The build process includes a "docs" target to build framework documentation as markdown using the .Rd files generated by Roxygen for the various modules and framework.
However, in grand Rstudio style, the maintainer of the Rd2md package recently gutted the interface, reduced the functionality to creating a single markdown file for all the functions in a package, and left the R world without a simple converter to take the Rd output (closer in format to Latex) and push it out to individual .Rd files.
It will be possible to use the new Rd2md package to dump each VE package into a single .md document which will serve the most basic purpose of getting the function documents into the website /docs area, but that will require some work.
More annoyingly, the new package also breaks the function-by-function parsing that was done previously and that allowed the functions to be grouped by their target audience (model users, model developers, internal functions, etc). It also breaks the very cute visual documentation output (showing the function utilization graphically).
I think there is a solution here, but it's not a CRAN package (has to be built from Github) and I don't want to make it a specific dependency for VE since its future is uncertain: https://github.com/Genentech/rd2markdown/
Ironically, there is (of all things) a Python package that can do .Rd to markdown conversion https://github.com/comet-ml/rd2md, but that doesn't solve the grouping problem (without more Python development).
The build process includes a "docs" target to build framework documentation as markdown using the .Rd files generated by Roxygen for the various modules and framework.
However, in grand Rstudio style, the maintainer of the Rd2md package recently gutted the interface, reduced the functionality to creating a single markdown file for all the functions in a package, and left the R world without a simple converter to take the Rd output (closer in format to Latex) and push it out to individual .Rd files.
It will be possible to use the new Rd2md package to dump each VE package into a single .md document which will serve the most basic purpose of getting the function documents into the website /docs area, but that will require some work.
More annoyingly, the new package also breaks the function-by-function parsing that was done previously and that allowed the functions to be grouped by their target audience (model users, model developers, internal functions, etc). It also breaks the very cute visual documentation output (showing the function utilization graphically).
I think there is a solution here, but it's not a CRAN package (has to be built from Github) and I don't want to make it a specific dependency for VE since its future is uncertain: https://github.com/Genentech/rd2markdown/
Ironically, there is (of all things) a Python package that can do .Rd to markdown conversion https://github.com/comet-ml/rd2md, but that doesn't solve the grouping problem (without more Python development).