gosset: An R package for analysis and synthesis of ranking data in agricultural experimentation

Kauê de Sousa; David Brown; Jonathan Steinke; Jacob van Etten

doi:10.1016/j.softx.2023.101402

gosset: An R package for analysis and synthesis of ranking data in agricultural experimentation

SoftwareX. 2023 May:22:None. doi: 10.1016/j.softx.2023.101402.

Authors

Kauê de Sousa^{1

2}, David Brown^{3

4

5}, Jonathan Steinke^{2

6}, Jacob van Etten²

Affiliations

¹ Department of Agricultural Sciences, Inland Norway University of Applied Sciences, 2318 Hamar, Norway.
² Digital Inclusion, Bioversity International, Parc Scientifique Agropolis II, 34397, Montpellier Cedex 5, France.
³ Laboratory of Geo-Information Science and Remote Sensing, Wageningen University & Research, Droevendaalsesteeg 3, 6708 PB, Wageningen, The Netherlands.
⁴ Digital Inclusion, Bioversity International, 30501, Turrialba, Costa Rica.
⁵ College of Agriculture and Life Sciences, Cornell University, 14853 Ithaca, NY, United States of America.
⁶ Thaer Institute of Agricultural and Horticultural Sciences, Humboldt University Berlin, Unter den Linden 6, 10099 Berlin, Germany.

Abstract

To derive insights from data, researchers working on agricultural experiments need appropriate data management and analysis tools. To ensure that workflows are reproducible and can be applied on a routine basis, programmatic tools are needed. Such tools are increasingly necessary for rank-based data, a type of data that is generated in on-farm experimentation and data synthesis exercises, among others. To address this need, we developed the R package gosset, which provides functionality for rank-based data and models. The gosset package facilitates data preparation, modeling and results presentation stages. It introduces novel functions not available in existing R packages for analyzing ranking data. This paper demonstrates the package functionality using the case study of a decentralized on-farm trial of common bean (Phaseolus vulgaris L.) varieties in Nicaragua.

Keywords: Bradley–Terry; Data science; On-farm trials; Plackett–Luce; Tricot approach.