Copyright	(c) Fabricio Olivetti de Franca 2020
License	GPL-3
Maintainer	fabricio.olivetti@gmail.com
Stability	experimental
Portability	POSIX
Safe Haskell	None

IT.Algorithms

Description

Definitions for a Solution, Population of solutions, fitness function and mutation function.

Synopsis

data Solution = Sol {
- _expr :: Expr
- _fit :: [Double]
- _constr :: Double
- _len :: Int
- _penalty :: Double
- _weights :: [Vector]
}
type Population = [Solution]
type Fitness = Expr -> Maybe Solution
type Constraint = Expr -> [Double] -> Double
type Mutation = Expr -> Rnd Expr

Documentation

data Solution Source #

data type containing a solution, its fitness and weight vector a refers to the type of Expr, b refers to a container of statistics.

Constructors

Sol

Fields

_expr :: Expr
The IT expression of type a
_fit :: [Double]
Fitness and other measures for evaluating the expression
_constr :: Double
Amount of Shape Constraint violation associated with the expression, always positive
_len :: Int
Expression size as per https://github.com/EpistasisLab/regression-benchmark/blob/dev/CONTRIBUTING.md
_penalty :: Double
penalty of fitness
_weights :: [Vector]
Weights associated with the expression (they count towards the length)

Instances

Instances details

Eq Solution Source #	These instances are only to find the best and worst individuals of a population.
Instance details Defined in IT.Algorithms Methods (==) :: Solution -> Solution -> Bool (/=) :: Solution -> Solution -> Bool
Ord Solution Source #
Instance details Defined in IT.Algorithms Methods compare :: Solution -> Solution -> Ordering (<) :: Solution -> Solution -> Bool (<=) :: Solution -> Solution -> Bool (>) :: Solution -> Solution -> Bool (>=) :: Solution -> Solution -> Bool max :: Solution -> Solution -> Solution min :: Solution -> Solution -> Solution
Show Solution Source #
Instance details Defined in IT.Algorithms Methods showsPrec :: Int -> Solution -> ShowS show :: Solution -> String showList :: [Solution] -> ShowS
NFData Solution Source #
Instance details Defined in IT.Algorithms Methods rnf :: Solution -> ()

type Population = [Solution] Source #

A population of 'Solution a b'

type Fitness = Expr -> Maybe Solution Source #

Fitness function that takes a list of expressions and returns an evaluated population. This function is a good candidate for parallelization. type Fitness a b = [Expr a] -> Population a b -- (Expr a, Double, b)

type Constraint = Expr -> [Double] -> Double Source #

Constraint is a function that receives an expression and its coefficients and return the penalty associated with the constraint violation.

type Mutation = Expr -> Rnd Expr Source #

Mutation function with signature 'Solution a b -> Rnd (Solution a b)'