Contents

11. Components

11.1. miplearn.components.primal.actions

class miplearn.components.primal.actions.EnforceProximity(tol: float)

Bases: miplearn.components.primal.actions.PrimalComponentAction

perform(model: miplearn.solvers.abstract.AbstractModel, var_names: numpy.ndarray, var_values: numpy.ndarray, stats: Optional[Dict]) → None
class miplearn.components.primal.actions.FixVariables

Bases: miplearn.components.primal.actions.PrimalComponentAction

perform(model: miplearn.solvers.abstract.AbstractModel, var_names: numpy.ndarray, var_values: numpy.ndarray, stats: Optional[Dict]) → None
class miplearn.components.primal.actions.PrimalComponentAction

Bases: abc.ABC

abstract perform(model: miplearn.solvers.abstract.AbstractModel, var_names: numpy.ndarray, var_values: numpy.ndarray, stats: Optional[Dict]) → None
class miplearn.components.primal.actions.SetWarmStart

Bases: miplearn.components.primal.actions.PrimalComponentAction

perform(model: miplearn.solvers.abstract.AbstractModel, var_names: numpy.ndarray, var_values: numpy.ndarray, stats: Optional[Dict]) → None

11.2. miplearn.components.primal.expert

class miplearn.components.primal.expert.ExpertPrimalComponent(action: miplearn.components.primal.actions.PrimalComponentAction)

Bases: object

before_mip(test_h5: str, model: miplearn.solvers.abstract.AbstractModel, stats: Dict[str, Any]) → None
fit(train_h5: List[str]) → None

11.3. miplearn.components.primal.indep

class miplearn.components.primal.indep.IndependentVarsPrimalComponent(base_clf: Any, extractor: miplearn.extractors.abstract.FeaturesExtractor, action: miplearn.components.primal.actions.PrimalComponentAction, clone_fn: Callable[[Any], Any] = <function clone>)

Bases: object

before_mip(test_h5: str, model: miplearn.solvers.abstract.AbstractModel, stats: Dict[str, Any]) → None
fit(train_h5: List[str]) → None

11.4. miplearn.components.primal.joint

class miplearn.components.primal.joint.JointVarsPrimalComponent(clf: Any, extractor: miplearn.extractors.abstract.FeaturesExtractor, action: miplearn.components.primal.actions.PrimalComponentAction)

Bases: object

before_mip(test_h5: str, model: miplearn.solvers.abstract.AbstractModel, stats: Dict[str, Any]) → None
fit(train_h5: List[str]) → None

11.5. miplearn.components.primal.mem

class miplearn.components.primal.mem.MemorizingPrimalComponent(clf: Any, extractor: miplearn.extractors.abstract.FeaturesExtractor, constructor: miplearn.components.primal.mem.SolutionConstructor, action: miplearn.components.primal.actions.PrimalComponentAction)

Bases: object

Component that memorizes all solutions seen during training, then fits a single classifier to predict which of the memorized solutions should be provided to the solver. Optionally combines multiple memorized solutions into a single, partial one.

before_mip(test_h5: str, model: miplearn.solvers.abstract.AbstractModel, stats: Dict[str, Any]) → None
fit(train_h5: List[str]) → None
class miplearn.components.primal.mem.MergeTopSolutions(k: int, thresholds: List[float])

Bases: miplearn.components.primal.mem.SolutionConstructor

Warm start construction strategy that first selects the top k solutions, then merges them into a single solution.

To merge the solutions, the strategy first computes the mean optimal value of each decision variable, then: (i) sets the variable to zero if the mean is below thresholds[0]; (ii) sets the variable to one if the mean is above thresholds[1]; (iii) leaves the variable free otherwise.

construct(y_proba: numpy.ndarray, solutions: numpy.ndarray) → numpy.ndarray
class miplearn.components.primal.mem.SelectTopSolutions(k: int)

Bases: miplearn.components.primal.mem.SolutionConstructor

Warm start construction strategy that selects and returns the top k solutions.

construct(y_proba: numpy.ndarray, solutions: numpy.ndarray) → numpy.ndarray
class miplearn.components.primal.mem.SolutionConstructor

Bases: abc.ABC

abstract construct(y_proba: numpy.ndarray, solutions: numpy.ndarray) → numpy.ndarray
10. Collectors & Extractors 12. Solvers