curricular-analytics-graph

Intended to be a CurricularVisualization.jl-compatible graph renderer.

This repo is set up in two parts:

• src/, which defines a generic graph renderer as a library.
• app/, a React app that uses the library to make a kitchen-sink demo.

Usage (as library)

The graph renderer is currently a work in progress.

$ npm install git+https://github.com/SheepTester-forks/curricular-analytics-graph.git

There are no pre-bundled files available. If you want to import this into your project, you will have to use a bundler like esbuild with --loader:.svg=file.

To create a graph, you need to

1. Import Graph from curricular-analytics-graph/src.

2. Provide an options object to the Graph constructor. A list of options is documented in GraphOptions. All options are optional, and by default, the graph will only have circles and arrows.

   To update the options object, directly modify Graph.options, then force a re-render using Graph.forceUpdate().

3. Provide a degree plan, in the form of a list of list of course objects, to Graph.setDegreePlan.

   Each course object must contain backwards and forwards, which are arrays of references to other course objects, representing the prerequisite relationships between courses. Course objects can contain other information, such as course IDs or names, which will be passed directly to the callbacks in the options object in the Graph constructor.

   These links don't store any metadata, so if you want to keep track of the type of prerequisite (e.g. prereq vs coreq vs strict coreq), you should maintain a separate object or Map mapping from pairs of course IDs to any metadata you wish to keep.

   To change the degree plan again later on, simply call setDegreePlan again with the new degree plan.

4. Add Graph.wrapper to the document.

import { Graph } from "curricular-analytics-graph/src";

// Define the properties of the object used to store course data (if using
// TypeScript). The only required fields used by `Graph` are `backwards` and
// `forwards`, which are arrays of references to other course objects in the
// plan.
type Course = {
  name: string;
  backwards: Course[];
  forwards: Course[];
};

const graph = new Graph<Course>({
  // The `courseName` option determines the text that displays underneath each
  // course node.
  courseName: ({ course }) => course.name,
});

// In practice, these links would be attached automatically rather than
// manually.
const plan = [
  [
    { name: "MATH 1", backwards: [], forwards: [] },
    { name: "MATH 2", backwards: [], forwards: [] },
  ],
  { name: "PHYS 10", backwards: [], forwards: [] },
];
plan[0][0].forwards.push(plan[1][0]);
plan[1][0].backwards.push(plan[0][0]);
plan[0][1].forwards.push(plan[1][0]);
plan[1][0].backwards.push(plan[0][1]);
graph.setDegreePlan(plan);

document.body.append(graph.wrapper);

For more examples about customizing the graph renderer, you can refer to these examples:

• Kitchen sink demo, live
• Using a graph in React, live

Development (app)

Requires Node 22+ (on Mac/Linux, I recommend nvm: curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/master/install.sh | bash then nvm install node) and Yarn (npm install -g yarn).

First, you also need to clone ExploratoryCurricularAnalytics in the same parent folder as this repository. Refer to the setup steps in that repo's README, but what's essential is that you run make protected (which means you need the protected data files).

$ cd ..
$ git clone https://github.com/SheepTester-forks/ExploratoryCurricularAnalytics.git
$ cd ExploratoryCurricularAnalytics/
# (Download the required data files)
$ make
$ make protected
$ cd ../curricular-analytics-graph/

Then, you can run yarn dev to start a development server.

# Build and minify the app
$ yarn build

# Start a local HTTP server that recompiles on the fly
$ yarn dev