search

swcarpentry / good-enough-practices-in-scientific-computing

Minimalist alternatives to "best practices" paper
https://swcarpentry.github.io/good-enough-practices-in-scientific-computing/
Other
159 stars 23 forks source link

Source code not in with data #88

Closed beroe closed 8 years ago

beroe commented 8 years ago

I am having an issue with my grad student putting copies of scripts in his data folders to keep everything "together". I would like to add a specific caveat against this potentially common practice, which has the dangers of source code proliferation (which copy am I editing vs which one am I running?) and also makes version control difficult/impossible without extensive use of .gitignore. (These are genome sequence, so we can't keep raw and filtered reads in the repo, for example.)

The question is where best to put this? When we are talking about Project Organization, up from with Software, or down with Version Control. I somewhat favor with Software, but P.O. also makes a lot of sense.

jkitzes commented 8 years ago

The current project organization section definitely recommends different subdirectories for code and data - is there something beyond the current recommendation that you'd suggest?

beroe commented 8 years ago

It is definitely implicit there, but I was hoping for an explicit statement NOT to do that.

gvwilson commented 8 years ago

Please check the latest version (generated PDF).