README.md

Autograder for EngComp

Autograding procedure for local grading using nbgrader. Setup to be used with Blackboard.

Workflow

1. Instructor creates assignment in a Jupyter notebook and uses nbgrader formating for autograding.
2. Generate the assignment via createAssignment() function within gradingFunctions.py.
3. Distribute assignment to students via Blackboard. Set to a single submission. ** code should take the newest submission if there are multiple - has not been tested **.
4. Collect assignments on Blackboard.
   • Generate a .zip file with student submissions (expecting a .txt and .ipynb per student).
   • Download a gradebook file with the specific assignment selected (expects a .csv).
5. Autograde the assignment using the autogradeAssignment() function in gradingFunctions.py. Assignment results will be displayed and saved into the scratch folder.

nbgrader Directory Structure

This will automatically be properly made when using the setupCourse() function, however, there are many nbgrader features that will not work because the must be configured manually within the nbgrader_config.py file.

Sample directory structure should look like:

opath/                      # overhead directory
├── how-to-grade.ipynb
├── gradingFunctions.py
├── coursename/
│   ├── nbgrader_config.py
│   ├── gradebook.db
│   ├── autograded/
│   │   ├── student_1/
│   │   │   └── ps1/
│   │   │       └── ps1.ipynb
│   │   └── student_2/
│   │       └── ps1/
│   │           └── ps1.ipynb
│   ├── feedback/
│   │   ├── student_1/
│   │   │   └── ps1/
│   │   │       └── ps1.ipynb
│   │   └── student_2/
│   │       └── ps1/
│   │           └── ps1.ipynb
│   ├── source/
│   │   └── ps1/
│   │       └── ps1.ipynb
│   ├── release/
│   │   └── ps1/
│   │       └── ps1.ipynb
│   └── submitted/
│       ├── student_1/
│       │   └── ps1/
│       │       └── ps1.ipynb
│       └── student_2/
│           └── ps1/
│               └── ps1.ipynb
└── temp/

Note This should not have to be manually made - after the first assignment has been made and graded, every file should appear.

Note About nbgrader's GUI:

formgrader is still fully functional (besides exchange-setup), however to access it for your course you must be located in the coursename directory before launching Jupyter Lab or Jupyter Notebook.