This is the GitHub organization page for Amherst College STAT/MATH 495 Advanced Data Analysis (Fall 2017). Whereas the course content and syllabus are on the course webpage, this organization centers around the problem sets.
Executive summary tl;dr
Students retrieve and submit the problem sets from this organization by using GitHub forks and pull requests.
Problem set submission process
The typical work flow for a problem set is below. All italics indicate GitHub terminology/lingo; this and more are explained in Prof. Jenny Bryan's Happy Git and GitHub for the useR
- On GitHub: The instructor will post a skeleton outline of each problem set in its own repository AKA repo, for example
PS01. This repository will contain the necessary data files and a template R Markdown file. Let's call this the master copy of the repo.
- On GitHub: Students will fork (i.e. make a copy of) the repo to their own GitHub account.
- GitHub -> locally: Students will clone (i.e. download) this forked repo locally as an RStudio project on their own machine.
- Locally: Students will complete the problem set on their own machine.
- Locally -> GitHub: Students will commit and push (i.e. upload) their work to the forked copy of the repo in their own GitHub account.
- On GitHub: Students will submit their work via pull request. This is a request to the owner of the master copy of the repo (in this case the instructor) to inspect and merge the proposed changes. When prompted to "Open a pull request", please give it title your name!
- Feedback will be delievered.
- The instructor will however not complete the final step of the typical pull request: they will not merge the proposed changes.
Why are we doing this?
Question: Why did you set up this complicated scheme? Why not just give all students write-access to the repositories (by making them a collaborator) and allow them to submit individual files?
Answer: Because much of the collaboration that occurs in the open-source world centers around pull requests to propose changes/improvements. For example, many of the crowd-sourced changes/improvements to the
ggplot2 R package for data visualization was done via one of (as of 2017-09-06) 610 pull requests. I would like to empower students to start taking their first steps of participation in this ecosystem.
Start small! Among my earliest pull requests; a very minor one. Open this link and click on the "Files Changed" tab.