pkgndep: a tool for analyzing dependency heaviness of R packages

Abstract

SummaryNumerous R packages have been developed for bioinformatics analysis in the last decade and dependencies among packages have become critical issues to consider. In this work, we proposed a new metric named dependency heaviness that measures the number of unique dependencies that a parent brings to a package, and we proposed possible solutions for reducing the complexity of dependencies by optimizing the use of heavy parents. We implemented the metric in a new R package pkgndep which provides an intuitive way for dependency heaviness analysis. Based on pkgndep, we additionally performed a global analysis of dependency heaviness on CRAN and Bioconductor ecosystems and we revealed top packages that have significant contributions of high dependency heaviness to their child packages. Availability and implementationThe package pkgndep and documentations are freely available from the Comprehensive R Archive Network (CRAN) https://cran.r-project.org/package=pkgndep. The dependency heaviness analysis for all 21,741 CRAN and Bioconductor packages retrieved on 2021-10-28 are available at https://pkgndep.github.io/.