alns

Documentation

Overview

Package alns defines interfaces for Adaptive Large Neighborhood Search.

Index

• type ProgressionEntry
• type Progressioner
• type Solution
• type Solutions
  • func (solutions Solutions[T]) Last() T
• type Solver

Types

type ProgressionEntry

type ProgressionEntry struct {
	ElapsedSeconds float64 `json:"elapsed_seconds"`
	Value          float64 `json:"value"`
}

ProgressionEntry is a single entry in the progression of the solver.

type Progressioner

type Progressioner interface {
	// Progression returns the progression of the solver.
	Progression() []ProgressionEntry
}

Progressioner is an interface that can be implemented by a solver to indicate that is can return the progression of the solver.

type Solution

type Solution[T any] interface {
	// Copy returns a copy of the solution which must be of type T the
	// `derived` type. This copy must be a deep copy.
	Copy() T

	// Score returns the score of the solution.
	Score() float64

	// Random returns the random number generator used by the solution.
	Random() *rand.Rand
}

Solution is a solution to a problem. It defines the minimum interface that a solution must implement to be used in Adaptive Local Neighborhood Search.

type Solutions

type Solutions[T any] <-chan T

Solutions is a channel of solutions.

func (Solutions[T]) Last

func (solutions Solutions[T]) Last() T

Last returns the last solution in the channel.

type Solver

type Solver[T Solution[T], Options any] interface {
	// Solve starts the solving process using the given options. It returns the
	// solutions as a channel.
	Solve(context.Context, Options, ...T) (Solutions[T], error)
}

Solver is an interface that can be implemented by a struct to indicate that it can solve a problem.