mbgrader (math batch grader) is a custom web application for batch grading assignments where students submit numbers and matrices as .csv files, or text in .txt files.

flask=1.1.1
flask-sqlalchemy=2.4.0
numpy=1.16.4
pandas=0.25.1
python=3.7.4
sqlalchemy=1.3.7
sqlite=3.29.0

Setup the database:

python init_db.py

Setup the Flask environment variable and run:

export FLASK_APP=app
flask run

Open a browser and navigate to http://127.0.0.1:5000/. Create a new assignment by entering the corresponding folder name locating in the submissions folder. There is an example assignment called example1. mbgrader assumes the folder structure:

submissions/
    assignment_name/
        student_id_1/
            var1.csv
            var2.csv
            var3.txt
        student_id_2/
            var1.csv
            var2.csv
            var3.txt

Create a new question and enter a preprocessing function if necessary. Requirements for the preprocessing function are:

• Function must be named fun.
• Input parameters:
  • s is the student number: int
  • r is the response: str, np.float64, or np.ndarray
• Return value must be the same type as the response.

For example, the preprocessing function for the question "What is your student number?":

import numpy as np

def fun(s,r):
    if isinstance(r,str):
        return "text"
    elif isinstance(r,np.ndarray):
        return np.array([0,0])
    else:
        if abs(r - s) > 0:
            return np.float64(1)
        else:
            return np.float64(0)

Create a new response from student responses by entering:

• New response variable name
• Comma-separated list of variables used to compute new response
• lambda to compute new variable where input is a list and entries are the selected variables
• Extension to indicate datatype

The script canvas2mbgrader.m reads MATLAB .mat and .fig files from the canvas folder and saves the data to the submissions folder. The script requires a file canvasIdstudentID.csv with Canvas IDs in the first column and Student IDs in the second column.

The script mbgrader2canvas.py reads the grades and feedback produced by mbgrader and submits it to Canvas. To do this, it will need a token from Canvas; you can generate a token from Canvas by going to 'Account', 'Settings', and clicking 'New Access Token'. This token needs to be stored in the top of the mbgrader directory in the file 'token.txt'. When running this script, it will request the Canvas Course ID, the Canvas Assignment ID, and the Assignment Name. The Canvas IDs can be obtained from the URL of the assignment submission page, e.g. https://canvas.ubc.ca/courses/12345/assignments/9876543 has a course ID of 12345 and an assignment ID of 9876543. The Assignment Name is what you called the assignment in mbgrader.