Data Science Project Template

A cookiecutter template for creating a data science project repository.

This template makes it easy to start your Jupyter notebook on a cloud VM. The following instructions have been tested on Google Cloud Platform with Debian 10.

There are three stages in deploying your own Jupyter notebook on the cloud.

1. Stage 1: Create a virtual machine and login using ssh
2. Stage 2: Use cookiecutter to start your project directory
3. Stage 3: From inside your project directory, configure and run your Jupyter notebook

Stage 1: Create a Virtual Machine

This template hast been tested with Google Cloud VM with Debian 10 (Buster) [instructions]. Either Allow HTTPS traffic (defaults to port 443) is selected or the port to be used in Stage 3 (see below) is open on the firewall [instructions].

Stage 2: Instantiate Project from Template

• Python (Miniconda3)

  curl -fSL -o ~/miniconda.sh https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh && \
      bash ~/miniconda.sh -b -u -p ~/miniconda3 && \
      ~/miniconda3/bin/conda init bash && \
      source ~/.bashrc

• Install dependencies

  pip install cookiecutter
  sudo apt-get update && \
      sudo apt-get install -y git

• Create [your-project-name] directory for "Your Project Name"

  cookiecutter gh:dddlab/reproducibility-demo

  Cookiecutter will ask you some questions including,
  • project_slug: name of your project directory: e.g. your-project-name
  • github_repo: the repository where your project will live
  • base_jupyter_notebook_image: a compatible notebook image. To find a starter image, see available image descriptions and their Docker Hub image tags links
  • jupyter_notebook_port: port number for https (default is 443)

Stage 3

Find README.md in [your-project-name] directory for instructions on how to launch your Jupyter notebook!