intlinprog
Mixed-integer linear programming (MILP)
Syntax
Description
Mixed-integer linear programming solver.
Finds the minimum of a problem specified by
f, x, intcon, b, beq, lb, and ub are vectors, and A and Aeq are matrices.
You can specify f, intcon, lb, and ub as vectors or arrays. See Matrix Arguments.
Note
intlinprog
applies only to the solver-based approach. For a discussion
of the two optimization approaches, see First Choose Problem-Based or Solver-Based Approach.
uses a
x
= intlinprog(problem
)problem
structure to encapsulate all solver inputs. You
can import a problem
structure from an MPS file using
mpsread
. You can also create a
problem
structure from an
OptimizationProblem
object by using prob2struct
.
Examples
Solve an MILP with Linear Inequalities
Solve the problem
Write the objective function vector and vector of integer variables.
f = [8;1]; intcon = 2;
Convert all inequalities into the form A*x <= b
by multiplying “greater than” inequalities by -1
.
A = [-1,-2; -4,-1; 2,1]; b = [14;-33;20];
Call intlinprog
.
x = intlinprog(f,intcon,A,b)
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [1e+00, 4e+00] Cost [1e+00, 8e+00] Bound [0e+00, 0e+00] RHS [1e+01, 3e+01] Presolving model 3 rows, 2 cols, 6 nonzeros 0s 3 rows, 2 cols, 6 nonzeros 0s Solving MIP model with: 3 rows 2 cols (0 binary, 1 integer, 0 implied int., 1 continuous) 6 nonzeros Nodes | B&B Tree | Objective Bounds | Dynamic Constraints | Work Proc. InQueue | Leaves Expl. | BestBound BestSol Gap | Cuts InLp Confl. | LpIters Time 0 0 0 0.00% -inf inf inf 0 0 0 0 0.0s T 0 0 0 0.00% -inf 59 Large 0 0 0 3 0.0s Solving report Status Optimal Primal bound 59 Dual bound 59 Gap 0% (tolerance: 0.01%) Solution status feasible 59 (objective) 0 (bound viol.) 1.7763568394e-15 (int. viol.) 0 (row viol.) Timing 0.00 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 1 LP iterations 3 (total) 0 (strong br.) 0 (separation) 0 (heuristics) Optimal solution found. Intlinprog stopped at the root node because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
x = 2×1
6.5000
7.0000
Solve an MILP with All Types of Constraints
Solve the problem
Write the objective function vector and vector of integer variables.
f = [-3;-2;-1]; intcon = 3;
Write the linear inequality constraints.
A = [1,1,1]; b = 7;
Write the linear equality constraints.
Aeq = [4,2,1]; beq = 12;
Write the bound constraints.
lb = zeros(3,1);
ub = [Inf;Inf;1]; % Enforces x(3) is binary
Call intlinprog
.
x = intlinprog(f,intcon,A,b,Aeq,beq,lb,ub)
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [1e+00, 4e+00] Cost [1e+00, 3e+00] Bound [1e+00, 1e+00] RHS [7e+00, 1e+01] Presolving model 2 rows, 3 cols, 6 nonzeros 0s 0 rows, 0 cols, 0 nonzeros 0s Presolve: Optimal Solving report Status Optimal Primal bound -12 Dual bound -12 Gap 0% (tolerance: 0.01%) Solution status feasible -12 (objective) 0 (bound viol.) 0 (int. viol.) 0 (row viol.) Timing 0.00 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 0 LP iterations 0 (total) 0 (strong br.) 0 (separation) 0 (heuristics) Optimal solution found. Intlinprog stopped at the root node because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
x = 3×1
0
6
0
Use Initial Point
Compare the number of steps to solve an integer programming problem both with and without an initial feasible point. The problem has eight variables, four linear equality constraints, and has all variables restricted to be positive.
Define the linear equality constraint matrix and vector.
Aeq = [22 13 26 33 21 3 14 26 39 16 22 28 26 30 23 24 18 14 29 27 30 38 26 26 41 26 28 36 18 38 16 26]; beq = [ 7872 10466 11322 12058];
Set lower bounds that restrict all variables to be nonnegative.
N = 8; lb = zeros(N,1);
Specify that all variables are integer-valued.
intcon = 1:N;
Set the objective function vector f
.
f = [2 10 13 17 7 5 7 3];
Solve the problem without using an initial point, and examine the display to see the number of branch-and-bound nodes.
[x1,fval1,exitflag1,output1] = intlinprog(f,intcon,[],[],Aeq,beq,lb);
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [3e+00, 4e+01] Cost [2e+00, 2e+01] Bound [0e+00, 0e+00] RHS [8e+03, 1e+04] Presolving model 4 rows, 8 cols, 32 nonzeros 0s 4 rows, 8 cols, 27 nonzeros 0s Objective function is integral with scale 1 Solving MIP model with: 4 rows 8 cols (0 binary, 8 integer, 0 implied int., 0 continuous) 27 nonzeros Nodes | B&B Tree | Objective Bounds | Dynamic Constraints | Work Proc. InQueue | Leaves Expl. | BestBound BestSol Gap | Cuts InLp Confl. | LpIters Time 0 0 0 0.00% 0 inf inf 0 0 0 0 0.0s 0 0 0 0.00% 1554.047531 inf inf 0 0 4 4 0.0s T 20753 210 8189 98.04% 1783.696925 1854 3.79% 30 8 9884 19222 3.0s Solving report Status Optimal Primal bound 1854 Dual bound 1854 Gap 0% (tolerance: 0.01%) Solution status feasible 1854 (objective) 0 (bound viol.) 9.63673585375e-14 (int. viol.) 0 (row viol.) Timing 3.08 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 21163 LP iterations 19608 (total) 223 (strong br.) 76 (separation) 1018 (heuristics) Optimal solution found. Intlinprog stopped because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
For comparison, find the solution using an initial feasible point.
x0 = [8 62 23 103 53 84 46 34]; [x2,fval2,exitflag2,output2] = intlinprog(f,intcon,[],[],Aeq,beq,lb,[],x0);
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [3e+00, 4e+01] Cost [2e+00, 2e+01] Bound [0e+00, 0e+00] RHS [8e+03, 1e+04] Assessing feasibility of MIP using primal feasibility and integrality tolerance of 1e-06 Solution has num max sum Col infeasibilities 0 0 0 Integer infeasibilities 0 0 0 Row infeasibilities 0 0 0 Row residuals 0 0 0 Presolving model 4 rows, 8 cols, 32 nonzeros 0s 4 rows, 8 cols, 27 nonzeros 0s MIP start solution is feasible, objective value is 3901 Objective function is integral with scale 1 Solving MIP model with: 4 rows 8 cols (0 binary, 8 integer, 0 implied int., 0 continuous) 27 nonzeros Nodes | B&B Tree | Objective Bounds | Dynamic Constraints | Work Proc. InQueue | Leaves Expl. | BestBound BestSol Gap | Cuts InLp Confl. | LpIters Time 0 0 0 0.00% 0 3901 100.00% 0 0 0 0 0.0s 0 0 0 0.00% 1554.047531 3901 60.16% 0 0 4 4 0.0s T 6266 708 2644 73.61% 1662.791423 3301 49.63% 20 6 9746 10699 1.6s T 9340 919 3970 80.72% 1692.410008 2687 37.01% 29 6 9995 16120 2.4s T 21750 192 9514 96.83% 1791.542628 1854 3.37% 20 6 9984 40278 5.8s Solving report Status Optimal Primal bound 1854 Dual bound 1854 Gap 0% (tolerance: 0.01%) Solution status feasible 1854 (objective) 0 (bound viol.) 1.42108547152e-13 (int. viol.) 0 (row viol.) Timing 5.86 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 22163 LP iterations 40863 (total) 538 (strong br.) 64 (separation) 2782 (heuristics) Optimal solution found. Intlinprog stopped because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
Without an initial point,
intlinprog
took about 30,000 branch-and-bound steps.Using an initial point,
intlinprog
took about 5,000 steps.
Giving an initial point does not always help. For this problem, giving an initial point saves time and computational steps. However, for some problems, giving an initial point can cause intlinprog
to take more steps.
Solve an MILP with Nondefault Options
Solve the problem
without showing iterative display.
Specify the solver inputs.
f = [-3;-2;-1];
intcon = 3;
A = [1,1,1];
b = 7;
Aeq = [4,2,1];
beq = 12;
lb = zeros(3,1);
ub = [Inf;Inf;1]; % enforces x(3) is binary
x0 = [];
Specify no display.
options = optimoptions('intlinprog','Display','off');
Run the solver.
x = intlinprog(f,intcon,A,b,Aeq,beq,lb,ub,x0,options)
x = 3×1
0
6
0
Solve MILP Using Problem-Based Approach
This example shows how to set up a problem using the problem-based approach and then solve it using the solver-based approach. The problem is
Create an OptimizationProblem
object named prob
to represent this problem. To specify a binary variable, create an optimization variable with integer type, a lower bound of 0, and an upper bound of 1.
x = optimvar('x',2,'LowerBound',0); xb = optimvar('xb','LowerBound',0,'UpperBound',1,'Type','integer'); prob = optimproblem('Objective',-3*x(1)-2*x(2)-xb); cons1 = sum(x) + xb <= 7; cons2 = 4*x(1) + 2*x(2) + xb == 12; prob.Constraints.cons1 = cons1; prob.Constraints.cons2 = cons2;
Convert the problem object to a problem structure.
problem = prob2struct(prob);
Solve the resulting problem structure.
[sol,fval,exitflag,output] = intlinprog(problem)
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [1e+00, 4e+00] Cost [1e+00, 3e+00] Bound [1e+00, 1e+00] RHS [7e+00, 1e+01] Presolving model 2 rows, 3 cols, 6 nonzeros 0s 0 rows, 0 cols, 0 nonzeros 0s Presolve: Optimal Solving report Status Optimal Primal bound -12 Dual bound -12 Gap 0% (tolerance: 0.01%) Solution status feasible -12 (objective) 0 (bound viol.) 0 (int. viol.) 0 (row viol.) Timing 0.00 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 0 LP iterations 0 (total) 0 (strong br.) 0 (separation) 0 (heuristics) Optimal solution found. Intlinprog stopped at the root node because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
sol = 3×1
0
6
0
fval = -12
exitflag = 1
output = struct with fields:
relativegap: 0
absolutegap: 0
numfeaspoints: 0
numnodes: 0
constrviolation: 0
algorithm: 'highs'
message: 'Optimal solution found....'
Both sol(1)
and sol(3)
are binary-valued. Which value corresponds to the binary optimization variable xb
?
prob.Variables
ans = struct with fields:
x: [2x1 optim.problemdef.OptimizationVariable]
xb: [1x1 optim.problemdef.OptimizationVariable]
The variable xb
appears last in the Variables
display, so xb
corresponds to sol(3) = 1
. See Algorithms.
Examine the MILP Solution and Process
Call intlinprog
with more outputs to see solution details and process.
The goal is to solve the problem
Specify the solver inputs.
f = [-3;-2;-1];
intcon = 3;
A = [1,1,1];
b = 7;
Aeq = [4,2,1];
beq = 12;
lb = zeros(3,1);
ub = [Inf;Inf;1]; % enforces x(3) is binary
Call intlinprog
with all outputs.
[x,fval,exitflag,output] = intlinprog(f,intcon,A,b,Aeq,beq,lb,ub)
Running HiGHS 1.7.0: Copyright (c) 2024 HiGHS under MIT licence terms Coefficient ranges: Matrix [1e+00, 4e+00] Cost [1e+00, 3e+00] Bound [1e+00, 1e+00] RHS [7e+00, 1e+01] Presolving model 2 rows, 3 cols, 6 nonzeros 0s 0 rows, 0 cols, 0 nonzeros 0s Presolve: Optimal Solving report Status Optimal Primal bound -12 Dual bound -12 Gap 0% (tolerance: 0.01%) Solution status feasible -12 (objective) 0 (bound viol.) 0 (int. viol.) 0 (row viol.) Timing 0.00 (total) 0.00 (presolve) 0.00 (postsolve) Nodes 0 LP iterations 0 (total) 0 (strong br.) 0 (separation) 0 (heuristics) Optimal solution found. Intlinprog stopped at the root node because the objective value is within a gap tolerance of the optimal value, options.AbsoluteGapTolerance = 1e-06. The intcon variables are integer within tolerance, options.ConstraintTolerance = 1e-06.
x = 3×1
0
6
0
fval = -12
exitflag = 1
output = struct with fields:
relativegap: 0
absolutegap: 0
numfeaspoints: 0
numnodes: 0
constrviolation: 0
algorithm: 'highs'
message: 'Optimal solution found....'
The output structure shows numnodes
is 0
. This means intlinprog
solved the problem before branching. This is one indication that the result is reliable. Also, the absolutegap
and relativegap
fields are 0
. This is another indication that the result is reliable.
Input Arguments
f
— Coefficient vector
real vector | real array
Coefficient vector, specified as a real vector or real array.
The coefficient vector represents the objective function f'*x
.
The notation assumes that f
is a column vector,
but you are free to use a row vector or array. Internally, linprog
converts f
to
the column vector f(:)
.
If you specify f = []
, intlinprog
tries
to find a feasible point without trying to minimize an objective function.
Example: f = [4;2;-1.7];
Data Types: double
intcon
— Vector of integer constraints
vector of integers
Vector of integer constraints, specified as a vector of positive
integers. The values in intcon
indicate the components
of the decision variable x
that are integer-valued. intcon
has
values from 1
through numel(f)
.
intcon
can also be an array. Internally, intlinprog
converts
an array intcon
to the vector intcon(:)
.
Example: intcon = [1,2,7]
means x(1)
, x(2)
,
and x(7)
take only integer values.
Data Types: double
A
— Linear inequality constraints
real matrix
Linear inequality constraints, specified as a real matrix. A
is an M
-by-N
matrix, where M
is the number of inequalities, and N
is the number of variables (length of f
). For large problems, pass A
as a sparse matrix.
A
encodes the M
linear inequalities
A*x <= b
,
where x
is the column vector of N
variables x(:)
, and b
is a column vector with M
elements.
For example, consider these inequalities:
x1 +
2x2 ≤
10
3x1 +
4x2 ≤
20
5x1 +
6x2 ≤ 30.
Specify the inequalities by entering the following constraints.
A = [1,2;3,4;5,6]; b = [10;20;30];
Example: To specify that the x-components add up to 1 or less, take A =
ones(1,N)
and b = 1
.
Data Types: double
b
— Linear inequality constraints
real vector
Linear inequality constraints, specified as a real vector. b
is
an M
-element vector related to the A
matrix.
If you pass b
as a row vector, solvers internally
convert b
to the column vector b(:)
.
For large problems, pass b
as a sparse vector.
b
encodes the M
linear
inequalities
A*x <= b
,
where x
is the column vector of N
variables x(:)
,
and A
is a matrix of size M
-by-N
.
For example, consider these inequalities:
x1
+ 2x2 ≤
10
3x1
+ 4x2 ≤
20
5x1
+ 6x2 ≤
30.
Specify the inequalities by entering the following constraints.
A = [1,2;3,4;5,6]; b = [10;20;30];
Example: To specify that the x components sum to 1 or less, use A =
ones(1,N)
and b = 1
.
Data Types: single
| double
Aeq
— Linear equality constraints
real matrix
Linear equality constraints, specified as a real matrix. Aeq
is an Me
-by-N
matrix, where Me
is the number of equalities, and N
is the number of variables (length of f
). For large problems, pass Aeq
as a sparse matrix.
Aeq
encodes the Me
linear equalities
Aeq*x = beq
,
where x
is the column vector of N
variables x(:)
, and beq
is a column vector with Me
elements.
For example, consider these equalities:
x1 +
2x2 +
3x3 =
10
2x1 +
4x2 +
x3 = 20.
Specify the equalities by entering the following constraints.
Aeq = [1,2,3;2,4,1]; beq = [10;20];
Example: To specify that the x-components sum to 1, take Aeq = ones(1,N)
and
beq = 1
.
Data Types: double
beq
— Linear equality constraints
real vector
Linear equality constraints, specified as a real vector. beq
is
an Me
-element vector related to the Aeq
matrix.
If you pass beq
as a row vector, solvers internally
convert beq
to the column vector beq(:)
.
For large problems, pass beq
as a sparse vector.
beq
encodes the Me
linear
equalities
Aeq*x = beq
,
where x
is the column vector of N
variables
x(:)
, and Aeq
is a matrix of size
Me
-by-N
.
For example, consider these equalities:
x1
+ 2x2 +
3x3 =
10
2x1
+ 4x2 +
x3 =
20.
Specify the equalities by entering the following constraints.
Aeq = [1,2,3;2,4,1]; beq = [10;20];
Example: To specify that the x components sum to 1, use Aeq = ones(1,N)
and
beq = 1
.
Data Types: single
| double
lb
— Lower bounds
[]
(default) | real vector or array
Lower bounds, specified as a vector or array of doubles. lb
represents
the lower bounds elementwise in lb
≤ x
≤ ub
.
Internally, intlinprog
converts an array lb
to
the vector lb(:)
.
Example: lb = [0;-Inf;4]
means x(1)
≥ 0
, x(3) ≥ 4
.
Data Types: double
ub
— Upper bounds
[]
(default) | real vector or array
Upper bounds, specified as a vector or array of doubles. ub
represents
the upper bounds elementwise in lb
≤ x
≤ ub
.
Internally, intlinprog
converts an array ub
to
the vector ub(:)
.
Example: ub = [Inf;4;10]
means x(2)
≤ 4
, x(3) ≤ 10
.
Data Types: double
x0
— Initial point
[]
(default) | real array
Initial point, specified as a real array. The number of elements in
x0
is the same as the number of elements of
f
, when f
exists.
Otherwise, the number is the same as the number of columns of
A
or Aeq
. Internally, the
solver converts an array x0
into a vector
x0(:)
.
Providing x0
can change the amount of time
intlinprog
takes to converge. It is difficult
to predict how x0
affects the solver. For
suggestions on using appropriate Heuristics
with
x0
, see Tips.
For the "legacy"
algorithm, x0
must be feasible with respect to all constraints. If
x0
is not feasible, the solver warns and
ignores x0
. If you do not have a feasible
x0
for this algorithm, set x0 =
[]
.
The "highs"
algorithm attempts to use any supplied
x0
, modifying the point if necessary for
feasibility.
Example: x0 = 100*rand(size(f))
Data Types: double
options
— Options for intlinprog
options created using optimoptions
Options for intlinprog
,
specified as the output of optimoptions
.
Some options are absent from the
optimoptions
display. These options appear in italics in the following
table. For details, see View Optimization Options.
All Algorithms | ||
---|---|---|
Option | Description | Default |
AbsoluteGapTolerance |
Nonnegative real.
| 1e-6 for "highs" , 0 for
"legacy" |
Algorithm | Choose the optimization algorithm:
| "highs" |
ConstraintTolerance | For the For the
| 1e-6 for "highs" , 1e-4 for
"legacy" |
Display |
Level of display (see Iterative Display):
| "iter" |
MaxFeasiblePoints | Strictly positive integer.
intlinprog stops if it finds
MaxFeasiblePoints integer
feasible points. | Inf |
MaxNodes | Strictly positive integer that is the maximum number of nodes intlinprog explores
in its branch-and-bound process. | 1e7 |
MaxTime | Nonnegative real that is the maximum time in seconds that intlinprog
runs. | 7200 |
ObjectiveCutOff | Real greater than -Inf . During the branch-and-bound
calculation, intlinprog discards any node where
the linear programming solution has an objective value exceeding ObjectiveCutOff . | Inf |
OutputFcn | One or more functions that an optimization
function calls at events. Specify as
For information on writing a custom output function, see intlinprog Output Function and Plot Function Syntax. | [] |
PlotFcn | Plots various measures of progress while
the algorithm executes; select from predefined
plots or write your own. Pass
For information on writing a custom plot function, see intlinprog Output Function and Plot Function Syntax. | [] |
RelativeGapTolerance | Real from For
the
When
For
the
Note Although you specify | 1e-4 |
Legacy Algorithm | ||
BranchRule | Rule for choosing the component for branching:
| 'reliability' |
CutGeneration | Level of cut generation (see Cut Generation):
| 'basic' |
CutMaxIterations | Number of passes through all cut generation
methods before entering the branch-and-bound
phase, an integer from 1
through 50 . Disable cut
generation by setting the
CutGeneration option to
'none' . | 10 |
Heuristics | Algorithm for searching for feasible points (see Heuristics for Finding Feasible Solutions):
| 'basic' |
HeuristicsMaxNodes | Strictly positive integer that bounds the number
of nodes intlinprog can
explore in its branch-and-bound search for
feasible points. Applies only to
'rss' and
'rins' . See Heuristics for Finding Feasible Solutions. | 50 |
IntegerPreprocess | Types of integer preprocessing (see Mixed-Integer Program Preprocessing):
| 'basic' |
IntegerTolerance | Real from 1e-10 through
1e-3 , where the maximum
deviation from integer that a component of the
solution x can have and still
be considered an integer.
IntegerTolerance is not a
stopping criterion. | 1e-5 |
LPMaxIterations | Strictly positive integer, the maximum number of simplex algorithm iterations per node during the branch-and-bound process. |
In this
expression, |
LPOptimalityTolerance | Nonnegative real where reduced costs must exceed
LPOptimalityTolerance for a
variable to be taken into the basis. | 1e-7 |
LPPreprocess | Type of preprocessing for the solution to the relaxed linear program (see Linear Program Preprocessing):
| 'basic' |
NodeSelection | Choose the node to explore next.
| 'simplebestproj' |
ObjectiveImprovementThreshold | Nonnegative real.
intlinprog changes the
current feasible solution only when it locates
another with an objective function value that is
at least
ObjectiveImprovementThreshold
lower: (fold – fnew)/(1 + |fold|) >
ObjectiveImprovementThreshold. | 0 |
RootLPAlgorithm | Algorithm for solving linear programs:
| 'dual-simplex' |
RootLPMaxIterations | Nonnegative integer that is the maximum number of simplex algorithm iterations to solve the initial linear programming problem. |
In this
expression, |
Example: options = optimoptions("intlinprog",MaxTime=120)
problem
— Structure encapsulating inputs and options
structure
Structure encapsulating the inputs and options, specified with the following fields.
f | Vector representing objective f'*x (required) |
intcon | Vector indicating variables that take integer values (required) |
Aineq | Matrix in linear inequality constraints Aineq*x ≤ bineq |
| Vector in linear inequality constraints Aineq*x ≤ bineq |
| Matrix in linear equality constraints Aeq*x = beq |
| Vector in linear equality constraints Aeq*x = beq |
lb | Vector of lower bounds |
ub | Vector of upper bounds |
x0 | Initial point |
solver | 'intlinprog' (required) |
| Options created using optimoptions (required) |
You must specify at least these fields in the problem structure. Other fields are optional:
f
intcon
solver
options
Example: problem.f = [1,2,3];
problem.intcon
= [2,3];
problem.options = optimoptions('intlinprog');
problem.Aineq = [-3,-2,-1];
problem.bineq
= -20;
problem.lb = [-6.1,-1.2,7.3];
problem.solver
= 'intlinprog';
Data Types: struct
Output Arguments
x
— Solution
real vector
Solution, returned as a vector that minimizes f'*x
subject
to all bounds, integer constraints, and linear constraints.
When a problem is infeasible or unbounded, x
is []
.
fval
— Objective value
real scalar
Objective value, returned as the scalar value f'*x
at
the solution x
.
When a problem is infeasible or unbounded, fval
is []
.
exitflag
— Algorithm stopping condition
integer
Algorithm stopping condition, returned as an integer identifying
the reason the algorithm stopped. The following lists the values of exitflag
and
the corresponding reasons intlinprog
stopped.
| The solution is
feasible with respect to the relative
|
|
|
|
|
|
|
|
|
| No feasible point found. |
| Root LP problem is unbounded. |
| Solver lost feasibility. |
The exit message can give more detailed information on the reason intlinprog
stopped,
such as exceeding a tolerance.
Exitflags 3
and -9
relate
to solutions that have large infeasibilities. These usually arise from linear constraint
matrices that have large condition number, or problems that have large solution components. To
correct these issues, try to scale the coefficient matrices, eliminate redundant linear
constraints, or give tighter bounds on the variables.
output
— Solution process summary
structure
Solution process summary, returned as a structure containing information about the optimization process.
| Relative percentage difference between upper
(
If Note Although you specify |
| Difference between upper and lower
bounds of the objective function that If |
| Number of integer feasible points found. If
|
| Number of nodes in the branch-and-bound algorithm for the
If |
| Constraint violation that is positive for violated constraints.
|
| Algorithm used, either
'highs' or
'legacy' . |
| Exit message. |
Limitations
Often, some supposedly integer-valued components of the solution
x(intCon)
are not precisely integers.intlinprog
deems as integers all solution values withinConstraintTolerance
of an integer (IntegerTolerance
for the"legacy"
algorithm).To round all supposed integers to be exactly integers, use the
round
function.x(intcon) = round(x(intcon));
Caution
Rounding solutions can cause the solution to become infeasible. Check feasibility after rounding:
max(A*x - b) % See if entries are not too positive, so have small infeasibility max(abs(Aeq*x - beq)) % See if entries are near enough to zero max(x - ub) % Positive entries are violated bounds max(lb - x) % Positive entries are violated bounds
intlinprog
does not enforce that solution components be integer-valued when their absolute values exceed2.1e9
. When your solution has such components,intlinprog
warns you. If you receive this warning, check the solution to see whether supposedly integer-valued components of the solution are close to integers.intlinprog
does not allow components of the problem, such as coefficients inf
,b
, orub
, to exceed1e20
in absolute value (1e25
for the"legacy"
algorithm), and does not allow the linear constraint matricesA
andAeq
to exceed or equal1e15
in absolute value. If you try to runintlinprog
with such a problem,intlinprog
issues an error.
More About
Enhanced Exit Messages
The next few items list the possible enhanced exit messages from
intlinprog
. Enhanced exit messages give a link for more
information as the first sentence of the message.
Solver stopped prematurely. Integer feasible point found.
intlinprog
did not
necessarily find an optimal solution. However, it did find at least one integer
feasible point. An integer feasible point is a point that satisfies all
constraints, including bounds, linear constraints, and integer
constraints.
Reached the maximum number of nodes
intlinprog
uses a branch-and-bound algorithm, whose
details are in Branch and Bound. Each branch in the
algorithm creates a new node. intlinprog
uses the
MaxNodes
option (a tolerance) as
a stopping criterion.
You can change the value of an option using dot notation:
options.MaxNodes = 5e4;
You can also change the value using optimoptions
:
options = optimoptions(options,'MaxNodes',5e4);
Exceeded the time limit
intlinprog
uses the MaxTime
option (a
tolerance) as a
stopping criterion.
You can change the value of an option using dot notation:
options.MaxTime = 5e4;
You can also change the value using optimoptions
:
options = optimoptions(options,'MaxTime',5e4);
Exceeded the iteration limit at a node
intlinprog
exceeded the iteration limit.
intlinprog
uses the LPMaxIterations
option (a tolerance) as a
stopping criterion.
You can change the value of an option using dot notation:
options.LPMaxIterations = 5e4;
You can also change the value using optimoptions
:
options = optimoptions(options,'LPMaxIterations',5e4);
Reached the maximum number of feasible points
intlinprog
found at least
MaxNumFeasPoints
integer feasible points.
intlinprog
did not necessarily find an optimal
solution.
An integer feasible point is a point that satisfies all constraints, including bounds, linear constraints, and integer constraints.
intlinprog
uses the MaxNumFeasPoints
option (a tolerance) as a
stopping criterion.
You can change the value of an option using dot notation:
options.MaxNumFeasPoints = 5e4;
You can also change the value using optimoptions
:
options = optimoptions(options,'MaxNumFeasPoints',5e4);
Objective value is within a gap tolerance
intlinprog
internally calculates both upper and lower
bounds on the value of the objective function at the solution. The
gap between these internally-calculated bounds is
the difference between the upper and lower bounds. The relative
gap is the ratio of the gap to one plus the absolute value of
the lower bound. intlinprog
uses the
TolGapAbs
option (a tolerance on the gap) and the
TolGapRel
option (a tolerance on the relative gap) as
stopping criteria.
No integer variables specified.
The intcon
argument was empty, so
intlinprog
solved a linear programming problem.
Solver stopped prematurely. No integer feasible point found.
intlinprog
did not find
any integer feasible points, and could not proceed. This does not necessarily
mean that the problem is infeasible, only that intlinprog
failed to find an integer feasible point. An integer feasible point is a point
that satisfies all constraints, including bounds, linear constraints, and
integer constraints.
Exceeded the iteration limit while solving the root LP problem
intlinprog
was unable to solve the relaxed LP because it
reached the RootLPMaxIterations
iteration limit.
intlinprog
uses the
RootLPMaxIterations
option (a tolerance) as a
stopping criterion.
Exceeded its allocated memory
intlinprog
ran out of memory. It is possible that
reformulating the problem could lead to a solution; see Williams [1].
No Feasible Solution Found
intlinprog
stopped because there is no solution to the
problem. Either the bounds and linear constraints are inconsistent, or the
integer constraints are inconsistent with the bounds and linear
constraints.
To determine the cause, rerun the problem with intcon = []
.
If the resulting linear program has no solution, then the bounds and linear
constraints are inconsistent. Otherwise, the integer constraints in the original
problem are inconsistent with the bounds and linear constraints.
Root LP problem is unbounded
The root linear programming (LP) problem is the original MILP problem but with
no integer constraints. intlinprog
stopped because the root
LP problem is unbounded.
The original problem might be infeasible. intlinprog
does
not check feasibility when the root LP problem is unbounded.
Problem is Unbounded
intlinprog
stopped because there is no solution to the
linear programming problem. For any target value there are feasible points with
objective value smaller than the target.
Definitions for Exit Messages
The next few items contain definitions for terms in the intlinprog
exit messages.
tolerance
Generally, a tolerance is a threshold which, if crossed, stops the iterations of a solver. For more information on tolerances, see Tolerances and Stopping Criteria.
Node
A node in a branch-and-bound or branch-and-cut tree is a value of
x
, and a component j
of
x
where x(j)
has a fractional part.
The branch-and-bound node has two subproblems: evaluate the linear programming
problem with the restriction
x(j)
≥ ceil(x(j))
, and
evaluate the linear programming problem with the restriction
x(j)
≤ floor(x(j))
. For
more information, see Branch-and-Bound.
Integer Within Tolerance
intlinprog
does not guarantee that the variables you
specify in intcon
are exactly integers, only that they are
within the TolInteger
tolerance of an integer value. See
Some “Integer” Solutions Are Not Integers.
Root Node
When intlinprog
stops at the root node, it did not have to
execute any branch-and-bound search. Either intlinprog
solved the problem during presolve, or it solved the problem during subsequent
processing, but without needing to branch.
The root node is the relaxed linear programming problem based on the original MILP. For details, see Branch-and-Bound.
Tips
To specify binary variables, set the variables to be integers in
intcon
, and give them lower bounds of0
and upper bounds of1
.Save memory by specifying sparse linear constraint matrices
A
andAeq
. However, you cannot use sparse matrices forb
andbeq
.To provide logical indices for integer components, meaning a binary vector with
1
indicating an integer, convert tointcon
form usingfind
. For example,logicalindices = [1,0,0,1,1,0,0]; intcon = find(logicalindices)
intcon = 1 4 5
This tip applies to the
"legacy"
algorithm.If you include an
x0
argument,intlinprog
uses that value in the'rins'
and guided diving heuristics until it finds a better integer-feasible point. So when you providex0
, you can obtain good results by setting the'Heuristics'
option to'rins-diving'
or another setting that uses'rins'
.intlinprog
replacesbintprog
. To update oldbintprog
code to useintlinprog
, make the following changes:Set
intcon
to1:numVars
, wherenumVars
is the number of variables in your problem.Set
lb
tozeros(numVars,1)
.Set
ub
toones(numVars,1)
.Update any relevant options. Use
optimoptions
to create options forintlinprog
.Change your call to
bintprog
as follows:[x,fval,exitflag,output] = bintprog(f,A,b,Aeq,Beq,x0,options) % Change your call to: [x,fval,exitflag,output] = intlinprog(f,intcon,A,b,Aeq,Beq,lb,ub,x0,options)
Alternative Functionality
App
The Optimize Live Editor task provides a visual interface for intlinprog
.
Version History
Introduced in R2014aR2024b: "highs"
algorithm supports output functions, plot
functions, and the MaxFeasiblePoints
option
The "highs"
algorithm now supports output functions and plot
functions. For details, see intlinprog Output Function and Plot Function Syntax. For an example using a built-in plot
function, see Factory, Warehouse, Sales Allocation Model: Solver-Based.
The optimValues Structure for output functions and plot functions
now uses the field name dualbound
instead of
lowerbound
. The meaning of dualbound
is the global lower bound for minimization problems, or the global upper bound
for maximization problems. The optimplotmilp
plot function
now uses the labels Primal Bound
and Dual
Bound
instead of Branch/bound UB
and
Branch/bound LB
.
The "highs"
algorithm now supports the
MaxFeasiblePoints
option. The
output
structure for the "highs"
algorithm now supports the numfeaspoints
field.
R2024b: "legacy"
algorithm will be removed
R2024a: New "highs"
algorithm is the default
intlinprog
gains an algorithm, an
Algorithm
option, and a new default algorithm. The
"highs"
algorithm is based on the open-source HiGHS
code. Currently, the "highs"
algorithm does not support
output functions or plot functions.
To use the previous algorithm, set Algorithm="legacy"
using
optimoptions
.
The allowable range of the ConstraintTolerance
option values
is now 1e-10
to 1e-3
for both algorithms,
as is the allowable range of the IntegerTolerance
option
values for the "legacy"
algorithm.
R2019a: Default BranchRule
is 'reliability'
The default value of the BranchRule
option is
'reliability'
instead of 'maxpscost'
.
In testing, this value gave better performance on many problems, both in
solution times and in number of explored branching nodes.
On a few problems, the previous branch rule performs better. To get the
previous behavior, set the BranchRule
option to
'maxpscost'
.
See Also
linprog
| mpsread
| optimoptions
| prob2struct
| Optimize
Topics
- Mixed-Integer Linear Programming Basics: Solver-Based
- Factory, Warehouse, Sales Allocation Model: Solver-Based
- Traveling Salesman Problem: Solver-Based
- Solve Sudoku Puzzles via Integer Programming: Solver-Based
- Mixed-Integer Quadratic Programming Portfolio Optimization: Solver-Based
- Optimal Dispatch of Power Generators: Solver-Based
- Mixed-Integer Linear Programming (MILP) Algorithms
- Tuning Integer Linear Programming
- Solver-Based Optimization Problem Setup
Commande MATLAB
Vous avez cliqué sur un lien qui correspond à cette commande MATLAB :
Pour exécuter la commande, saisissez-la dans la fenêtre de commande de MATLAB. Les navigateurs web ne supportent pas les commandes MATLAB.
Select a Web Site
Choose a web site to get translated content where available and see local events and offers. Based on your location, we recommend that you select: .
You can also select a web site from the following list:
How to Get Best Site Performance
Select the China site (in Chinese or English) for best site performance. Other MathWorks country sites are not optimized for visits from your location.
Americas
- América Latina (Español)
- Canada (English)
- United States (English)
Europe
- Belgium (English)
- Denmark (English)
- Deutschland (Deutsch)
- España (Español)
- Finland (English)
- France (Français)
- Ireland (English)
- Italia (Italiano)
- Luxembourg (English)
- Netherlands (English)
- Norway (English)
- Österreich (Deutsch)
- Portugal (English)
- Sweden (English)
- Switzerland
- United Kingdom (English)