This report provides class progress statistics for the Getting Started section of the Computational Physics course KD5081 at Northumbria University.

Friendly request

If you use this notebook as a basis for your own work please let me know - hearing from people who find what I've written about useful encourages me to spend more time in the future documenting my ideas/workflows.

Student workflow

From a student's point of view the process is:

1. Create a Github account (it is free).
2. Go to the course issues page and click on “New Issue”.
3. Next to “Python basics checklist” click “Get started”.
4. Without editing any text, click “Submit new issue”. This will generate an issue with a pre-defined checklist.
5. Work through the tutorials and questions for the relevant section of the course (part one - “Getting Started”) and tick off the items on the checklist accordingly.

Issue template

Steps 2 to 4 in the student workflow above enable students to generate the checklist via a Github issue template which can be found here.

Automation

This notebook is hosted on Github and automatically runs after every commit, and once an hour. Automation is via a Github action. This workflow was designed following the clear advice given by the Living with machines project.

Summary plot

The end result of this notebook is a class progress summary plot, which can also be found on this page.

Adapting this for a different course / checklist

If you want to adapt this workflow for your own Github-hosted course repository, the following should work (please let me know if it does or does not!):

1) Adapt the issue template) with a checklist for your own curriculum:

• The empty check boxes - [ ] need to be immediately followed by the checklist item e.g. "Open and save a Jupyter Notebook" and then a newline, or the regex below will not work.
• If you change the issue label (currently "python_checklist") you will need to update this notebook accordingly (see below).

2) Upload the issue template to your repository folder .github/ISSUE_TEMPLATE/.

3) Copy this notebook across to your course repository. You will need to edit the values in the cell below Course specific inputs.

4) Copy the Github action to your repository folder .github/workflows/. You will need to update two fields at a minimum: the author name and the path to this notebook.

Course specific inputs

course_repo = "nu-cem/CompPhys"
issue_template_path = '.github/ISSUE_TEMPLATE/python-basics-checklist.md'
issue_template_label = "python_checklist"
output_file_path = "../images/progress_report.png"

Prepare system

Install libraries, get Github authentication token and create tasks list.

from github import Github
import matplotlib.pyplot as plt
import numpy as np
from dotenv import load_dotenv
import os
import re

load_dotenv()
GH_TOKEN = os.getenv("GH_TOKEN")  # this is needed to interact with the Github API

Access course repo

g = Github(GH_TOKEN)

repo = g.get_repo(course_repo)

Get checkbox items from the issue template

issue_template_content = repo.get_contents(issue_template_path).decoded_content.decode()

tasks = re.findall(r'\[ ] (.*)\n', issue_template_content)

Retrieve and parse raised issue information

checklist_issues = repo.get_issues(labels=[issue_template_label])

totals = [0]*len(tasks)
issue_count = 0

for issue in checklist_issues:
    issue_count += 1
    body = issue.body
    for i,task in enumerate(tasks):
        if task in body:
            splits = body.split(task,maxsplit=2)
            if splits[0][-4:-1] == '[x]':
                totals[i] += 1
        else:
            print("problem: '{}' not in string in issue # {}".format(task, issue.number))

Plot results

fig, ax = plt.subplots(1,1,figsize=(20,len(tasks)))

ax.barh(np.arange(len(tasks)),totals, align='center')
ax.set_yticks(np.arange(len(tasks)))
ax.set_xticks(np.arange(issue_count+0.1))
ax.set_yticklabels(tasks)
ax.invert_yaxis()
ax.set_xlabel('# of checked boxes')

for item in ([ax.title, ax.xaxis.label, ax.yaxis.label] +
             ax.get_xticklabels() + ax.get_yticklabels()):
    item.set_fontsize(20)

plt.tight_layout()
plt.savefig(output_file_path,dpi=250)
plt.show()