Multi Language [Data Analysis] Project Template

About

The purpose of this work is to create a language agnostic project template for data analysis.

Insipration

I work with Julia, Python, and R. While the language-specific conventions are different the projects are usually structured similarly.

Usually, I start with what I want to accomplish before I pick what language to work with. So, I decided that it would helpful to automate and standardize the steps prior to picking a language.

Operating Systems

Currently focused on Ubuntu-based Linux distros, such as Linux Mint but the setup script should work for most Linux distros.

Language-Specific Conventions

Julia

• uses Pgk to manage dependencies
• tests are housed under ./test

Python

• uses virtual environments or other software such as poetry, virtualenv, etc. to manage dependencies
• tests are housed under ./tests

R

• uses the renv package to manage dependencies
• tests are housed under ./tests

Similar works

There are many existing works that create language-specific project structures. Below are some examples.

• Julia:
  • BestieTemplate
  • DrWatson
  • PkgTemplates
• Python
  • cookiecutter
  • GitHub Topics
• R
  • usethis
  • ProjectTemplate
  • GitHub User

How to Use

1. git clone repo
2. Open a terminal and navigate to ^
3. within the terminal run chmod +x create-project.sh

Optional Flags

• -l: Programming language flag
  • Only affects the testing directory
  • If no language is specified, then both the ./test and ./tests directories are created
    • Simply delete the unused testing directory when a language for the project is chosen
  • Takes the following args (case insensitive)
    • r
    • py
    • python
    • jl
    • julia
  • WARNING only one option should be specified, there is no guardrail if more than one option is passed
• -p: Project directory
  • Takes any existing path; e.g., ./create-project.sh -p ~/Documents/proj-root
  • Otherwise the results will be created in the current directory

Examples

Get the script.

mkdir git
cd git
git clone https://github.com/SaintRod/multi-lang-project-template.git
cd multi-lang-project-template
chmod +x create-project.sh

A) No clue what language will be used, and the project hasn't been started

The following will create the results in the current directory. Copy/cut and paste the results to the project root directory when it's been created.

./create-project.sh

B) No clue what language will be used, and the project root directory exists

./create-project.sh -p ~/.../project-root

C) You know what language will be used, and the project root directory exists

# R
./create-project.sh -l r -p ~/.../project-root
./create-project.sh -l R -p ~/.../project-root

# Python
./create-project.sh -l py -p ~/.../project-root
./create-project.sh -l python -p ~/.../project-root
./create-project.sh -l PY -p ~/.../project-root
./create-project.sh -l PYTHON -p ~/.../project-root
./create-project.sh -l pYtHoN -p ~/.../project-root

# Julia
./create-project.sh -l jl -p ~/.../project-root
./create-project.sh -l julia -p ~/.../project-root
./create-project.sh -l JL -p ~/.../project-root
./create-project.sh -l jL -p ~/.../project-root

D) You know what language will be used, but the project root directory doesn't exist

In this scenario the results will be created in the current directory. When the project directory exists, copy/cut and paste the results there.

# R
./create-project.sh -l r
./create-project.sh -l R

# Python
./create-project.sh -l py
./create-project.sh -l python
./create-project.sh -l PY
./create-project.sh -l PYTHON
./create-project.sh -l pYtHoN

# Julia
./create-project.sh -l jl
./create-project.sh -l julia
./create-project.sh -l JL
./create-project.sh -l jL

Results

project-root/
├── config
├── data
│   ├── input
│   └── output
├── docs        # technical documentation
├── LICENSE
├── notebooks   # literate programming medium
├── README.md
├── scripts     # files that import from ./src
├── src         # functions or modules
├── test        # for Julia
└── tests       # for R and Python

Additional details:

• docs - Technical documentation created programmatically or manually by users
• notebooks - Any literate programming medium such as Jupyter Labs, Quarto markdown, RMarkdown, or Pluto notebooks
• test and tests - If both testing directories are present, it implies no language was specified

Contributing

Contributors welcome!

Fork the repo, make changes, and submit a PR :)