Discussion place for R packages built with "extras", special libraries and bindings.
This GitHub diminutive organization holds a collection of minimal R packages, designed to illustrate special cases.
Use this repo to discuss issues, contribute fixes or enhancements, and contribute your own examples.
These examples vary in quality and currency. (A definite problem right now is that some will point to the wrong organization for CI badges and installation instructions - please submit a PR if you use one and make updates!).
| Package | Description |
|---|---|
| depbioc | minimal R package with import from a Bioconductor package |
| depsf | Minimal R package importing sf with CI testing on Travis |
| earcut.cpp | Minimal C++ bindings of Mapbox earcut.hpp for R |
| gdalmin | Minimal R package with bindings to GDAL, built from source a la rgdal |
| gdalosx | Test a system package install of GDAL on OSX with Travis |
| ncdep | CI testing for NetCDF with R |
| rgdalwinhdf4 | Build rgdal on Windows with Appveyor, and read HDF4 |
| rglmin | minimal R package importing rgl |
These packages were created to prove a point, or demonstrate something in the works on x machine sense
| Package | Description |
|---|---|
| brwsrchk | Does R CMD check warn / error on browser() calls? |
| dputtruncat | R package to show dput truncation |
These are probably not worth mentioning, see gdalmin and gdalosx and rgdalwinhdf4 as more current examples for gdal stuff.
gdal.rhub, hdf4.rhub, hdf5.rhub, anc
The repos grew naturally out of informal and "scatter gun" testing and exploration. When I finally learn how to get something to work I like to have a tiny example, because I know it worked once and an actual package is good proof. It's extremely easy to move on from that simple example, complicate things and get lost - and then you're in git hell or worse. It's also hard to fork a functional package and strip it back to just that part you need, it's exactly this getting started core that I'm often looking for.
Ideally these might be canonical, the perfect starting point if you need library IJK in an R package - just fork it, test it, and starting specializing it. These examples are definitely not good enough for that, yet.
https://github.com/craigcitro/r-travis/wiki/Recipes a bunch of examples for Travis-CI
https://github.com/ropenscilabs/tic (this one's fantastic, stands out as minimal example gold mine on its own)
https://github.com/rstudio/bookdown-demo (minimal bookdown)
https://github.com/rstudio/blogdown (Blogdown will create a functional project for you)
https://github.com/yihui/rmini
http://kbroman.org/pkg_primer/pages/minimal.html
https://github.com/b4winckler/reverse
https://github.com/yixuan/miniGPU