This page is devoted to document and make easier the use of RAFF- Robust Algebraic Fitting Function. Our intent is to provide a package to determine fitting functions for a dataset with ability to detect possible outliers of the dataset. All the code was made in Julia language, version 1.0.

This package is not an implentation of classical least squares solvers. It is an optimization-based package, based on algorithms for Lower Order-Value Optimization (LOVO) which were introduced in [1] and revisited in [2] to fit the user-provided models to experimental data. Recently, a good review can be found in [3]. To find possible outliers, LOVO methods depend on the number of outliers as input information. RAFF differs in this point and has no dependence on this number of outliers to perform the fitting process. In order to find a robust adjustment, a voting system is used, which is also responsible for the detection of possible outliers.

Current Status

The current status of this project is beta quality, don't use for anything important. We provide support to serial and parallel running.

Developed by

This project was developed by the optimization group at Department of Mathematics, State University of Maringá, Brazil.

The authors of this package were sponsored by Fundação Araucária, project number 002/17 - 47223.


[1] Andreani, R., Dunder, C. & Martínez, J.M. Math Meth Oper Res (2005) 61: 365. https://doi.org/10.1007/s001860400410

[2] Andreani, R., Martínez, J.M., Martínez, L. et al. J Glob Optim (2009) 43: 1. https://doi.org/10.1007/s10898-008-9280-3

[3] Martínez, J.M. TOP (2012) 20: 75. https://doi.org/10.1007/s11750-010-0169-1

Citing this package

If you would like to cite this package, please use

Castelani, E. V., Lopes, R., Shirabayashi, W., & Sobral, F. N. C. (2019). RAFF.jl: Robust Algebraic Fitting Function in Julia. Journal of Open Source Software, 4(39), 1385. https://doi.org/10.21105/joss.01385