Citing the package about umap

[tkonopka]

Thanks for asking. Yes, citations should go to the original authors. The umap-learn repository lists recommendations here. In brief, you can cite the manuscript on arxiv or the paper in JOSS.

Of course, if you find this R package useful, it would be very nice to credit that as well. There is no formal publication, but you can cite it as an R package (see citation("umap")). An advantage of citing the software (as opposed to a manuscript) is that it makes clear what version you've used. But customs in this regard vary across disciplines, even across journals and institutions within one discipline. If you are at all constrained by number of references, do give preference to the original authors.

In case you run analyses and produce reports with Rmarkdown, you can consider including that as an appendix/supplement to a publication and including a sessionInfo() at the end. That would credit all the packages in the analysis, including base R, the dependencies this package uses, and any others.