This is my attempt to make use of my Wyvern project.
This project is very coupled with Wyvern and will only take in .CSV files created using it. I have added it as a submodule.
Prior to using Gryph Graph, you need to scrape the requisite data using Wyvern. You can just specify an output name (see below) and wyvern.py should be able to use the default URL. At the time of writing, this URL works. Manually specifying the URL will only become relevant in the future if the Guelph website changes. If you don't want to make a new scrape, there exists a pre-made courses.csv in the wyvern submodule.
python3 graph.py courses.csv output
graph.py will take your .CSV and create:
-
A dotfile representation of the dependency graph (output.gv)
-
An SVG file of it (output.svg), and
-
A PNG of it (output.png)
You may automate this process using the run.sh script.
Though I recommend running Wyvern first, filtering out what you don't need, and then using that. If you load all ~1800 courses of UofG, you'll get unreadable garbage. It will also be hard to tell which analyses are correct or not (see below).
Work in progress: the prerequisite and restriction parsing is somewhat inaccurate as I am doing a narive regex matching. If a course has the prerequisite string (CIS\*3750 or CIS\*3760), (CIS\*2460 or STAT\*2040), my regex match will return all four courses as prerequisites, which is terribly inaccurate. For most courses it should be fine, but I am definitely going to make a proper parsing algorithm in the future.
There are a lot of things in both Wyvern and GryphGraph that I want to work on. So far the only way to really narrow down a faculty is to manually edit the CSV in Excel. Fortunately with the number of parameters I scraped, it should hopefully be suitable for basic cases.