search

Data4Democracy / drug-spending

Project to understand pharmaceutical spending, currently focused on US government programs.
73 stars 46 forks source link

Restructuring of repo #36

Closed jenniferthompson closed 7 years ago

jenniferthompson commented 7 years ago

I think I have it cleaned up - anything y'all see that I missed or should be different? I made folder names generally consistent. I have empty folders created on my drive for some things that aren't showing up here (R/d4ddrugspending, eg) because there aren't any files in them yet.

jenniferthompson commented 7 years ago

Not sure if it would be better to go ahead and merge @cduvallet's PR first, then incorporate that here, or the other way around?

dhuppenkothen commented 7 years ago

For the empty folders, you could make a bare-bones README.md in each that will work as a sort of directory for the repo structure? Committing that, the folders will appear.

dhuppenkothen commented 7 years ago

Let's get the directory structure sorted first. Once that's done, I may be able to write a couple of simple Python scripts that allow one to import code, which would I think help @cduvallet with not duplicating code that's already there.

jenniferthompson commented 7 years ago

Ah, yep, good call on READMEs; we'll definitely want those anyway. I'll try to do that today sometime. I defer to you two on Python organization!

mattgawarecki commented 7 years ago

Review comments:

  1. I think we can safely get rid of cms/, since that data's up in data.world. (Probably just a result of accidentally removing .gitignore.)
  2. I've seen in other projects where all-caps filenames are used for basically all documentation (licensing, readmes, etc.). Maybe this applies to datadictionaries/data_dictionary_template.md? Thinking maybe just datadictionaries/TEMPLATE.md.
  3. Probably safe to remove the suffix on datadictionaries/manufacturers_datadict.md (i.e., just datadictionaries/manufacturers.md)
  4. Re: python/datawrangling/centerwatch -- I'm fine with it going in python/d4ddrugspending since it's sort of a sub-module. Just for clarification, it's a Scrapy spider that gathers data from a site called "Centerwatch."
  5. If we do the above, we can safely get rid of python/datawrangling.
  6. If you don't foresee us writing READMEs for directories that still need to exist, you can simply create an empty file (conventionally, .gitkeep) and Git will track that file, which guarantees the directory will exist when people clone the repo.
  7. The rest of the structure looks good to me, I think.
cduvallet commented 7 years ago

I think it makes more sense to merge this PR first. I can then move around my files to be in the right place, import code from @dhuppenkothen's scripts, etc. This way I can also be the first guinea pig for the new structure ;)

jenniferthompson commented 7 years ago

@mattgawarecki @dhuppenkothen Made the changes - let me know if I missed anything! I left python/datawrangling for now but feel free to remove it if it's never going to be useful.

ETA - going to work on edits to the main README to reflect these changes later this afternoon/evening, so it's not 100% ready to go yet, but that's the only change I have planned unless you two see anything else.

jenniferthompson commented 7 years ago

README is edited! Ready to merge when y'all are.