The goal of this project is to analyze the changes in the number of vignettes over time, ideally to quantify the effect of knitr and rmarkdown.
I created a personal CRAN mirror and cloned all available Bioconductor git repos on 2021-10-09. For each current and available package under the CRAN archive, I counted how many sweave and markdown vignettes were present. For Bioconductor, I counted how many of each vignette type was present at each commit.
The RDS files with this data are in:
For more information on how I ran this code, see the _targets.R file and the underlying functions in the R directory.
The quickest way to get started with this data is to clone it and then use targets to load the data of interest:
targets::tar_load(cran_has_vignettes)
targets::tar_load(bioc_has_vignettes)I'm relatively sure the code is counting vignettes correctly, given that only 200 Bioconductor vignettes have zero, and when I check some of those manually, there are no vignettes listed on the package web-page.