 /* along with SCIP; see the file COPYING. If not visit scipopt.org. */ 13 /* */ 14 /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ 15 16 /**@file type_benders.h 17 * @ingroup TYPEDEFINITIONS 18 * @brief type definitions for Benders' decomposition methods 19 * @author Stephen J. Maher 20 */ 21 22 /*---+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8----+----9----+----0----+----1----+----2*/ 23 24 #ifndef __SCIP_TYPE_BENDERS_H__ 25 #define __SCIP_TYPE_BENDERS_H__ 26 27 #include "scip/def.h" 28 #include "scip/type_retcode.h" 29 #include "scip/type_scip.h" 30 31 #ifdef __cplusplus 32 extern "C" { 33 #endif 34 35 enum SCIP_BendersEnfoType 36 { 37 SCIP_BENDERSENFOTYPE_LP = 1, /**< the Benders' subproblems are solved during the enforcement of an LP solution */ 38 SCIP_BENDERSENFOTYPE_RELAX = 2, /**< the Benders' subproblems are solved during the enforcement of a relaxation solution */ 39 SCIP_BENDERSENFOTYPE_PSEUDO = 3, /**< the Benders' subproblems are solved during the enforcement of a pseudo solution */ 40 SCIP_BENDERSENFOTYPE_CHECK = 4 /**< the Benders' subproblems are solved during the checking of a solution for feasibility */ 41 }; 42 typedef enum SCIP_BendersEnfoType SCIP_BENDERSENFOTYPE; /**< indicates the callback in cons_benders and cons_benderslp that triggered the subproblem solve */ 43 44 enum SCIP_BendersSolveLoop 45 { 46 SCIP_BENDERSSOLVELOOP_CONVEX = 0, /**< the relaxation is solved in this iteration of the loop */ 47 SCIP_BENDERSSOLVELOOP_CIP = 1, /**< the CIP is solved in this iteration of the loop */ 48 SCIP_BENDERSSOLVELOOP_USERCONVEX = 2, /**< the user defined solve function is called */ 49 SCIP_BENDERSSOLVELOOP_USERCIP = 3 /**< the user defined solve function is called */ 50 }; 51 typedef enum SCIP_BendersSolveLoop SCIP_BENDERSSOLVELOOP; /**< identifies the type of problem solved in each solve loop */ 52 53 enum SCIP_BendersSubStatus 54 { 55 SCIP_BENDERSSUBSTATUS_UNKNOWN = 0, /**< the subsystem status is unknown */ 56 SCIP_BENDERSSUBSTATUS_OPTIMAL = 1, /**< the subsystem is solved to be optimal */ 57 SCIP_BENDERSSUBSTATUS_AUXVIOL = 2, /**< the subproblem is optimal, but the auxiliary variable is violated */ 58 SCIP_BENDERSSUBSTATUS_INFEAS = 3 /**< the subproblem is solved to be infeasible */ 59 }; 60 typedef enum SCIP_BendersSubStatus SCIP_BENDERSSUBSTATUS; 61 62 enum SCIP_BendersSubType 63 { 64 SCIP_BENDERSSUBTYPE_CONVEXCONT = 0, /**< the subproblem has convex constraints and continuous variables */ 65 SCIP_BENDERSSUBTYPE_CONVEXDIS = 1, /**< the subproblem has convex constraints and discrete variables */ 66 SCIP_BENDERSSUBTYPE_NONCONVEXCONT = 2, /**< the subproblem has non-convex constraints and continuous variables */ 67 SCIP_BENDERSSUBTYPE_NONCONVEXDIS = 3, /**< the subproblem has non-convex constraints and discrete variables */ 68 SCIP_BENDERSSUBTYPE_UNKNOWN = 4, /**< the default type before the type is known */ 69 }; 70 typedef enum SCIP_BendersSubType SCIP_BENDERSSUBTYPE; 71 72 typedef struct SCIP_Benders SCIP_BENDERS; /**< Benders' decomposition data */ 73 typedef struct SCIP_BendersData SCIP_BENDERSDATA; /**< locally defined Benders' decomposition data */ 74 typedef struct SCIP_SubproblemSolveStat SCIP_SUBPROBLEMSOLVESTAT; /**< the solving statistics of the subproblems */ 75 76 77 /** copy method for Benders' decomposition plugins (called when SCIP copies plugins). If there is an active Benders' 78 * decomposition, all copies are not valid. As such, there is no valid parameter that is passed to the callback 79 * function 80 * 81 * input: 82 * - scip : SCIP main data structure 83 * - benders : the Benders' decomposition itself 84 * - threadsafe : must the Benders' decomposition copy be thread safe 85 */ 86 #define SCIP_DECL_BENDERSCOPY(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_Bool threadsafe) 87 88 /** destructor of Benders' decomposition to free user data (called when SCIP is exiting) 89 * 90 * input: 91 * - scip : SCIP main data structure 92 * - benders : the Benders' decomposition itself 93 */ 94 #define SCIP_DECL_BENDERSFREE(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 95 96 /** initialization method of Benders' decomposition (called after problem was transformed and the Benders' decomposition 97 * is active) 98 * 99 * input: 100 * - scip : SCIP main data structure 101 * - benders : the Benders' decomposition itself 102 */ 103 #define SCIP_DECL_BENDERSINIT(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 104 105 /** deinitialization method of Benders' decomposition (called before transformed problem is freed and the Benders' 106 * decomposition is active) 107 * 108 * input: 109 * - scip : SCIP main data structure 110 * - benders : the Benders' decomposition itself 111 */ 112 #define SCIP_DECL_BENDERSEXIT(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 113 114 /** presolving initialization method of the Benders' decomposition (called when presolving is about to begin) 115 * 116 * This function is called immediately after the auxiliary variables are created in the master problem. The callback 117 * provides the user an opportunity to add variable data to the auxiliary variables. 118 * 119 * input: 120 * - scip : SCIP main data structure 121 * - benders : the Benders' decomposition itself 122 */ 123 #define SCIP_DECL_BENDERSINITPRE(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 124 125 /** presolving deinitialization method of the Benders' decomposition (called after presolving has been finished) 126 * 127 * input: 128 * - scip : SCIP main data structure 129 * - benders : the Benders' decomposition itself 130 */ 131 #define SCIP_DECL_BENDERSEXITPRE(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 132 133 /** solving process initialization method of Benders' decomposition (called when branch and bound process is about to begin) 134 * 135 * This method is called when the presolving was finished and the branch and bound process is about to begin. 136 * The Benders' decomposition may use this call to initialize its branch and bound specific data. 137 * 138 * input: 139 * - scip : SCIP main data structure 140 * - benders : the Benders' decomposition itself 141 */ 142 #define SCIP_DECL_BENDERSINITSOL(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 143 144 /** solving process deinitialization method of Benders' decomposition (called before branch and bound process data is freed) 145 * 146 * This method is called before the branch and bound process is freed. 147 * The Benders' decomposition should use this call to clean up its branch and bound data. 148 * 149 * input: 150 * - scip : SCIP main data structure 151 * - benders : the Benders' decomposition itself 152 */ 153 #define SCIP_DECL_BENDERSEXITSOL(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders) 154 155 /** the method for creating the Benders' decomposition subproblem. This method is called during the initialisation stage 156 * (after the master problem was transformed). 157 * 158 * @note When the create subproblem callback is invoked, the mapping between the master problem and subproblem 159 * variables must be available. The create subproblem callback is invoked immediately after BENDERSINIT. So, it is 160 * possible to construct the variable mapping within the BENDERSINIT callback. 161 * 162 * This method must register the SCIP instance for the subproblem with the Benders' decomposition core by calling 163 * SCIPaddBendersSubproblem. Typically, the user must create the SCIP instances for the subproblems. These can be 164 * created within a reader or probdata and then registered with the Benders' decomposition core during the call of this 165 * callback. If there are any settings required for solving the subproblems, then they should be set here. However, 166 * some settings will be overridden by the standard solving method included in the Benders' decomposition framework. 167 * If a special solving method is desired, the user can implement the bendersSolvesubXyz callback. In this latter case, 168 * it is possible to provide a NULL pointer to SCIPaddBendersSubproblem. This will ensure that no internal solving 169 * methods available within the Benders' decomposition core are invoked during the solving process. 170 * 171 * If the user defines a subproblem solving method, then in BENDERSCREATESUB, the user must specify whether the 172 * subproblem is convex. This is necessary because the dual solutions from convex problems can be used to generate cuts. 173 * The classical Benders' optimality and feasibility cuts require that the subproblems are convex. If the subproblem is 174 * convex, then the user must call SCIPbendersSetSubproblemIsConvex() 175 * 176 * If the user does NOT implement a subproblem solving method, then the convexity of the problem is determined 177 * internally. 178 * 179 * input: 180 * - scip : SCIP main data structure 181 * - benders : the Benders' decomposition data structure 182 * - probnumber : the subproblem problem number 183 */ 184 #define SCIP_DECL_BENDERSCREATESUB(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, int probnumber) 185 186 /** called before the subproblem solving loop for Benders' decomposition. The pre subproblem solve function gives the 187 * user an oppportunity to perform any global set up for the Benders' decomposition. 188 * 189 * input: 190 * - scip : SCIP main data structure 191 * - benders : the Benders' decomposition data structure 192 * - sol : the solution that will be checked in the subproblem. Can be NULL. 193 * - type : the enforcement type that called the Benders' decomposition solve. 194 * - checkint : should the integer subproblems be checked. 195 * - infeasible : flag to return whether the master problem in infeasible with respect to the added cuts 196 * - auxviol : set to TRUE only if the solution is feasible but the aux vars are violated 197 * - skipsolve : flag to return whether the current subproblem solving loop should be skipped 198 * - result : a result to be returned to the Benders' constraint handler if the solve is skipped. If the 199 * solve is not skipped, then the returned result is ignored. 200 * 201 * possible return values for *result (if more than one applies, the first in the list should be used): 202 * - SCIP_DIDNOTRUN : the subproblem was not solved in this iteration. Other decompositions will be checked. 203 * - SCIP_CONSADDED : a constraint has been added to the master problem. No other decompositions will be checked. 204 * - SCIP_SEPARATED : a cut has been added to the master problem. No other decompositions will be checked. 205 * - SCIP_FEASIBLE : feasibility of the solution is reported to SCIP. Other decompositions will be checked. 206 * - SCIP_INFEASIBLE : infeasibility of the solution is reported to SCIP. No other decompositions will be checked. 207 */ 208 #define SCIP_DECL_BENDERSPRESUBSOLVE(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_SOL* sol,\ 209 SCIP_BENDERSENFOTYPE type, SCIP_Bool checkint, SCIP_Bool* infeasible, SCIP_Bool* auxviol, SCIP_Bool* skipsolve,\ 210 SCIP_RESULT* result) 211 212 /** the solving method for a convex Benders' decomposition subproblem. This call back is provided to solve problems 213 * for which the dual soluitons are used to generate Benders' decomposition cuts. In the classical Benders' 214 * decomposition implementation, this would be an LP. However, it can be any convex problem where the dual solutions 215 * are given by a single vector of reals. 216 * 217 * In the Benders' decomposition subproblem solving process, there are two solving loops. The first is where the convex 218 * subproblems, and the convex relaxations of subproblems, are solved. If no cuts are generated after this solving 219 * loop, then the second loop solves subproblems defined as CIPs. This callback is executed during the FIRST solving 220 * loop only. 221 * 222 * In the classical Benders' decomposition implementation, if the subproblems are all LPs the only the 223 * BENDERSSOLVESUBCONVEX need to be implemented. If the subproblems are MIPs, then it is useful to only implement a 224 * single SCIP instance for the subproblem and then change the variable types of the appropriate variables to 225 * CONTINUOUS for the CONVEX subproblem solve and to INTEGER for the CIP subproblem solve. 226 * 227 * The solving methods are separated so that they can be called in parallel. 228 * 229 * NOTE: The solving methods must be thread safe. 230 * 231 * This method is called from within the execution method. 232 * 233 * input: 234 * - scip : SCIP main data structure 235 * - benders : the Benders' decomposition data structure 236 * - sol : the solution that will be checked in the subproblem. Can be NULL. 237 * - probnumber : the subproblem problem number 238 * - onlyconvexcheck : flag to indicate that only the convex relaxations will be checked in this solving loop. This is 239 * a feature of the Large Neighbourhood Benders' Search 240 * - objective : variable to return the objective function value of the subproblem 241 * - result : the result from solving the subproblem 242 * 243 * possible return values for *result (if more than one applies, the first in the list should be used): 244 * - SCIP_DIDNOTRUN : the subproblem was not solved in this iteration 245 * - SCIP_FEASIBLE : the subproblem is solved and is feasible 246 * - SCIP_INFEASIBLE : the subproblem is solved and is infeasible 247 * - SCIP_UNBOUNDED : the subproblem is solved and is unbounded 248 */ 249 #define SCIP_DECL_BENDERSSOLVESUBCONVEX(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_SOL* sol,\ 250 int probnumber, SCIP_Bool onlyconvexcheck, SCIP_Real* objective, SCIP_RESULT* result) 251 252 /** the solving method for a Benders' decomposition subproblem as a CIP. This call back is provided to solve problems 253 * for which the dual solutions are not well defined. In this case, the cuts are typically generated from the primal 254 * solution to the CIP. In the classical Benders' decomposition implementation, this would be a MIP. However, it can 255 * be any CIP. 256 * 257 * In the Benders' decomposition subproblem solving process, there are two solving loops. The first is where the convex 258 * subproblems, and the convex relaxations of subproblems, are solved. If no cuts are generated after this solving 259 * loop, then the second loop solves subproblems defined as CIPs. This callback is executed during the SECOND solving 260 * loop only. 261 * 262 * The solving methods are separated so that they can be called in parallel. 263 * 264 * NOTE: The solving methods must be thread safe. 265 * 266 * This method is called from within the execution method. 267 * 268 * input: 269 * - scip : SCIP main data structure 270 * - benders : the Benders' decomposition data structure 271 * - sol : the solution that will be checked in the subproblem. Can be NULL. 272 * - probnumber : the subproblem problem number 273 * - objective : variable to return the objective function value of the subproblem 274 * - result : the result from solving the subproblem 275 * 276 * possible return values for *result (if more than one applies, the first in the list should be used): 277 * - SCIP_DIDNOTRUN : the subproblem was not solved in this iteration 278 * - SCIP_FEASIBLE : the subproblem is solved and is feasible 279 * - SCIP_INFEASIBLE : the subproblem is solved and is infeasible 280 * - SCIP_UNBOUNDED : the subproblem is solved and is unbounded 281 */ 282 #define SCIP_DECL_BENDERSSOLVESUB(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_SOL* sol, int probnumber,\ 283 SCIP_Real* objective, SCIP_RESULT* result) 284 285 /** the post-solve method for Benders' decomposition. The post-solve method is called after the subproblems have 286 * been solved but before they have been freed. After the solving of the Benders' decomposition subproblems, the 287 * subproblem solving data is freed in the SCIP_DECL_BENDERSFREESUB callback. However, it is not necessary to implement 288 * SCIP_DECL_BENDERSFREESUB. 289 * 290 * If SCIP_DECL_BENDERSFREESUB is not implemented, then the Benders' decomposition framework will perform a default 291 * freeing of the subproblems. If a subproblem is an LP, then they will be in probing mode for the subproblem 292 * solve. So the freeing process involves ending the probing mode. If the subproblem is a MIP, then the subproblem is 293 * solved by calling SCIPsolve. As such, the transformed problem must be freed after each subproblem solve. 294 * 295 * This callback provides the opportunity for the user to clean up any data structures that should not exist beyond the current 296 * iteration. 297 * The user has full access to the master and subproblems in this callback. So it is possible to construct solution for 298 * the master problem in the method. 299 * Additionally, if there are any subproblems that are infeasibility and this can not be resolved, then the it is 300 * possible to merge these subproblems into the master problem. The subproblem indices are given in the mergecands 301 * array. The merging can be perform by a user defined function or by calling SCIPmergeBendersSubproblemIntoMaster. If a 302 * subproblem was merged into the master problem, then the merged flag must be set to TRUE. 303 * 304 * input: 305 * - scip : SCIP main data structure 306 * - benders : the Benders' decomposition data structure 307 * - sol : the solution that was checked by solving the subproblems. Can be NULL. 308 * - type : the enforcement type that called the Benders' decomposition solve. 309 * - mergecands : the subproblems that are candidates for merging into the master problem, the first 310 * npriomergecands are the priority candidates (they should be merged). The remaining 311 * (nmergecands - npriomergecands) are subproblems that could be merged if desired. 312 * - npriomergecands : the number of priority merge candidates. 313 * - nmergecands : the total number of subproblems that are candidates for merging into the master problem 314 * - checkint : should the integer subproblems be checked. 315 * - infeasible : indicates whether at least one subproblem is infeasible 316 * - merged : flag to indicate whether a subproblem was merged into the master problem. 317 */ 318 #define SCIP_DECL_BENDERSPOSTSOLVE(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_SOL* sol,\ 319 SCIP_BENDERSENFOTYPE type, int* mergecands, int npriomergecands, int nmergecands, SCIP_Bool checkint,\ 320 SCIP_Bool infeasible, SCIP_Bool* merged) 321 322 /** frees the subproblem so that it can be resolved in the next iteration. As stated above, it is not necessary to 323 * implement this callback. If the callback is implemented, the subproblems should be freed by calling 324 * SCIPfreeTransform(). However, if the subproblems are LPs, then it could be more efficient to put the subproblem 325 * into probing mode prior to solving and then exiting the probing mode during the callback. To put the subproblem into 326 * probing mode, the subproblem must be in SCIP_STAGE_SOLVING. This can be achieved by using eventhandlers. 327 * 328 * If SCIP_DECL_BENDERSFREESUB is not implemented, then the Benders' decomposition framework will perform a default 329 * freeing of the subproblems. If a subproblem is an LP, then they will be in probing mode for the subproblem 330 * solve. So the freeing process involves ending the probing mode. If the subproblem is a MIP, then the subproblem is 331 * solved by calling SCIPsolve. As such, the transformed problem must be freed after each subproblem solve. 332 * 333 * NOTE: The freeing methods must be thread safe. 334 * 335 * input: 336 * - scip : SCIP main data structure 337 * - benders : the Benders' decomposition data structure 338 * - probnumber : the subproblem problem number 339 */ 340 #define SCIP_DECL_BENDERSFREESUB(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, int probnumber) 341 342 /** the variable mapping from the subproblem to the master problem. It is neccessary to have a mapping between every 343 * master problem variable and its counterpart in the subproblem. This mapping must go both ways: from master to sub 344 * and sub to master. 345 * 346 * This method is called when generating the cuts. The cuts are generated by using the solution to the subproblem to 347 * eliminate a solution to the master problem. 348 * 349 * input: 350 * - scip : SCIP main data structure 351 * - benders : the Benders' decomposition structure 352 * - var : the variable for which the corresponding variable in the master or subproblem is required 353 * - mappedvar : pointer to store the variable that is mapped to var 354 * - probnumber : the number of the subproblem that the desired variable belongs to, -1 for the master problem 355 */ 356 #define SCIP_DECL_BENDERSGETVAR(x) SCIP_RETCODE x (SCIP* scip, SCIP_BENDERS* benders, SCIP_VAR* var,\ 357 SCIP_VAR** mappedvar, int probnumber) 358 359 #ifdef __cplusplus 360 } 361 #endif 362 363 #endif 364