Skip to content

README.md

Latest commit

 

History

History
41 lines (27 loc) · 2.52 KB

README.md

File metadata and controls

41 lines (27 loc) · 2.52 KB
Raw

DegreeViz

What's DegreeViz?

As a first year McGill student, I was left confused on how to ensure that all my pre-requesities are met, and I'm taking all my courses when needed. To help, I created DegreeViz.

With a simple UI, this allows users to select their McGill program from the dynamic search bar, and automatically get a generated directional graph that shows prerequisite relationships.

The most recent features that have been added are:

  • ability to add minors and/or other majors to the graph
  • save plans to a user profile
  • basic modifications of the graph for customization (add, delete, modify nodes and edges, etc.)

Deployment Locally

To run DegreeViz locally, you can install all required packages with pip install -r requirements.txt, or any other package manager of choice. Then, you can simply run the command

python app.py

and that should give a link to run the site locally in a dev environment.

Here's a basic run-down of what's where: As a Flask app, the core request handling is stored in app.py. Runtime helper code lives in the degreeviz package:

  • degreeviz/programs/graph_builder.py builds graph data from a selected program.
  • degreeviz/programs/llm_prerequisites.py calls Gemini to simplify prerequisite text.
  • degreeviz/programs/requirements.py parses program requirement buckets.
  • degreeviz/catalogue/course_scraper.py refreshes the static McGill course catalogue JSON.
  • tools/ contains manual maintenance commands, such as refreshing courses, programs, and honours mappings.

Finally, the /templates directory holds all HTML files for the front-end, along with the /static directory, which has browser JS, CSS, images, and static JSON data.

What I'm working on

DegreeViz is still a work in progress, and I'm open to contribitors which can submit PR requests so I can merge them in. Here's a few things that are currently on my drawing board for this project:

  • Improving saved-plan migrations as the graph schema evolves
  • Re-doing the UI to be more modern. I'm not a front-end dev and so this is more challenging for me
  • Incorporating better CI/CD tests to get basic tests before deployment
  • Have better monitoring. Although I have basic monitoring, recently the Gemini API migrated to a new package and structure, which broke my website when making API calls. Unfortunately, the only way I found out about it was by going on the site myself, which is not ideal...

If you have any other ideas, please feel free to reach me at andrey.ambartsumov@mail.mcgill.ca!