RAFF.jl: Robust Algebraic Fitting Function in Julia
Emerson Vitor Castelani, Ronaldo Lopes, Wesley Vagner Inês Shirabayashi, F. N. C. Sobral · The Journal of Open Source Software · 2019
RAFF.jl is a Julia package for the adjustment of a function to a dataset coming from some experiment.This package is an alternative to classical adjustment techniques such as linear and nonlinear regression.The goal of this package is to find robust adjustments free from the influence of possible outliers (discrepant points of the adjustment). MotivationLet f : R n → R be a function whose mathematical description is not available.This function can be, for example, a black-box, a proprietary computer program or an experiment.Suppose that a dataset S = {(x 1 , y 1 ), . . ., (x m , y m )} is available, where y i is an approximation of f (x i ) (from an experimental procedure, numerical approximation, etc.) and we want to approximate f by a known model ϕ.Model ϕ can be defined as ϕ(x, θ), where x are the n independent variables of f and θ represents some parameters of ϕ.RAFF.jl (Robust Algebraic Fitting Function) is a Julia package developed to find parameters θ for ϕ in order to adjust it to the observed values S of the unknown function f .Following Liu & Wang (2008) and Keleş (2018), in general, the adjustment can be related to