Description

This utility can be used to download data about courses and student submissions from Canvas Learning Management System by Instructure.

Getting Started

Dependencies

To run the program, install the dependencies listed in the requirements.txt file:
pip install -r requirements.txt --user
Note that the requirements include CanvasAPI, which is the library used in this utility to download json data and files.

Canvas API

To export data from Canvas LMS you need two pieces of information:

• API KEY, which can be generated in your Canvas log in page under Profile > Settings > Approved Integrations
• API URL i.e. your organization's Canvas Base URL, which is the URL of your Canvas LMS page where you log in to

You need administration privileges on the courses you want to export.

Running the script

Export all data:

python export.py

Print out course, assignment, submission IDs and files without downloading them:

python export.py --no_download

Optional Configuration

This script will prompt the user for getting the API URL and API KEY. However, you can configure create a .env in this repository using the example file example_dotenv: cp example_dotenv .env

The parameters that can be configured in .env are:

• Canvas API URL
• Canvas API key
• Output directory (default: "export")
• Comma separated list of Course IDs that should be skipped
• Comma separated list of Assignment IDs that should be skipped

Tip: run python export.py --no_download to get the Course IDs and Assignment IDs without downloading them

Export

The exported data include raw json data that can be downloaded using Canvas API, and submission attachments to submissions. In particular, the exported json files include: courses, assignments, submissions. The rubrics and full rubric assessments are included in the exported files

Reading the data using Python

Check the Process_export.ipynb notebook and an example of dataframes extracted to the pickle file example_dataframes.pkl:

with open('example_dataframes.pkl', 'rb') as f: 
	df_files, df_rubrics, df_rubricsubmitted = pickle.load(f)

Possible further development

• Consider to save announcements, discussions, pages, modules, other attachment files
• Consider to implement anonymization

Credits

David Katsandres's script canvas-student-data-export.