docplex.mp.relaxer module

class docplex.mp.relaxer.FunctionalPrioritizer(fn, override=False)

Bases: Prioritizer

class docplex.mp.relaxer.MappingPrioritizer(priority_mapping, default_priority=Priority.MANDATORY, override=False)

Bases: Prioritizer

class docplex.mp.relaxer.MatchNamePrioritizer(priority_for_unnamed=Priority.MANDATORY, priority_for_non_matches=Priority.MANDATORY, case_sensitive=False, override=False)

Bases: Prioritizer

get_priority(ct)

Looks for known priority names inside constraint names.

class docplex.mp.relaxer.NamedPrioritizer(priority=Priority.MEDIUM, override=False)

Bases: Prioritizer

class docplex.mp.relaxer.Prioritizer(override=False)

Bases: object

Abstract base class for prioritizers. This class acts like a functor to be called on each model constraint.

class docplex.mp.relaxer.Relaxer(prioritizer='all', verbose=False, precision=1e-05, override=False, **kwargs)

Bases: PublishResultAsDf, object

This class is an abstract algorithm, in the sense that it operates on interfaces.

It takes a prioritizer, which an implementation of ConstraintPrioritizer. For convenience, predefined prioritizer types are accessible through names:

• all relaxes all constraints using a MEDIUM priority; this is the default.

• named relaxes all constraints with a user name but not the others.

• match looks for priority names within constraint names; unnamed constraints are not relaxed.

Note

All predefined prioritizers apply various forms of logic, but, when a constraint has been assigned a priority by the user, this priority is always used. For example, the named prioritizer relaxes all named constraints with MEDIUM, but if an unnamed constraint was assigned a HIGH priority, then HIGH will be used.

See also

docplex.mp.priority.Priority

get_relaxation(ct)

Returns the infeasibility computed for this constraint.

Parameters:

ct – A constraint.

Returns:

The amount by which the constraint has been relaxed by the relaxer. The method returns 0 if the constraint has not been relaxed.

is_relaxed(ct)

Returns true if the constraint ct has been relaxed

Parameters:

ct – The constraint to check.

Returns:

True if the constraint has been relaxed, else False.

iter_relaxations()

Iterates on relaxations.

Relaxations are built as a dictionary with constraints as keys and numeric violations as values, so this iterator returns (ct, violation) pairs.

property number_of_relaxations

This property returns the number of relaxations found.

relax(mdl, relax_mode=None, **kwargs)

Runs the relaxation loop.

Parameters:

• mdl – The model to be relaxed.

• relax_mode – the relaxation mode. Accept either None (in which case the default mode is used, or an instance of RelaxationMode enumerated type, or a string that can be translated to a relaxation mode.

• kwargs – Accepts named arguments similar to solve.

Returns:

If the relaxation succeeds, the method returns a solution object, an instance of SolveSolution; otherwise returns None.

See also

docplex.mp.model.Model.solve(), docplex.mp.solution.SolveSolution, docplex.mp.constants.RelaxationMode

relaxations()

Returns a dictionary with all relaxed constraints.

Returns:

A dictionary where the keys are the relaxed constraints,

and the values are the numerical slacks.

property relaxed_objective_value

Returns the objective value of the relaxed solution.

Raises:

DOCplexException – If the relaxation has not been successful.

class docplex.mp.relaxer.TOutputTables(Constraint, Priority, Amount)

Bases: tuple

Amount

Alias for field number 2

Constraint

Alias for field number 0

Priority

Alias for field number 1

class docplex.mp.relaxer.UniformPrioritizer(priority=Priority.MEDIUM, override=False)

Bases: Prioritizer