MRSortModel Class Reference

Datastructure of the model to learn. More...

#include <MRSortModel.h>

Collaboration diagram for MRSortModel:

Public Member Functions
	MRSortModel (Criteria &criteria, Profiles &profiles, Categories &categories, float lambda, std::string id="model")
	MRSortModel (int n_cat, int n_crit, std::string id="model")
	MRSortModel (const MRSortModel &mrsort)
std::string	getId () const
float	getScore () const
void	setScore (float score)
Category	categoryAssignment (std::vector< Perf > &alt, std::vector< std::vector< Perf >> &profiles_pt)
AlternativesPerformance	categoryAssignments (PerformanceTable &pt)
float	computeConcordance (std::vector< Perf > &prof, std::vector< Perf > &alt)
std::unordered_map< std::string, std::unordered_map< std::string, float > >	computeConcordanceTable (PerformanceTable &pt)

Public Attributes
Criteria	criteria
Profiles	profiles
float	lambda
Categories	categories

Friends
std::ostream &	operator<< (std::ostream &out, const MRSortModel &mrsort)

Detailed Description

Datastructure of the model to learn.

The MRSortModel is the datastructure that represents a complete modelisation of the problem to learn. The state of a model is defined by lambda, a Criteria object and a Profile object. Lambda is a float representing the global threshold of the problem, the Profile object hold the values that defined the profiles delimiting all the categories, and Criteria holds the weights of each Criterion.

The model is the "atomic" object that we use in the metaheuristic, it is the datastructure running trough the different pipeline and modified by the learning algorithms.

The performance of the model is evaluated by running the ground truth dataset into the model and looking at the category assignment.

Constructor & Destructor Documentation

MRSortModel() [1/3]

MRSortModel::MRSortModel	(	Criteria &	criteria,
		Profiles &	profiles,
		Categories &	categories,
		float	lambda,
		std::string	id = "model"
	)

MRSortModel standard constructor

Parameters

criteria	Criteria object
profiles	Profiles containing the profiles
categories	Categories object used for the profiles
lambda	threshold
id	optional name of the model

MRSortModel() [2/3]

MRSortModel::MRSortModel	(	int	n_cat,
		int	n_crit,
		std::string	id = "model"
	)

MRSortModel generator constructor. This constructor initializes the profiles, lambda and criteria weights at random.

Parameters

n_cat	number of categories
n_crit	number of criteria
id	mrsort model's id

MRSortModel() [3/3]

MRSortModel::MRSortModel

(

const MRSortModel &

mrsort

)

MRSortModel constructor by copy

Parameters

mrsort

MRSortModel object to copy

Member Function Documentation

categoryAssignment()

Category MRSortModel::categoryAssignment	(	std::vector< Perf > &	alt,
		std::vector< std::vector< Perf >> &	profiles_pt
	)

categoryAssignment assign the category given the alternative and the current state of the model.

Parameters

alt	PerfVect of the alternative
profiles_pt	PerfTable of the profiles

Returns

category_assignment Category object associated to the alternative

categoryAssignments()

AlternativesPerformance MRSortModel::categoryAssignments

(

PerformanceTable &

pt

)

categoryAssignments assign the categories given the performance table and the current state of the model.

Parameters

pt	PerformanceTable

Returns

category_assignment AlternativeAssignment object containing the pt given in args and the category assignment.

computeConcordance()

float MRSortModel::computeConcordance	(	std::vector< Perf > &	prof,
		std::vector< Perf > &	alt
	)

computeConcordance computes the concordance value between a profile and an alternative

Parameters

prof	profile
alt	alternative

Returns

concordance value

computeConcordanceTable()

std::unordered_map<std::string, std::unordered_map<std::string, float> > MRSortModel::computeConcordanceTable

(

PerformanceTable &

pt

)

computeConcordanceTable computes the concordance table of a performance table

Parameters

pt	PerformanceTable

Returns

concordance table giving the concordance value for each profile, alternative tupple, displayed in a map { prof_id1 : {alt_id1 : c1, alt_id2: c2, ...}, ...}

getId()

std::string MRSortModel::getId

(

)

const

getId getter of id parameter

Returns

id

getScore()

float MRSortModel::getScore

(

)

const

getScore getter of score parameter

Returns

score

setScore()

void MRSortModel::setScore

(

float

score

)

setScore setter of score parameter

Parameters

score

new score

---

The documentation for this class was generated from the following file:

• /Users/ThomasVindard 1/Desktop/3A/Projet/fastPL/include/types/MRSortModel.h