equinor / data-science-template Goto Github PK

I think, we should allow people to use the repo as a Template directly by making this as Template repository, as shown here https://docs.github.com/en/free-pro-team@latest/github/creating-cloning-and-archiving-repositories/creating-a-template-repository

The benefits of this way are described here https://docs.github.com/en/free-pro-team@latest/github/creating-cloning-and-archiving-repositories/creating-a-repository-from-a-template#about-repository-templates

Add templating to other areas of the project (package, notebook, ...)

Add azure-pipelines.yaml and instructions for configuring CI for generated projects

Hey,

When viewing the readme.md within the {{cookiecutter.repo_name}} folder this still specifies Anaconda and conda as environment control manager whereas the root readme.md now prefers virtual environment.

Virtual environment and non-Anaconda seems to be the way we are going now so should we update the readme.md within {{cookiecutter.repo_name}} to match?

Add support for pipelines within Azure / Omnia:

• DataBricks
• Data Factory
• ML Services
• Functions
• Web Apps
• API Management

The existing sample package is a bit outdated and examples could be improved.

Setup CI so sphinx project is validated and build
CD for (optional) auto deployment to blob storage
Perhaps throw in a make file for this and some other bits while at it.

This is a great project setup. It would be even better/more usable if it could be used as a "cookie-cutter" template :)

Within the readme.md within {{cookiecutter.package_name}}

cd docs/writeup
make html

There are a proportion using this on a windows operating system and the make command does not work here.

Either needs a suggestion for windows or a clarification that this does not work on a windows operating system.

Please don't remove the Sphinx template as it's a very good idea to include from start of project.