Training Set Determination for Genomic Selection R package TSDFGS version 2.0 CRAN Task View: Agriculture (https://CRAN.R-project.org/view=Agriculture) We propose an optimality criterion to determine the required training set, r-score, which is derived directly from Pearson's correlation between the genomic estimated breeding values and phenotypic values of the test set <doi:10.1007/s00122-019-03387-0>. This package provides two main functions to determine a good training set and its size. Software Depends: R (>= 2.10) Imports: dplyr, ggplot2, latex2exp, lifecycle, parallel, Rcpp (>= 1.0.8.3) LinkingTo: Rcpp, RcppEigen Jen-Hsiang Ou <jen-hsiang.ou@imbim.uu.se> Comprehensive R Archive Network (CRAN) Jen-Hsiang Ou [aut, cre] (ORCID: <https://orcid.org/0000-0001-9305-2931>), Po-Ya Wu [aut] (ORCID: <https://orcid.org/0000-0002-7342-2867>), Chen-Tuo Liao [aut, ths] (ORCID: <https://orcid.org/0000-0001-9777-3701>) GPL (>= 3) 2022-06-07 application/tgz https://CRAN.R-project.org/package=TSDFGS doi:10.32614/CRAN.package.TSDFGS