`src`: solver API

This module solves vehicle routing problems using branch&cut&price methods

`Model`()	Define a routing model.
`Point`(id[, name, id_customer, ...])	Define a point of graph (customer or depot).
`Customer`(id[, name, id_customer, penalty, ...])	Define a point customer of graph.
`Depot`(id[, name, cost, service_time, ...])	Define a point depot of graph.
`Link`(start_point_id, end_point_id[, name, ...])	Define a link of graph.
`VehicleType`(id[, start_point_id, ...])	Define a vehicle type with different attributes.
`Parameters`([time_limit, upper_bound, ...])	Define all parameters from model
`Solution`([json_input, status])	Contains all elements of solution after running model.solve().
`Route`(json_input)	Define a route from solution

Model

class src.solver.Model

Define a routing model.

add_customer(id, name='', id_customer=0, service_time=0.0, penalty=0.0, tw_begin=0.0, tw_end=0.0, demand=0, incompatible_vehicles=[]): Add customer in dictionary points

add_depot(id, name='', service_time=0.0, cost=0.0, tw_begin=0.0, tw_end=0.0, incompatible_vehicles=[]): Add depot in dictionary points

add_link(start_point_id=0, end_point_id=0, name='', is_directed=False, distance=0.0, time=0.0, fixed_cost=0.0): Add Link in dictionary links

add_point(id, name='', id_customer=0, service_time=0.0, penalty_or_cost=0.0, tw_begin=0.0, tw_end=0.0, demand=0, incompatible_vehicles=[]): Add Point in dictionary points, if we want to add Depot, id_customer must be equal to 0, otherwise it cannot be greater than 1022 for a Customer

add_vehicle_type(id: int, start_point_id=-1, end_point_id=-1, name='', capacity=0, fixed_cost=0.0, var_cost_dist=0.0, var_cost_time=0.0, max_number=1, tw_begin=0.0, tw_end=0.0): Add VehicleType in dictionary vehicle_types

check_depots(): Update the model if there are defined intermediate depots not used by vehicles

delete_customer(id: int): Delete a customer by giving his id

delete_depot(id: int): Delete a depot by giving his id

delete_link(start_point_id: int, end_point_id: int): Delete a link by giving start point id and end point id

delete_vehicle_type(id: int): Delete a vehicle type by giving his id

export(name='instance', all_elements=False): Export the model for debugging model, we can specify the file name. If you put all_elements to True, it exports the model with preprocessing elements.

property links: contains the set of links

property max_total_vehicles_number: the maximum total vehicles number

property message: int : indicates the message associated with status

property parameters: getter function of parameters

property points

contains the set of customers and depots

Type:

PointsDict : dictionary contains only points

Informations:

It’s possible to resolve the problem until 1022 points.
For the moment, the capacity of depot is not considered

set_json(): Set model in json format with all elements of model

set_parameters(time_limit=300, upper_bound=1000000, heuristic_used=False, time_limit_heuristic=20, config_file='', solver_name='CLP', print_level=-1, action='solve', cplex_path=''): Set parameters of the model. For more advanced parameters please indicates a configuration file on config_file variable. solver_name : [CLP,CPLEX] action : [solve,enumAllFeasibleRoutes], print_level = [-2,-1,0]

solve()

Solve the routing problem by using the shared library bapcod.

Additional informations:: VRPSolverEasy is compatible with Windows 64x, Linux and macOS only

property status: int : indicates the status of solution

property vehicle_types

contains the set of vehicle types

Type:

VehicleTypesDict : dictionary contains only vehicle type

Informations:

For academic version, with CLP it’s possible to resolve the problem with only one vehicle type.

Ids of vehicle types must be greater than one

After you finish solving, you can find the status of the model and know whether you have reached optimality or not using the status attribute.

model.status
Status	numerical value
`INFEASIBLE`	-2
`INTERRUPTED_BY_ERROR`	-1
`OPTIMAL_SOL_FOUND`	0
`BETTER_SOL_FOUND`	1
`BETTER_SOL_DOES_NOT_EXISTS`	2
`BETTER_SOL_NOT_FOUND`	3

If you change the parameters to run an enumeration of all feasible solutions, the interpretation of status is different :

model.status
Status	numerical value
`ENUMERATION_SUCCEEDED`	0
`ENUMERATION_NOT_SUCCEEDED`	1
`ENUMERATION_INFEASIBLE`	2

Point

class src.solver.Point(id, name='', id_customer=0, penalty_or_cost=0.0, service_time=0, tw_begin=0, tw_end=0, demand=0, incompatible_vehicles=[])

Define a point of graph (customer or depot).

Additional informations:

service_time : It can represent the time of loading or unloading.
penalty_or_cost : If the point is a customer we can specify a penalty for not visiting the customer otherwise, if the point is a depot we can specify a cost to using the depot.
tw_begin : time windows begin
tw_end : time windows end
incompatible_vehicles : id of vehicles that cannot serve the customer or are not accepted in a depot.

__init__(id, name='', id_customer=0, penalty_or_cost=0.0, service_time=0, tw_begin=0, tw_end=0, demand=0, incompatible_vehicles=[])

__repr__(): Return repr(self).

__weakref__: list of weak references to the object (if defined)

property cost: getter function of cost

property demand: getter function of demand

get_point(debug=False): Get all components of a Point which are different of default value

property id: getter function of id

property id_customer: getter function of id customer

property incompatible_vehicles: getter function of incompatible_vehicles

property name: getter function of name

property penalty: getter function of penalty

property service_time: getter function of service_time

property time_windows: getter function of time windows

property tw_begin: getter function of time windows begin

property tw_end: getter function of time windows end

Customer

class src.solver.Customer(id, name='', id_customer=0, penalty=0.0, service_time=0, tw_begin=0, tw_end=0, demand=0, incompatible_vehicles=[])

Define a point customer of graph.

Additional informations:

id_customer(int): must be inferior or equal to 1022
penalty(float): represents the penalty of non visited customer
tw_begin(float): time window begin
tw_end(float): time window end
demand(int): must be an integer

__init__(id, name='', id_customer=0, penalty=0.0, service_time=0, tw_begin=0, tw_end=0, demand=0, incompatible_vehicles=[])

Depot

class src.solver.Depot(id, name='', cost=0, service_time=0, tw_begin=0, tw_end=0, incompatible_vehicles=[])

Define a point depot of graph.

Additional informations:: capacity: must be an integer

__init__(id, name='', cost=0, service_time=0, tw_begin=0, tw_end=0, incompatible_vehicles=[])

Link

class src.solver.Link(start_point_id, end_point_id, name='', is_directed=False, distance=0.0, time=0.0, fixed_cost=0.0)

Define a link of graph.

Additional informations:: is_directed – it’s equal to True if we cannot return at start point with the same time and distance

__init__(start_point_id, end_point_id, name='', is_directed=False, distance=0.0, time=0.0, fixed_cost=0.0)

__repr__(): Return repr(self).

__weakref__: list of weak references to the object (if defined)

property distance: getter function of distance

property end_point_id: getter function of end_point_id

property fixed_cost: getter function of fixed_cost

get_link(debug=False): Get all components of a Link which are different of default value

property is_directed: getter function of is_directed

property name: getter function of name

property start_point_id: getter function of start_point_id

property time: getter function of time

VehicleType

class src.solver.VehicleType(id: int, start_point_id=-1, end_point_id=-1, name='', capacity=0, fixed_cost=0.0, var_cost_dist=0.0, var_cost_time=0.0, max_number=1, tw_begin=0, tw_end=0)

Define a vehicle type with different attributes.

__init__(id: int, start_point_id=-1, end_point_id=-1, name='', capacity=0, fixed_cost=0.0, var_cost_dist=0.0, var_cost_time=0.0, max_number=1, tw_begin=0, tw_end=0)

__repr__(): Return repr(self).

__weakref__: list of weak references to the object (if defined)

property capacity: –int : capacity of vehicle type

property end_point_id: the id of the end point

property fixed_cost: float : fixed cost of using vehicle type

get_vehicle_type(debug=False): Get all components of a vehicle type which are differents of default value

property id: int : cannot be less than 1

property max_number: number of vehicles available

property name: str : name of vehicle type

property start_point_id: the id of the starting point

property tw_begin: time windows begin of vehicle type

property tw_end: time windows end of vehicle type

property var_cost_dist: variable cost per unit of distance :type var_cost_dist : float :

property var_cost_time: variable cost per unit of time

Parameters

class src.solver.Parameters(time_limit=300.0, upper_bound=1000000, heuristic_used=False, time_limit_heuristic=20.0, config_file='', solver_name='CLP', print_level=-1, action='solve', cplex_path='')

Define all parameters from model

__init__(time_limit=300.0, upper_bound=1000000, heuristic_used=False, time_limit_heuristic=20.0, config_file='', solver_name='CLP', print_level=-1, action='solve', cplex_path='')

__repr__(): Return repr(self).

__weakref__: list of weak references to the object (if defined)

property action: str : indicates if we want to solve the problem (“solve”) or enumerate all feasible routes(“enumAllFeasibleRoutes”)

property config_file: str : indicates the path of the config file for more advanced settings

property cplex_path: str : path of library cplex 22.1. You can specify a path if you want to use cplex and replace the bapcod-shared library by the library using cplex

get_parameters(debug=False): Get all parameters which are different of default value

property heuristic_used: bool : getter function of heuristic_used

property print_level: indicates the level of print from Bapcod during the resolution, we can choose (-2,-1,0)

property solver_name: str : indicates the solver used during the resolution, we can choose “CLP” or “CPLEX” solver

property time_limit: float : represents the limit time of resolution

property time_limit_heuristic: float : getter function of time_limit_heuristic

property upper_bound: float : represents the cost we want to reach

Solution

class src.solver.Solution(json_input=None, status=-23)

Contains all elements of solution after running model.solve().

__init__(json_input=None, status=-23)

__repr__(): Return repr(self).

__str__(): Return str(self).

__weakref__: list of weak references to the object (if defined)

export(name='instance'): Export solution for sharing or debugging model, we can specify the name of the file

is_defined(): return true if a solution is defined

property json: str : formatted string to print statistics in json format

property routes: list(Route) : contains the set of routes

property value: float : return the total cost of the solution

Statistics

class src.solver.Statistics(json_input='')

Define a statistics from solution

__init__(json_input='')

__repr__(): Return repr(self).

__weakref__: list of weak references to the object (if defined)

property best_lb: float : best lower bound find during the resolution

property json_input: str : formatted string to print statistics in json format

property nb_branch_and_bound_nodes: int : number of branch and bound nodes from tree structure.

property root_lb: float : the root lower bound find during the resolution

property root_time: float : getter function of root_time

property solution_time: float : time computed to find the solution

Route

class src.solver.Route(json_input)

Define a route from solution

__init__(json_input)

__repr__(): Return repr(self).

__str__(): Return str(self).

__weakref__: list of weak references to the object (if defined)

property cap_consumption: list(float) : the loads at each point

property incoming_arc_names: list(str) : the names of incoming arc

property point_ids: list(int) : if of each point visited

property point_names: list(str) : names of visited points

property route: str : formatted string to print route in json format

property route_cost: float : cost incurred by variable and fixed costs

property time_consumption: list(float) : the time at each point

property vehicle_type_id: int : id of vehicle type making the trip

src: solver API