Skip to content

tasdemir-lab/project-template

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PROJECT TEMPLATE

This repo is a minimalistic template for empirical economics projects. It usefully serves as a bare bone directory structure in empirical research projects in economics. One can easily extent the structure to allow for more complicated cases.

├── code
├── data [not tracked]
│   ├── final
│   ├── inter
│   └── raw
│       ├── raw_data-1
│       │   └── doc
│       └── raw_data-2
│           └── doc
├── docs [not tracked]
├── out
├── readings [not tracked]
├── report
└── temp [not tracked]

The template has been constructed with a specific workflow in mind, and has the following features/recommendations:

  • All code go into code and modified/analysis-ready data goes into data/final directory.

  • All (if any) intermediate data should be in data/inter

  • All raw (original) data go into data/raw directory. This folder is not tracked by git. After cloning the repo, the user should run 00-data-download.R or 00-data-download.py file to download raw data from a Google Drive folder.

  • All output (tables, figures etc.) must go into out directory.

  • The following is some advanced workflow, and no need to adapt for beginners:

    • Ideally each data/raw/data_x directory should have a makefile that runs the code in that directory and subdirectories. This makefile works as a master code file. Another option would be to use a master code file. When there are code files running different programming languages, makefile option is a better way to go.
    • A main makefile should be in the root directory. This makefile works as a master code file. Another option would be to use a master code file. When there are scripts in different programming languages, makefile option is a better way to go.
  • The project directory should NEVER be in a cloud sync directory such as Dropbox, Box or Google Drive.

  • If you need to use certain libraries or configuration files to be included in the project folder, then create a .lib or .config folder in the project root.

  • If you add new directories to the root project folder, make sure that you are not sharing anything that you would not want to share via Github. So, check and, if necessary, modify .gitignore file.

  • .gitignore file follows a particular logic. First unfollow everything, then follow the ones you want to be tracked by git.

The following coding principles are necessary for a reproducibility and easy collaboration:

  • Each small task, such as creating a figure, table or estimation of a specific model should be done by a separate code, such as table-2-1.R. If code pieces must be run in an order, numbering files sequentially e.g. 04-table-3-2.do or table-3-2.Ris a useful practice.
  • In collaboration with the research team, following the modified Gitflow workflow by DIME is highly recommended. Please also consult Development Research in Practice by DIME Analytics of World Bank.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published