suggestion

[Antonio Pedro]

Hi all,

I've been using ProjectTemplate with great pleasure and usefulness for *all* my data analysis projects now. Recently, I've trying to move them all to github as well. Thus I think myself and other no so familiar with version control systems will benefit even more from the ProjectTemplate by inclusion of a short section on how to use it in conjunction with some control version system, such as github. It could also including tips such as which type of files should be upload to repository. 

Thanks John and everybody who are contributing to the project,

All the best,

Antonio Pedro.

[John Myles White]

Hi Antonio Pedro,

I upload everything that's not confidential data to GitHub for my analyses. That means that I can just enter a project directory and run

git add .
git commit -m "Committing new project"

to get started. From there, I always follow the remote branch instructions provided by GitHub since I can never remember those commands.

I don't do anything smarter than that for analysis projects, but maybe other people will have suggestions. Julia is the only project I'm involved in that required substantial use of branches and ProjectTemplate is one of the few GitHub repos I've ever had that needed to use tags. So I'm not sure if they're worth learning about for most users.

-- John

[Hilary Parker]

Hey guys,

I've been thinking about this a bit because I'm publishing code organized using ProjectTemplate onto github right now.

One easy thing might be some sort of quick tutorial of how to turn the README file into a README.markdown file (which then will be displayed on a github repository page if I understand it correctly?). I am going to try to do something fancy with knitr so that the markdown file walks someone through the entire analysis, complete with graphs, code chunks, etc. Possibly overkill, but it might be nice.

For reproducibility in this readme file I'll also explain "Hey set your config file to munging: on if you really want to reproduce EVERYTHING, and munging: off if you want to save a lot of time and work with the cleaned dataset" again with code chunks.

This is a bit stream-of-thought because I'm in the middle of the project, I'll be sure to post about it when I'm finished and have the workflow nailed down.

Hilary