Implementing Constraint Handlers in SCIP
|
|
- Ruby Miller
- 6 years ago
- Views:
Transcription
1 Implementing Constraint Handlers in SCIP Gregor Hendel Ambros Gleixner, Felipe Serrano September 30
2 Outline Why use constraints? Motivation Callback: Default Constraint Handlers of SCIP Implementation hints for the Nosubtour Constraint Handler Hendel,Gleixner,Serrano Constraint Handler Implementation 1 / 43
3 Outline Why use constraints? Motivation Callback: Default Constraint Handlers of SCIP Implementation hints for the Nosubtour Constraint Handler Hendel,Gleixner,Serrano Constraint Handler Implementation 2 / 43
4 Traveling Salesman Problem (TSP) Denition Given a graph G = (V, E) with edge lengths c e. Find a Hamiltonian cycle (cycle containing all nodes, tour) of minimum length. K 8 Hendel,Gleixner,Serrano Constraint Handler Implementation 3 / 43
5 Traveling Salesman Problem (TSP) Denition Given a graph G = (V, E) with edge lengths c e. Find a Hamiltonian cycle (cycle containing all nodes, tour) of minimum length. K 8 Hendel,Gleixner,Serrano Constraint Handler Implementation 3 / 43
6 Motivation For General Constraints: MIP Formulation min s.t. c e x e e E e δ(v) e δ(s) x e = 2 v V x e 2 S V, S x e {0, 1} e E. Hendel,Gleixner,Serrano Constraint Handler Implementation 4 / 43
7 Motivation For General Constraints: MIP Formulation min s.t. c e x e e E e δ(v) e δ(s) x e = 2 v V x e 2 S V, S x e {0, 1} e E. CIP Formulation min s.t. c e x e e E x e = 2 v V e δ(v) nosubtour(g, x) x e {0, 1} e E. nosubtour(g, x) C {e E x e = 1} : C is a cycle of length C < V Hendel,Gleixner,Serrano Constraint Handler Implementation 4 / 43
8 Separation Problem Separation Problem Given graph G = (V, E) and x [0, 1] E. Decide whether x satises all subtour elimination constraints. If not, nd violated subtour elimination constraint. Hendel,Gleixner,Serrano Constraint Handler Implementation 5 / 43
9 Separation Problem Separation Problem Given graph G = (V, E) and x [0, 1] E. Decide whether x satises all subtour elimination constraints. If not, nd violated subtour elimination constraint. Trivial observation: Consider G = (V, E) with edge capacities x e. x violates at least one subtour elimination constraint cut δ(s) with capacity x (δ(s)) < 2. Hendel,Gleixner,Serrano Constraint Handler Implementation 5 / 43
10 Separation Problem Separation Problem Given graph G = (V, E) and x [0, 1] E. Decide whether x satises all subtour elimination constraints. If not, nd violated subtour elimination constraint. Idea of separation algorithm: s, t V : Find (s, t)-cut δ(s) of minimum capacity If all cut capacities 2, all subtour elimination constraints satised ( ) V 2 times MaxFlow-MinCut algo Hendel,Gleixner,Serrano Constraint Handler Implementation 5 / 43
11 Separation Problem Separation Problem Given graph G = (V, E) and x [0, 1] E. Decide whether x satises all subtour elimination constraints. If not, nd violated subtour elimination constraint. Idea of separation algorithm: s, t V : Find (s, t)-cut δ(s) of minimum capacity If all cut capacities 2, all subtour elimination constraints satised V 1 times MaxFlow-MinCut algo within Gomory-Hu-Algorithm Hendel,Gleixner,Serrano Constraint Handler Implementation 5 / 43
12 Result of Gomory-Hu-Algorithm Gomory-Hu-Tree T For all s, t V : capacity of minimum (s, t)-cut in G : minimum label f e of all edges e in unique (s, t)-path in T minimum (s, t)-cut in G : bipartition of V obtained be deleting this edge from T Hendel,Gleixner,Serrano Constraint Handler Implementation 6 / 43
13 Outline Why use constraints? Motivation Callback: Default Constraint Handlers of SCIP Implementation hints for the Nosubtour Constraint Handler Hendel,Gleixner,Serrano Constraint Handler Implementation 7 / 43
14 Default Constraint Handlers Hendel,Gleixner,Serrano Constraint Handler Implementation 8 / 43
15 Special Linear Constraints min x 1 + x 2 + x 3 + x 4 + x 5 + y 1 + y 2 s.t. y x 1 y x 2 3 x x x 3 8 x 2 + x 3 + x 4 = 1 3 x x x y y 2 4 x 1, x 2, x 3, x 4 {0, 1} x 5 Z + y 1, y 2 R + Hendel,Gleixner,Serrano Constraint Handler Implementation 9 / 43
16 Special Linear Constraints min x 1 + x 2 + x 3 + x 4 + x 5 + y 1 + y 2 s.t. y x 1 y x 2 3 x x x 3 8 x 2 + x 3 + x 4 = 1 3 x x x y y 2 4 x 1, x 2, x 3, x 4 {0, 1} x 5 Z + y 1, y 2 R + Hendel,Gleixner,Serrano Constraint Handler Implementation 9 / 43
17 Special Linear Constraints min x 1 + x 2 + x 3 + x 4 + x 5 + y 1 + y 2 s.t. y x 1 y x 2 3 x x x 3 8 x 2 + x 3 + x 4 = 1 3 x x x y y 2 4 x 1, x 2, x 3, x 4 {0, 1} x 5 Z + y 1, y 2 R + Hendel,Gleixner,Serrano Constraint Handler Implementation 9 / 43
18 Special Linear Constraints min x 1 + x 2 + x 3 + x 4 + x 5 + y 1 + y 2 s.t. y x 1 y x 2 3 x x x 3 8 x 2 + x 3 + x 4 = 1 3 x x x y y 2 4 x 1, x 2, x 3, x 4 {0, 1} x 5 Z + y 1, y 2 R + Hendel,Gleixner,Serrano Constraint Handler Implementation 9 / 43
19 Special Linear Constraints min x 1 + x 2 + x 3 + x 4 + x 5 + y 1 + y 2 s.t. y x 1 y x 2 3 x x x 3 8 x 2 + x 3 + x 4 = 1 3 x x x y y 2 4 x 1, x 2, x 3, x 4 {0, 1} x 5 Z + y 1, y 2 R + Hendel,Gleixner,Serrano Constraint Handler Implementation 9 / 43
20 Knapsack Constraints Feasible region of 0-1 knapsack problem: { x {0, 1} N : j N a j x j b } weights of the variables: a j Z + for all j N capacity of the knapsack: b Z + Hendel,Gleixner,Serrano Constraint Handler Implementation 10 / 43
21 The Integral Constraint Handler The integral constraint handler ensures the integrality of discrete problem variables in a solution does not need constraints because integrality restrictions are encoded in the type of a variable is the shortest example of a working constraint handler I know Hendel,Gleixner,Serrano Constraint Handler Implementation 11 / 43
22 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties/Parameters Hendel,Gleixner,Serrano Constraint Handler Implementation 12 / 43
23 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties/Parameters Hendel,Gleixner,Serrano Constraint Handler Implementation 12 / 43
24 Private Data Constraint data: information needed to dene single constraint 1 struct SCIP_ConsData 2 { 3 SCIP_VAR** vars; // vari ables in knapsack 4 int nvars; // number of v ariabl es 5 SCIP_Longint* weights; // weights of v ariab les 6 SCIP_Longint capacity; // c a p a c i t y of knapsack 7 }; Constraint handler data: information belonging to constraint handler itself 1 struct SCIP_ConshdlrData 2 { 3 int maxrounds; // max nr. of sepa rounds per node 4 int maxsepacuts; // max nr. of cuts per sepa round 5 }; Hendel,Gleixner,Serrano Constraint Handler Implementation 13 / 43
25 Interface Methods Creating a single knapsack constraint. 1 SCIP_RETCODE SCIPcreateConsBasicKnapsack( 2 SCIP* scip, / < SCIP data s t r u c t u r e / 3 SCIP_CONS** cons, / < p o i n t e r to hold the created c o n s t r a i n t 4 const char* name, / < name of c o n s t r a i n t / 5 int nvars, / < number of items i n the knapsack / 6 SCIP_VAR** vars, / < array with item vari ables / 7 SCIP_Longint* weights, / < a r r a y with item weights / 8 SCIP_Longint capacity / < c a p a c i t y of knapsack / 9 ) 10 { 11 SCIP_CONSDATA* consdata; 12 SCIP_CALL( consdatacreate(scip, &consdata, nvars, vars, 13 weights, capacity) ); 14 SCIP_CALL( SCIPcreateConsBasic(scip, cons, name, conshdlr, 15 consdata) ); 16 return SCIP_OKAY; 17 } Hendel,Gleixner,Serrano Constraint Handler Implementation 14 / 43
26 Interface Methods Including the knapsack constraint handler. 1 SCIP_RETCODE SCIPincludeConshdlrKnapsack( 2 SCIP* scip // SCIP data s t r u c t u r e 3 ) 4 { 5 SCIP_CONSHDLR* conshdlr; 6 SCIP_CONSHDLRDATA* conshdlrdata; 7 8 SCIP_CALL( conshdlrdatacreate(scip, &conshdlrdata) ); / i n c l u d e c o n s t r a i n t handler i n t o SCIP / 12 SCIP_CALL( SCIPincludeConshdlrBasic(scip, &conshdlr, 13 CONSHDLR_NAME, CONSHDLR_DESC, 14 [...] 15 consenfolpknapsack, consenfopsknapsack, 16 conscheckknapsack, conslockknapsack, 17 conshdlrdata) ); / s e t non fundamental c a l l b a c k s v i a s p e c i f i c s e t t e r f u n c t i o n s / } Hendel,Gleixner,Serrano Constraint Handler Implementation 15 / 43
27 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties/Parameters Hendel,Gleixner,Serrano Constraint Handler Implementation 16 / 43
28 Callback Methods What is a callback method? Hendel,Gleixner,Serrano Constraint Handler Implementation 17 / 43
29 Callback Methods What is a callback method? Nice quote from someone who cited an older Wikipedia version: In computer programming, a callback is a reference to executable code, or a piece of executable code, that is passed as an argument to other code. This allows a lower-level software layer to call a subroutine (or function) dened in a higher-level layer. source: Hendel,Gleixner,Serrano Constraint Handler Implementation 17 / 43
30 Callback Methods What is a callback method? Nice quote from someone who cited an older Wikipedia version: In computer programming, a callback is a reference to executable code, or a piece of executable code, that is passed as an argument to other code. This allows a lower-level software layer to call a subroutine (or function) dened in a higher-level layer. source: All the functionality of constraint handlers (as well as all other plugin types of SCIP) is dened through callback functions. Hendel,Gleixner,Serrano Constraint Handler Implementation 17 / 43
31 CONSCHECK Most important callback... usually called by primal heuristics checks given solution for feasibility wrt all constraints of its type possible result values SCIP_FEASIBLE SCIP_INFEASIBLE Given solution: (x 1, x 2, x 3 ) = (1, 0, 1) Knapsack constraints: 3x 1 + 6x 2 + 4x 3 8 2x x 3 3 Result: SCIP_INFEASIBLE Hendel,Gleixner,Serrano Constraint Handler Implementation 18 / 43
32 CONSENFOLP and CONSENFOPS CONSENFOLP: checks LP solution for feasibility CONSENFOPS: checks Pseudo solution for feasibility LP solution solution of LP relaxation min x 1 x 2 + x 3 s.t. 3 x x x 3 4 x 1, x 2, x 3 {0, 1} LP solution: (0, 1 2, 0) Pseudo Solution solution of LP relaxation with only bound constraints used if LP solving disabled, or numerical diculties occurred min x 1 x 2 + x 3 s.t. x 1, x 2, x 3 {0, 1} Pseudo Solution: (0, 1, 0) Hendel,Gleixner,Serrano Constraint Handler Implementation 19 / 43
33 CONSENFOLP and CONSENFOPS LP solution may violate a constraint not contained in the relaxation. Enforcement callbacks are necessary for a correct implementation! In addition, they can resolve an infeasibility by... reducing a variable's domain, separating a cutting plane (may use integrality), adding a (local) constraint, creating a branching, concluding that the subproblem is infeasible and can be cut o, or just saying solution infeasible. Hendel,Gleixner,Serrano Constraint Handler Implementation 20 / 43
34 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
35 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
36 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
37 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
38 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
39 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
40 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
41 CONSENFOLP and CONSENFOPS Presolving Domain propagation Conict analysis Node selection Solve LP Pricing Primal heuristics Processing Branching Enforce constraints Cuts Enforcement result of constraint handler: reduced domain added cut added constraint branched cuto infeasible feasible Hendel,Gleixner,Serrano Constraint Handler Implementation 21 / 43
42 The Integral Constraint Handler The integral constraint handler ensures the integrality of discrete problem variables in a solution does not need constraints because integrality restrictions are encoded in the type of a variable is the shortest example of a working constraint handler I know Hendel,Gleixner,Serrano Constraint Handler Implementation 22 / 43
43 The Integral Constraint Handler The integral constraint handler ensures the integrality of discrete problem variables in a solution does not need constraints because integrality restrictions are encoded in the type of a variable is the shortest example of a working constraint handler I know Question How does the integral constraint handler enforce the LP solution? Hendel,Gleixner,Serrano Constraint Handler Implementation 22 / 43
44 The Integral Constraint Handler The integral constraint handler ensures the integrality of discrete problem variables in a solution does not need constraints because integrality restrictions are encoded in the type of a variable is the shortest example of a working constraint handler I know Question How does the integral constraint handler enforce the LP solution? Answer It calls the branching rules to perform variable branching. Hendel,Gleixner,Serrano Constraint Handler Implementation 22 / 43
45 CONSLOCK Provides dual information for single constraints (useful for presolving, primal heuristics,... ) For each variable of a constraint, returns whether... increasing its value, and/or decreasing its value may lead to a violation of the constraint 3 x 1 5 x x 3 7 increasing: x 1 and x 3 decreasing: x 2 Hendel,Gleixner,Serrano Constraint Handler Implementation 23 / 43
46 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties/Parameters Hendel,Gleixner,Serrano Constraint Handler Implementation 24 / 43
47 CONSINIT..., CONSEXIT... Init Presolving Init Solve Problem Transforming Transformed Solving Free Transform Transformed Free Solve CONSINIT and CONSEXIT: called after problem was transformed / before transformed problem is freed initialize and free statistics in SCIP_ConshdlrData Hendel,Gleixner,Serrano Constraint Handler Implementation 25 / 43
48 CONSINIT..., CONSEXIT... Init Presolving Init Solve Problem Transforming Transformed Solving Free Transform Transformed Free Solve CONSINITPRE and CONSEXITPRE: called before presolving starts / after presolving is nished initialize and free presolving data Hendel,Gleixner,Serrano Constraint Handler Implementation 25 / 43
49 CONSINIT..., CONSEXIT... Init Presolving Init Solve Problem Transforming Transformed Solving Free Transform Transformed Free Solve CONSINITSOL and CONSEXITSOL: called before branch-and-bound process starts / before branch-and-bound process is freed initialize and release branch-and-bound specic data Hendel,Gleixner,Serrano Constraint Handler Implementation 25 / 43
50 CONSINIT..., CONSEXIT... Init Presolving Init Solve Problem Transforming Transformed Solving Free Transform Transformed Free Solve CONSINITLP: called before rst LP relaxation is solved add linear relaxation of all "initial" constraints to the LP relaxation Hendel,Gleixner,Serrano Constraint Handler Implementation 25 / 43
51 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties/Parameters Hendel,Gleixner,Serrano Constraint Handler Implementation 26 / 43
52 CONSSEPALP, CONSSEPASOL Feasible region of 0-1 knapsack problem: { x {0, 1} N : j N a j x j b } Minimal Cover: C N j C a j > b j C\{i} a j b i C Minimal Cover Inequality x j C 1 j C 5x 1 + 6x 2 + 2x 3 + 2x 4 8 Minimal cover: C = {2, 3, 4} Minimal cover inequality: x 2 + x 3 + x 4 2 separated in knapsack constraint handler Hendel,Gleixner,Serrano Constraint Handler Implementation 27 / 43
53 CONSSEPALP, CONSSEPASOL Separation is implemented in separators and constraint handlers. 1. Separators with SEPA_PRIORITY 0 (decreasing order) 2. Constraint handlers (decreasing order of CONSHDLR_SEPAPRIORITY) 3. Separators with SEPA_PRIORITY < 0 (decreasing order) Hendel,Gleixner,Serrano Constraint Handler Implementation 28 / 43
54 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Domain propagation (during subproblem processing) Reason for domain reductions (for conict analysis) Presolving (before processing root node) Hendel,Gleixner,Serrano Constraint Handler Implementation 29 / 43
55 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Called whenever... SCIP enters/leaves subtree where local constraint exists constraint is enabled/disabled (no propagation, no separation) Hendel,Gleixner,Serrano Constraint Handler Implementation 29 / 43
56 Ingredients of a Constraint Handler Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Displaying, parsing problems Copying problems between dierent SCIP environments Hendel,Gleixner,Serrano Constraint Handler Implementation 29 / 43
57 Outline Why use constraints? Motivation Callback: Default Constraint Handlers of SCIP Implementation hints for the Nosubtour Constraint Handler Hendel,Gleixner,Serrano Constraint Handler Implementation 30 / 43
58 Exercise: TSP MIP Formulation min s.t. c e x e e E e δ(v) e δ(s) x e = 2 v V x e 2 S V, S x e {0, 1} e E. CIP Formulation min s.t. c e x e e E x e = 2 v V e δ(v) nosubtour(g, x) x e {0, 1} e E. nosubtour(g, x) C {e E x e = 1} : C is a cycle of length C < V Hendel,Gleixner,Serrano Constraint Handler Implementation 31 / 43
59 Exercise: TSP Conict Analysis Cut Pool Solution Pool Implication Graph Presolve Management LP Relaxation Search Tree Hendel,Gleixner,Serrano Constraint Handler Implementation 32 / 43
60 Exercise: TSP MIP Default Plugins Conict Analysis Cut Pool Solution Pool Implication Graph Presolve Management LP Relaxation Search Tree Hendel,Gleixner,Serrano Constraint Handler Implementation 32 / 43
61 Exercise: TSP New Solution TSP Event Handler File Reader 2-Opt MIP Heuristic Default Plugins Conict Analysis Cut Pool Solution Pool Implication Graph Presolve Management LP Relaxation Search Tree Hendel,Gleixner,Serrano Constraint Handler Implementation 32 / 43
62 Exercise: TSP New Solution Event Handler TSP File Reader Nosubtour Constraint Hdlr 2-Opt MIP Heuristic Default Plugins Conict Analysis Cut Pool Solution Pool Implication Graph Presolve Management LP Relaxation Search Tree Hendel,Gleixner,Serrano Constraint Handler Implementation 32 / 43
63 Exercise: TSP New Solution Event Handler TSP File Reader Bonus: LP Greedy Heuristic Nosubtour Constraint Hdlr 2-Opt Heuristic MIP Default Plugins Conict Analysis Cut Pool Solution Pool Implication Graph Presolve Management LP Relaxation Search Tree Hendel,Gleixner,Serrano Constraint Handler Implementation 32 / 43
64 Exercise: TSP Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSINIT..., CONSEXIT... CONSSEPALP, CONSSEPASOL CONSPROP, CONSRESPROP, CONSPRESOL CONSACTIVE, CONSDEACTIVE, CONSENABLE, CONSDISABLE CONSTRANS, CONSDELETE, CONSFREE CONSPRINT, CONSPARSE, CONSCOPY CONSDELVARS, CONSGETVARS, CONSGETNVARS CONSGETDIVEBDCHGS Further Ingredients Private data Interface methods Properties Hendel,Gleixner,Serrano Constraint Handler Implementation 33 / 43
65 Exercise: TSP Callback Methods Fundamental: CONSCHECK CONSENFOLP, CONSENFOPS CONSLOCK Additional: CONSSEPALP, CONSSEPASOL CONSTRANS, CONSDELETE Further Ingredients Private data Interface methods Properties Hendel,Gleixner,Serrano Constraint Handler Implementation 33 / 43
66 CONSCHECK Feasibility Check Given a point y Q E, decide whether y satises all subtour elimination constraints. Hendel,Gleixner,Serrano Constraint Handler Implementation 34 / 43
67 CONSCHECK Feasibility Check Given a point y Q E, decide whether y satises all subtour elimination constraints. Check Priorities make this task easier: Constraint handlers check y in decreasing order of their check priority (CONSHDLR_CHECKPRIORITY). Check priority of nosubtour(g, x) smaller than linear and integral check priorities The solution y already satises integrality and (linear) degree constraints. min s.t. c e x e e E x e = 2 v V e δ(v) nosubtour(g, x) x e {0, 1} e E. Hendel,Gleixner,Serrano Constraint Handler Implementation 34 / 43
68 CONSCHECK Feasibility Check Given a point y Q E, decide whether y satises all subtour elimination constraints. Check Priorities make this task easier: Constraint handlers check y in decreasing order of their check priority (CONSHDLR_CHECKPRIORITY). Check priority of nosubtour(g, x) smaller than linear and integral check priorities The solution y already satises integrality and (linear) degree constraints. min s.t. c e x e e E x e = 2 v V e δ(v) nosubtour(g, x) x e {0, 1} e E. Hendel,Gleixner,Serrano Constraint Handler Implementation 34 / 43
69 CONSCHECK Feasibility Check Given a point y Q E, decide whether y satises all subtour elimination constraints. Check Priorities make this task easier: Constraint handlers check y in decreasing order of their check priority (CONSHDLR_CHECKPRIORITY). Check priority of nosubtour(g, x) smaller than linear and integral check priorities The solution y already satises integrality and (linear) degree constraints. min s.t. c e x e e E x e = 2 v V e δ(v) nosubtour(g, x) x e {0, 1} e E. y is the characteristic vector of a collection of cycles when passed to the nosubtour constraint handler. Hendel,Gleixner,Serrano Constraint Handler Implementation 34 / 43
70 Understanding SCIP Callbacks SCIP callback declarations are realized as macros: What you see in cons_nosubtour.c: 1 static 2 SCIP_DECL_CONSCHECK(consCheckNosubtour) 3 { 4 // Your S o p h i s t i c a t e d Code 5 } What the compiler sees (after preprocessor and some formatting): 1 static 2 SCIP_RETCODE conschecknosubtour(scip* scip, SCIP_CONSHDLR* conshdlr, 3 SCIP_CONS** conss, int nconss, SCIP_SOL* sol, 4 SCIP_Bool checkintegrality, SCIP_Bool checklprows, 5 SCIP_Bool printreason, SCIP_RESULT* result) 6 { 7 // Your S o p h i s t i c a t e d Code 8 } SCIP_DECL_CONSCHECK is a macro dened in type_cons.h. It is responsible of expanding the callback function with the needed arguments and return value. Hendel,Gleixner,Serrano Constraint Handler Implementation 35 / 43
71 Understanding SCIP Callbacks even more A compilable check callback 1 static 2 SCIP_DECL_CONSCHECK(consCheckNosubtour) 3 { 4 / TODO: add some more code / 5 *result = SCIP_INFEASIBLE; 6 7 return SCIP_OKAY; 8 } all callbacks must return a SCIP_RETCODE! Use SCIP_OKAY if everything worked often, callbacks require to correctly assign a result-pointer, examples: *result = SCIP_[IN]FEASIBLE *result = SCIP_SEPARATED *result = SCIP_DIDNOTFIND Find the full callback denition and possible return types in type_<pluginname>.h, eg, type_cons.h for constraint handlers or type_heur.h for primal heuristics. Hendel,Gleixner,Serrano Constraint Handler Implementation 36 / 43
72 The Graph implementation The graph data structure for this exercise is dened in GomoryHuTree.h: 1 typedef struct Graph 2 { 3 int nuses; 4 int nnodes; / < number of nodes of the graph / 5 int nedges; / < number of edges / 6 int nedgesnonzero; 7 8 GRAPHNODE *nodes; / < a r r a y c o n t a i n i n g the nodes of the graph / 9 10 GRAPHEDGE *edges; / < a r r a y c o n t a i n i n g a l l h a l f e d g e s... / } GRAPH; Hendel,Gleixner,Serrano Constraint Handler Implementation 37 / 43
73 The Graph implementation Every graph edge is a half-edge in GomoryHuTree.h: 1 typedef struct GraphEdge 2 { 3 double cap; / < c a p a c i t y used i n maxflow / 4 double rcap; / < r e s i d u a l capacity used in maxflow / 5 double length; / < length of the edge measured by some f i x e d metric 6 7 struct GraphEdge *next; / < i n i n c i d e n c e l i s t of node from which edge i s em 8 struct GraphEdge *back; / < p o i n t e r to r e v e r s e edge / 9 10 GRAPHNODE *adjac; / < p o i n t e r to adjacent node / SCIP_VAR* var; } GRAPHEDGE; Every edge has a variable pointer edge->var. Hendel,Gleixner,Serrano Constraint Handler Implementation 38 / 43
74 The Graph implementation Nodes of the graph only have two relevant members (for us): 1 typedef struct GraphNode 2 { 3 int id; / < number of the node / struct GraphEdge *first_edge;/ < i n l i s t of i n c i d e n t edges / } GRAPHNODE; Hendel,Gleixner,Serrano Constraint Handler Implementation 39 / 43
75 The Graph implementation Question How to traverse all incident edges of a node? Answer: 1 GRAPHEDGE* edge = node->first_edge; 2 while( edge!= NULL ) 3 { 4 / do something with t h i s edge / 5 6 edge = edge->next; 7 } Hendel,Gleixner,Serrano Constraint Handler Implementation 40 / 43
76 The Gomory-Hu Black Box Only one method from GomoryHuTree.h is relevant for us: ghc_tree(). It has the following arguments: GRAPH* gr: the graph data structure SCIP_Bool** cuts: 2-dimensional array to return cuts, needs to be allocated by the user int* ncuts: pointer to store the number of violated cuts double minviol feasibility tolerance, use SCIPfeastol(scip) The method returns TRUE if it found at least one violated cut, and false otherwise. ghc_tree() lls cuts with all minimum (s, t)-cuts with capacity 2 minviol Hendel,Gleixner,Serrano Constraint Handler Implementation 41 / 43
77 Documentation Constraint Integer Programming Tobias Achterberg, Ph.D. dissertation, TU Berlin, Doxygen documentation, HowTos, FAQ source code scip.h, pub_*.h, type_*.h Hendel,Gleixner,Serrano Constraint Handler Implementation 42 / 43
78 Please,... build groups of 3 people laptop owner C expert IP expert raise your hand, if you get stuck. Hendel,Gleixner,Serrano Constraint Handler Implementation 43 / 43
79 Implementing Constraint Handlers in SCIP Gregor Hendel Ambros Gleixner, Felipe Serrano September 30
The MIP-Solving-Framework SCIP
The MIP-Solving-Framework SCIP Timo Berthold Zuse Institut Berlin DFG Research Center MATHEON Mathematics for key technologies Berlin, 23.05.2007 What Is A MIP? Definition MIP The optimization problem
More informationSCIP Workshop 2014, Berlin, September 30, Introduction to SCIP
SCIP Workshop 2014, Berlin, September 30, 2014 Introduction to SCIP ZIB: Fast Algorithms Fast Computers Konrad-Zuse-Zentrum für Informationstechnik Berlin non-university research institute of the state
More informationBasic Concepts of Constraint Integer Programming
Basic Concepts of Constraint Integer Programming Ambros Gleixner Zuse Institute Berlin September 30, 2015 Outline SCIP Solving Constraint Integer Programs 4 methodologies in optimization An integrated
More informationIntroduction to SCIP. Gregor Hendel, SCIP Introduction Day September 26, rd IMI-ISM-ZIB MODAL Workshop
Introduction to SCIP Gregor Hendel, hendel@zib.de SCIP Introduction Day September 26, 2018 3rd IMI-ISM-ZIB MODAL Workshop What is SCIP? SCIP (Solving Constraint Integer Programs)... provides a full-scale
More informationSCIP Workshop Visit scip.zib.de/workshop2018 for all information. Tomorrow and on Thursday: Scientific Talks
SCIP Workshop 2018 Structure of the SCIP Introduction Day 9:30 11:00 Introduction to SCIP 11:00 11:30 Coffee Break 11:30 12:30 Installation of SCIP and GCG 12:30 14:00 Lunch 14:00 17:30 Programming Exercises
More informationPrimal Heuristics in SCIP
Primal Heuristics in SCIP Timo Berthold Zuse Institute Berlin DFG Research Center MATHEON Mathematics for key technologies Berlin, 10/11/2007 Outline 1 Introduction Basics Integration Into SCIP 2 Available
More informationMethods and Models for Combinatorial Optimization Exact methods for the Traveling Salesman Problem
Methods and Models for Combinatorial Optimization Exact methods for the Traveling Salesman Problem L. De Giovanni M. Di Summa The Traveling Salesman Problem (TSP) is an optimization problem on a directed
More informationFundamentals of Integer Programming
Fundamentals of Integer Programming Di Yuan Department of Information Technology, Uppsala University January 2018 Outline Definition of integer programming Formulating some classical problems with integer
More informationAddressing degeneracy in the dual simplex algorithm using a decompositon approach
Addressing degeneracy in the dual simplex algorithm using a decompositon approach Ambros Gleixner, Stephen J Maher, Matthias Miltenberger Zuse Institute Berlin Berlin, Germany 16th July 2015 @sj_maher
More informationMotivation for Heuristics
MIP Heuristics 1 Motivation for Heuristics Why not wait for branching? Produce feasible solutions as quickly as possible Often satisfies user demands Avoid exploring unproductive sub trees Better reduced
More informationMVE165/MMG631 Linear and integer optimization with applications Lecture 9 Discrete optimization: theory and algorithms
MVE165/MMG631 Linear and integer optimization with applications Lecture 9 Discrete optimization: theory and algorithms Ann-Brith Strömberg 2018 04 24 Lecture 9 Linear and integer optimization with applications
More informationCombinatorial Optimization - Lecture 14 - TSP EPFL
Combinatorial Optimization - Lecture 14 - TSP EPFL 2012 Plan Simple heuristics Alternative approaches Best heuristics: local search Lower bounds from LP Moats Simple Heuristics Nearest Neighbor (NN) Greedy
More informationSolving lexicographic multiobjective MIPs with Branch-Cut-Price
Solving lexicographic multiobjective MIPs with Branch-Cut-Price Marta Eso (The Hotchkiss School) Laszlo Ladanyi (IBM T.J. Watson Research Center) David Jensen (IBM T.J. Watson Research Center) McMaster
More informationAdvanced Use of GAMS Solver Links
Advanced Use of GAMS Solver Links Michael Bussieck, Steven Dirkse, Stefan Vigerske GAMS Development 8th January 2013, ICS Conference, Santa Fe Standard GAMS solve Solve william minimizing cost using mip;
More informationMVE165/MMG630, Applied Optimization Lecture 8 Integer linear programming algorithms. Ann-Brith Strömberg
MVE165/MMG630, Integer linear programming algorithms Ann-Brith Strömberg 2009 04 15 Methods for ILP: Overview (Ch. 14.1) Enumeration Implicit enumeration: Branch and bound Relaxations Decomposition methods:
More informationHandling first-order linear constraints with SCIP
Handling first-order linear constraints with SCIP James Cussens, University of York KU Leuven, 2015-02-16 James Cussens, University of York FO with SCIP KU Leuven, 2015-02-16 1 / 18 MIP Mixed integer programs
More informationSCIP. 1 Introduction. 2 Model requirements. Contents. Stefan Vigerske, Humboldt University Berlin, Germany
SCIP Stefan Vigerske, Humboldt University Berlin, Germany Contents 1 Introduction.................................................. 673 2 Model requirements..............................................
More informationExploiting Degeneracy in MIP
Exploiting Degeneracy in MIP Tobias Achterberg 9 January 2018 Aussois Performance Impact in Gurobi 7.5+ 35% 32.0% 30% 25% 20% 15% 14.6% 10% 5.7% 7.9% 6.6% 5% 0% 2.9% 1.2% 0.1% 2.6% 2.6% Time limit: 10000
More informationThe Gurobi Solver V1.0
The Gurobi Solver V1.0 Robert E. Bixby Gurobi Optimization & Rice University Ed Rothberg, Zonghao Gu Gurobi Optimization 1 1 Oct 09 Overview Background Rethinking the MIP solver Introduction Tree of Trees
More informationWelcome to the Webinar. What s New in Gurobi 7.5
Welcome to the Webinar What s New in Gurobi 7.5 Speaker Introduction Dr. Tobias Achterberg Director of R&D at Gurobi Optimization Formerly a developer at ILOG, where he worked on CPLEX 11.0 to 12.6 Obtained
More informationMathematical Tools for Engineering and Management
Mathematical Tools for Engineering and Management Lecture 8 8 Dec 0 Overview Models, Data and Algorithms Linear Optimization Mathematical Background: Polyhedra, Simplex-Algorithm Sensitivity Analysis;
More informationDETERMINISTIC OPERATIONS RESEARCH
DETERMINISTIC OPERATIONS RESEARCH Models and Methods in Optimization Linear DAVID J. RADER, JR. Rose-Hulman Institute of Technology Department of Mathematics Terre Haute, IN WILEY A JOHN WILEY & SONS,
More informationIntroduction to Mathematical Programming IE496. Final Review. Dr. Ted Ralphs
Introduction to Mathematical Programming IE496 Final Review Dr. Ted Ralphs IE496 Final Review 1 Course Wrap-up: Chapter 2 In the introduction, we discussed the general framework of mathematical modeling
More informationOutline. Modeling. Outline DMP204 SCHEDULING, TIMETABLING AND ROUTING. 1. Models Lecture 5 Mixed Integer Programming Models and Exercises
Outline DMP204 SCHEDULING, TIMETABLING AND ROUTING 1. Lecture 5 Mixed Integer Programming and Exercises Marco Chiarandini 2. 3. 2 Outline Modeling 1. Min cost flow Shortest path 2. Max flow Assignment
More informationTheorem 2.9: nearest addition algorithm
There are severe limits on our ability to compute near-optimal tours It is NP-complete to decide whether a given undirected =(,)has a Hamiltonian cycle An approximation algorithm for the TSP can be used
More informationColumn Generation Based Primal Heuristics
Column Generation Based Primal Heuristics C. Joncour, S. Michel, R. Sadykov, D. Sverdlov, F. Vanderbeck University Bordeaux 1 & INRIA team RealOpt Outline 1 Context Generic Primal Heuristics The Branch-and-Price
More informationDM515 Spring 2011 Weekly Note 7
Institut for Matematik og Datalogi Syddansk Universitet May 18, 2011 JBJ DM515 Spring 2011 Weekly Note 7 Stuff covered in Week 20: MG sections 8.2-8,3 Overview of the course Hints for the exam Note that
More informationThe SCIP Optimization Suite 3.2
The SCIP Optimization Suite 3.2 Gerald Gamrath 1, Tobias Fischer 2, Tristan Gally 2, Ambros M. Gleixner 1, Gregor Hendel 1, Thorsten Koch 1, Stephen J. Maher 1, Matthias Miltenberger 1, Benjamin Müller
More informationColumn Generation Method for an Agent Scheduling Problem
Column Generation Method for an Agent Scheduling Problem Balázs Dezső Alpár Jüttner Péter Kovács Dept. of Algorithms and Their Applications, and Dept. of Operations Research Eötvös Loránd University, Budapest,
More informationCombinatorial Optimization Lab No. 10 Traveling Salesman Problem
Combinatorial Optimization Lab No. 10 Traveling Salesman Problem Industrial Informatics Research Center http://industrialinformatics.cz/ May 29, 2018 Abstract In this lab we review various ways how to
More informationAlgorithms II MIP Details
Algorithms II MIP Details What s Inside Gurobi Optimizer Algorithms for continuous optimization Algorithms for discrete optimization Automatic presolve for both LP and MIP Algorithms to analyze infeasible
More informationFebruary 19, Integer programming. Outline. Problem formulation. Branch-andbound
Olga Galinina olga.galinina@tut.fi ELT-53656 Network Analysis and Dimensioning II Department of Electronics and Communications Engineering Tampere University of Technology, Tampere, Finland February 19,
More information56:272 Integer Programming & Network Flows Final Examination -- December 14, 1998
56:272 Integer Programming & Network Flows Final Examination -- December 14, 1998 Part A: Answer any four of the five problems. (15 points each) 1. Transportation problem 2. Integer LP Model Formulation
More informationHow to use your favorite MIP Solver: modeling, solving, cannibalizing. Andrea Lodi University of Bologna, Italy
How to use your favorite MIP Solver: modeling, solving, cannibalizing Andrea Lodi University of Bologna, Italy andrea.lodi@unibo.it January-February, 2012 @ Universität Wien A. Lodi, How to use your favorite
More informationA Generic Separation Algorithm and Its Application to the Vehicle Routing Problem
A Generic Separation Algorithm and Its Application to the Vehicle Routing Problem Presented by: Ted Ralphs Joint work with: Leo Kopman Les Trotter Bill Pulleyblank 1 Outline of Talk Introduction Description
More informationBenders in a nutshell Matteo Fischetti, University of Padova
Benders in a nutshell Matteo Fischetti, University of Padova ODS 2017, Sorrento, September 2017 1 Benders decomposition The original Benders decomposition from the 1960s uses two distinct ingredients for
More information12.1 Formulation of General Perfect Matching
CSC5160: Combinatorial Optimization and Approximation Algorithms Topic: Perfect Matching Polytope Date: 22/02/2008 Lecturer: Lap Chi Lau Scribe: Yuk Hei Chan, Ling Ding and Xiaobing Wu In this lecture,
More information15.083J Integer Programming and Combinatorial Optimization Fall Enumerative Methods
5.8J Integer Programming and Combinatorial Optimization Fall 9 A knapsack problem Enumerative Methods Let s focus on maximization integer linear programs with only binary variables For example: a knapsack
More information5.3 Cutting plane methods and Gomory fractional cuts
5.3 Cutting plane methods and Gomory fractional cuts (ILP) min c T x s.t. Ax b x 0integer feasible region X Assumption: a ij, c j and b i integer. Observation: The feasible region of an ILP can be described
More informationExact solutions to mixed-integer linear programming problems
Exact solutions to mixed-integer linear programming problems Dan Steffy Zuse Institute Berlin and Oakland University Joint work with Bill Cook, Thorsten Koch and Kati Wolter November 18, 2011 Mixed-Integer
More informationLast topic: Summary; Heuristics and Approximation Algorithms Topics we studied so far:
Last topic: Summary; Heuristics and Approximation Algorithms Topics we studied so far: I Strength of formulations; improving formulations by adding valid inequalities I Relaxations and dual problems; obtaining
More informationInteger Programming. Xi Chen. Department of Management Science and Engineering International Business School Beijing Foreign Studies University
Integer Programming Xi Chen Department of Management Science and Engineering International Business School Beijing Foreign Studies University Xi Chen (chenxi0109@bfsu.edu.cn) Integer Programming 1 / 42
More informationSolutions to Assignment# 4
Solutions to Assignment# 4 Liana Yepremyan 1 Nov.12: Text p. 651 problem 1 Solution: (a) One example is the following. Consider the instance K = 2 and W = {1, 2, 1, 2}. The greedy algorithm would load
More informationApproximation Algorithms
Approximation Algorithms Prof. Tapio Elomaa tapio.elomaa@tut.fi Course Basics A 4 credit unit course Part of Theoretical Computer Science courses at the Laboratory of Mathematics There will be 4 hours
More informationApplied Mixed Integer Programming: Beyond 'The Optimum'
Applied Mixed Integer Programming: Beyond 'The Optimum' 14 Nov 2016, Simons Institute, Berkeley Pawel Lichocki Operations Research Team, Google https://developers.google.com/optimization/ Applied Mixed
More informationCS599: Convex and Combinatorial Optimization Fall 2013 Lecture 14: Combinatorial Problems as Linear Programs I. Instructor: Shaddin Dughmi
CS599: Convex and Combinatorial Optimization Fall 2013 Lecture 14: Combinatorial Problems as Linear Programs I Instructor: Shaddin Dughmi Announcements Posted solutions to HW1 Today: Combinatorial problems
More informationA constant-factor approximation algorithm for the asymmetric travelling salesman problem
A constant-factor approximation algorithm for the asymmetric travelling salesman problem London School of Economics Joint work with Ola Svensson and Jakub Tarnawski cole Polytechnique F d rale de Lausanne
More informationSolutions for the Exam 6 January 2014
Mastermath and LNMB Course: Discrete Optimization Solutions for the Exam 6 January 2014 Utrecht University, Educatorium, 13:30 16:30 The examination lasts 3 hours. Grading will be done before January 20,
More informationOn Mixed-Integer (Linear) Programming and its connection with Data Science
On Mixed-Integer (Linear) Programming and its connection with Data Science Andrea Lodi Canada Excellence Research Chair École Polytechnique de Montréal, Québec, Canada andrea.lodi@polymtl.ca January 16-20,
More informationThe Gurobi Optimizer. Bob Bixby
The Gurobi Optimizer Bob Bixby Outline Gurobi Introduction Company Products Benchmarks Gurobi Technology Rethinking MIP MIP as a bag of tricks 8-Jul-11 2010 Gurobi Optimization 2 Gurobi Optimization Incorporated
More informationAlgorithms for Integer Programming
Algorithms for Integer Programming Laura Galli November 9, 2016 Unlike linear programming problems, integer programming problems are very difficult to solve. In fact, no efficient general algorithm is
More informationConstruction of Minimum-Weight Spanners Mikkel Sigurd Martin Zachariasen
Construction of Minimum-Weight Spanners Mikkel Sigurd Martin Zachariasen University of Copenhagen Outline Motivation and Background Minimum-Weight Spanner Problem Greedy Spanner Algorithm Exact Algorithm:
More informationInteger Programming Chapter 9
1 Integer Programming Chapter 9 University of Chicago Booth School of Business Kipp Martin October 30, 2017 2 Outline Branch and Bound Theory Branch and Bound Linear Programming Node Selection Strategies
More informationAlgorithms for Euclidean TSP
This week, paper [2] by Arora. See the slides for figures. See also http://www.cs.princeton.edu/~arora/pubs/arorageo.ps Algorithms for Introduction This lecture is about the polynomial time approximation
More informationColumn Generation and its applications
Column Generation and its applications Murat Firat, dept. IE&IS, TU/e BPI Cluster meeting Outline Some real-life decision problems Standard formulations Basics of Column Generation Master formulations
More informationAlgorithm Design and Analysis
Algorithm Design and Analysis LECTURE 29 Approximation Algorithms Load Balancing Weighted Vertex Cover Reminder: Fill out SRTEs online Don t forget to click submit Sofya Raskhodnikova 12/7/2016 Approximation
More informationAssignment 3b: The traveling salesman problem
Chalmers University of Technology MVE165 University of Gothenburg MMG631 Mathematical Sciences Linear and integer optimization Optimization with applications Emil Gustavsson Assignment information Ann-Brith
More informationMachine Learning for Software Engineering
Machine Learning for Software Engineering Introduction and Motivation Prof. Dr.-Ing. Norbert Siegmund Intelligent Software Systems 1 2 Organizational Stuff Lectures: Tuesday 11:00 12:30 in room SR015 Cover
More informationA List Heuristic for Vertex Cover
A List Heuristic for Vertex Cover Happy Birthday Vasek! David Avis McGill University Tomokazu Imamura Kyoto University Operations Research Letters (to appear) Online: http://cgm.cs.mcgill.ca/ avis revised:
More informationAPPROXIMATION ALGORITHMS FOR GEOMETRIC PROBLEMS
APPROXIMATION ALGORITHMS FOR GEOMETRIC PROBLEMS Subhas C. Nandy (nandysc@isical.ac.in) Advanced Computing and Microelectronics Unit Indian Statistical Institute Kolkata 70010, India. Organization Introduction
More informationUnit.9 Integer Programming
Unit.9 Integer Programming Xiaoxi Li EMS & IAS, Wuhan University Dec. 22-29, 2016 (revised) Operations Research (Li, X.) Unit.9 Integer Programming Dec. 22-29, 2016 (revised) 1 / 58 Organization of this
More informationDual-fitting analysis of Greedy for Set Cover
Dual-fitting analysis of Greedy for Set Cover We showed earlier that the greedy algorithm for set cover gives a H n approximation We will show that greedy produces a solution of cost at most H n OPT LP
More informationlpsymphony - Integer Linear Programming in R
lpsymphony - Integer Linear Programming in R Vladislav Kim October 30, 2017 Contents 1 Introduction 2 2 lpsymphony: Quick Start 2 3 Integer Linear Programming 5 31 Equivalent and Dual Formulations 5 32
More informationA LARGE SCALE INTEGER AND COMBINATORIAL OPTIMIZER
A LARGE SCALE INTEGER AND COMBINATORIAL OPTIMIZER By Qun Chen A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Industrial Engineering) at the
More informationComputational Integer Programming. Lecture 12: Branch and Cut. Dr. Ted Ralphs
Computational Integer Programming Lecture 12: Branch and Cut Dr. Ted Ralphs Computational MILP Lecture 12 1 Reading for This Lecture Wolsey Section 9.6 Nemhauser and Wolsey Section II.6 Martin Computational
More informationto the Traveling Salesman Problem 1 Susanne Timsj Applied Optimization and Modeling Group (TOM) Department of Mathematics and Physics
An Application of Lagrangian Relaxation to the Traveling Salesman Problem 1 Susanne Timsj Applied Optimization and Modeling Group (TOM) Department of Mathematics and Physics M lardalen University SE-721
More informationThe Heuristic (Dark) Side of MIP Solvers. Asja Derviskadic, EPFL Vit Prochazka, NHH Christoph Schaefer, EPFL
The Heuristic (Dark) Side of MIP Solvers Asja Derviskadic, EPFL Vit Prochazka, NHH Christoph Schaefer, EPFL 1 Table of content [Lodi], The Heuristic (Dark) Side of MIP Solvers, Hybrid Metaheuristics, 273-284,
More informationval(y, I) α (9.0.2) α (9.0.3)
CS787: Advanced Algorithms Lecture 9: Approximation Algorithms In this lecture we will discuss some NP-complete optimization problems and give algorithms for solving them that produce a nearly optimal,
More informationPrimal Heuristics for Branch-and-Price Algorithms
Primal Heuristics for Branch-and-Price Algorithms Marco Lübbecke and Christian Puchert Abstract In this paper, we present several primal heuristics which we implemented in the branch-and-price solver GCG
More informationExact Algorithms for NP-hard problems
24 mai 2012 1 Why do we need exponential algorithms? 2 3 Why the P-border? 1 Practical reasons (Jack Edmonds, 1965) For practical purposes the difference between algebraic and exponential order is more
More informationTwo-layer Network Design by Branch-and-Cut featuring MIP-based Heuristics
Two-layer Network Design by Branch-and-Cut featuring MIP-based Heuristics Sebastian Orlowski, Zuse Institute Berlin, Takustr. 7, D-14195 Berlin, orlowski@zib.de Arie M.C.A. Koster, Zuse Institute Berlin,
More informationPractice Final Exam 2: Solutions
lgorithm Design Techniques Practice Final Exam 2: Solutions 1. The Simplex lgorithm. (a) Take the LP max x 1 + 2x 2 s.t. 2x 1 + x 2 3 x 1 x 2 2 x 1, x 2 0 and write it in dictionary form. Pivot: add x
More informationCloud Branching MIP workshop, Ohio State University, 23/Jul/2014
Cloud Branching MIP workshop, Ohio State University, 23/Jul/2014 Timo Berthold Xpress Optimization Team Gerald Gamrath Zuse Institute Berlin Domenico Salvagnin Universita degli Studi di Padova This presentation
More informationInteger Programming Theory
Integer Programming Theory Laura Galli October 24, 2016 In the following we assume all functions are linear, hence we often drop the term linear. In discrete optimization, we seek to find a solution x
More informationQuestions? You are given the complete graph of Facebook. What questions would you ask? (What questions could we hope to answer?)
P vs. NP What now? Attribution These slides were prepared for the New Jersey Governor s School course The Math Behind the Machine taught in the summer of 2011 by Grant Schoenebeck Large parts of these
More informationTutorial on Integer Programming for Visual Computing
Tutorial on Integer Programming for Visual Computing Peter Wonka and Chi-han Peng November 2018 1 1 Notation The vector space is denoted as R,R n,r m n,v,w Matricies are denoted by upper case, italic,
More informationThe Subtour LP for the Traveling Salesman Problem
The Subtour LP for the Traveling Salesman Problem David P. Williamson Cornell University November 22, 2011 Joint work with Jiawei Qian, Frans Schalekamp, and Anke van Zuylen The Traveling Salesman Problem
More informationInteger Programming! Using linear programming to solve discrete problems
Integer Programming! Using linear programming to solve discrete problems Solving Discrete Problems Linear programming solves continuous problem! problems over the reai numbers.! For the remainder of the
More informationPrimal Dual Schema Approach to the Labeling Problem with Applications to TSP
1 Primal Dual Schema Approach to the Labeling Problem with Applications to TSP Colin Brown, Simon Fraser University Instructor: Ramesh Krishnamurti The Metric Labeling Problem has many applications, especially
More informationInteger Programming ISE 418. Lecture 7. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 7 Dr. Ted Ralphs ISE 418 Lecture 7 1 Reading for This Lecture Nemhauser and Wolsey Sections II.3.1, II.3.6, II.4.1, II.4.2, II.5.4 Wolsey Chapter 7 CCZ Chapter 1 Constraint
More informationThe SCIP Optimization Suite 4.0
Zuse Institute Berlin Takustrasse 7 D-14195 Berlin-Dahlem Germany STEPHEN J. MAHER, TOBIAS FISCHER, TRISTAN GALLY, GERALD GAMRATH, AMBROS GLEIXNER, ROBERT LION GOTTWALD, GREGOR HENDEL, THORSTEN KOCH, MARCO
More information3 INTEGER LINEAR PROGRAMMING
3 INTEGER LINEAR PROGRAMMING PROBLEM DEFINITION Integer linear programming problem (ILP) of the decision variables x 1,..,x n : (ILP) subject to minimize c x j j n j= 1 a ij x j x j 0 x j integer n j=
More informationCSE 417 Branch & Bound (pt 4) Branch & Bound
CSE 417 Branch & Bound (pt 4) Branch & Bound Reminders > HW8 due today > HW9 will be posted tomorrow start early program will be slow, so debugging will be slow... Review of previous lectures > Complexity
More informationInteger Programming and Network Modeis
H.A. Eiselt C.-L. Sandblom Integer Programming and Network Modeis With Contributions by K. Spielberg, E. Richards, B.T. Smith, G. Laporte, B.T. Boffey With 165 Figures and 43 Tables &m Springer CONTENTS
More informationModern Benders (in a nutshell)
Modern Benders (in a nutshell) Matteo Fischetti, University of Padova (based on joint work with Ivana Ljubic and Markus Sinnl) Lunteren Conference on the Mathematics of Operations Research, January 17,
More information7KH9HKLFOH5RXWLQJSUREOHP
7K9KO5RXWJSUREOP Given a set of vehicles with a certain capacity located at a depot and a set of customers with different demands at various locations, the vehicle routing problem (VRP) is how to satisfy
More informationA Steady-State Genetic Algorithm for Traveling Salesman Problem with Pickup and Delivery
A Steady-State Genetic Algorithm for Traveling Salesman Problem with Pickup and Delivery Monika Sharma 1, Deepak Sharma 2 1 Research Scholar Department of Computer Science and Engineering, NNSS SGI Samalkha,
More informationBenders Decomposition
Benders Decomposition Using projections to solve problems thst@man.dtu.dk DTU-Management Technical University of Denmark 1 Outline Introduction Using projections Benders decomposition Simple plant location
More informationOutline. 1 The matching problem. 2 The Chinese Postman Problem
Outline The matching problem Maximum-cardinality matchings in bipartite graphs Maximum-cardinality matchings in bipartite graphs: The augmenting path algorithm 2 Let G = (V, E) be an undirected graph.
More informationHeuristics in MILP. Group 1 D. Assouline, N. Molyneaux, B. Morén. Supervisors: Michel Bierlaire, Andrea Lodi. Zinal 2017 Winter School
Heuristics in MILP Group 1 D. Assouline, N. Molyneaux, B. Morén Supervisors: Michel Bierlaire, Andrea Lodi Zinal 2017 Winter School 0 / 23 Primal heuristics Original paper: Fischetti, M. and Lodi, A. (2011).
More informationAlgorithms for Decision Support. Integer linear programming models
Algorithms for Decision Support Integer linear programming models 1 People with reduced mobility (PRM) require assistance when travelling through the airport http://www.schiphol.nl/travellers/atschiphol/informationforpassengerswithreducedmobility.htm
More informationAdvanced Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras
Advanced Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras Lecture 18 All-Integer Dual Algorithm We continue the discussion on the all integer
More informationDecision Problems. Observation: Many polynomial algorithms. Questions: Can we solve all problems in polynomial time? Answer: No, absolutely not.
Decision Problems Observation: Many polynomial algorithms. Questions: Can we solve all problems in polynomial time? Answer: No, absolutely not. Definition: The class of problems that can be solved by polynomial-time
More informationLecture 14: Linear Programming II
A Theorist s Toolkit (CMU 18-859T, Fall 013) Lecture 14: Linear Programming II October 3, 013 Lecturer: Ryan O Donnell Scribe: Stylianos Despotakis 1 Introduction At a big conference in Wisconsin in 1948
More informationModelling of LP-problems (2WO09)
Modelling of LP-problems (2WO09) assignor: Judith Keijsper room: HG 9.31 email: J.C.M.Keijsper@tue.nl course info : http://www.win.tue.nl/ jkeijspe Technische Universiteit Eindhoven meeting 1 J.Keijsper
More informationsbb COIN-OR Simple Branch-and-Cut
sbb COIN-OR Simple Branch-and-Cut John Forrest Lou Hafer lou@cs.sfu.ca CORS/INFORMS Joint Meeting Banff, May 2004 p.1/12 Outline Basic Branch-and-Bound Branching Node Selection Cuts and Heuristics Using
More informationMartin W.P. Savelsbergh. Georgia Institute of Technology. School of Industrial and Systems Engineering
A MINTO short course 1 Martin W.P. Savelsbergh George L. Nemhauser Introduction Georgia Institute of Technology School of Industrial and Systems Engineering Atlanta, GA 30332-0205 (February 1, 1995) MINTO
More informationGurobi Guidelines for Numerical Issues February 2017
Gurobi Guidelines for Numerical Issues February 2017 Background Models with numerical issues can lead to undesirable results: slow performance, wrong answers or inconsistent behavior. When solving a model
More informationSolutions for Operations Research Final Exam
Solutions for Operations Research Final Exam. (a) The buffer stock is B = i a i = a + a + a + a + a + a 6 + a 7 = + + + + + + =. And the transportation tableau corresponding to the transshipment problem
More information56:272 Integer Programming & Network Flows Final Exam -- December 16, 1997
56:272 Integer Programming & Network Flows Final Exam -- December 16, 1997 Answer #1 and any five of the remaining six problems! possible score 1. Multiple Choice 25 2. Traveling Salesman Problem 15 3.
More information