dsdk_cookie

This repo should be used as a template for new data science projects at Penn Medicine, especially for any project that is expected to push data back into the EMR

Instructions

Cruft

To use this repo you first need a python enviroment with cruft installed. To install cruft run

pip install cruft

or the equivalent for your environment.

Creating a new project

The project name must be a valid postgres schema name, python module name, and nomad task name.

Choose a project / repo name:

1. Short but does not abbreviate words
2. All lower case
3. No hyphens
4. No underscores

Once cruft is installed you can run the following to create your new project:

cruft create https://github.com/pennsignals/dsdk_cookie

Push to github

Once your project is created you should push it to github by making a new private repo and pushing it like:

cd {your-project-name}
git init
git add -A
git commit -m "first commit"
git branch -M main
git remote add origin [email protected]:pennsignals/{name}.git
git push -u origin main

Updating an existing project

Occasionally updates will be made to dsdk_cookie which should be rolled into your project to keep it up to date. To check if your project is up to date run the following:

cruft check

If your project is out of date you can run the following to update it:

cruft update

Further instructions

For further instructions on how to use your project can be found in the readme.md for your project at https://github.com/pennsignals/{your-repo-name}#readme

TODO

• Add cruft check to CICD in {{cookiecutter.name}}/.github/workflows. Maybe as part of test.yml? Maybe it's own thing?
• Add .github/test.yml to root of this repo and have it run cookiecutter . to ensure that the template isn't broken when updates are made. See #3
• Update {{cookiecutter.name}}/readme.md with instructions on how to use cruft update.