EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3458. goalp: Weighted and Lexicographic Goal Programming Interface for R

Invited abstract in session WC-30: Optimization Frameworks, stream Software for Optimization.

Wednesday, 12:30-14:00
Room: 064 (building: 208)

Authors (first author is the speaker)

1. David Palma
Management, Leeds University Business School
2. Richard Hodgett
Leeds University Business School, University of Leeds

Abstract

Goal programming (GP) is a special kind of linear programming (LP) technique where users seek to find a solution that deviates as little as possible from a set of goals, expressed as a series of linear equalities or inequalities. Applications include: planning manufacturing to match demand as closely as possible, given a limited number of raw materials; scheduling workers’ shifts to match their stated availability as closely as possible; etc.
Solving GP problems requires expanding the original goals into a larger LP problem. While the expansion is relatively straightforward, it is easy to make mistakes when utilising it manually in large problems. Furthermore, there are few software implementations of GP, with users mostly relying on general-purpose LP software. This makes solving large GP problems tedious and prone to errors.
We fill this gap by introducing goalp, an R package for GP. It allows solving basic, weighted, and lexicographic GP problems, as well as a mixture of them. The package can take as input a human-readable representation of the goals, or a set of matrices. Then it automatically sets up the corresponding LP problem, which it later solves using the powerful lp_solve solver. goalp also allows mixing goals with traditional linear optimisation constraints, i.e. those not allowing for deviations. Furthermore, it allows for continuous, discrete, and dichotomous decision variables.
goalp is currently available via CRAN, including a user guide (vignette).

Keywords

Status: accepted

Back to the list of papers