Open brandmaier opened 2 years ago
Some more thoughts on this. In my recent repro project, I added a line like this
To reproduce the analysis, run:
make docker make -B DOCKER=TRUE
This will create the results of the Bayesian models as presented in the results section of the manuscript as well as reported in the supplementary materials.
If you do not have docker and make installed on your system, install all required software dependencies as listed in the Rmd files' YAML header manually and then first compile fit_model.Rmd and then results.Rmd.
I was wondering if we could dynamically create this part of the README per default. For example, the README could contain a list of all dependent R packages created from the YAML header. And the list of Rmd files could also be automatically generated. This would then help people who do not know the workflow to help setting up their local computer to reproduce the script (for example, if they do not have docker and Make and no system rights to install it).
A repro project should contain a readme.md with basic instructions on how to reproduce and a link to the repro package and our papers.