Assignment Feedback

Prerequisites

Python

This tool is written in python (v3.11) and has thus to be run with python 3.
It has the following requirements:

Packages:

• click
• dash
• dash-bootstrap-components
• flask
• openpyxl
• rich
• pandas
• selenium
• tabulate

The packages may be installed all at once with the file requirements.txt:

pip install -r requirements.txt

or individually:

pip install click
pip install dash
pip install dash-bootstrap-components
pip install flask
pip install openpyxl
pip install rich
pip install pandas
pip install selenium
pip install tabulate

Note: Both approaches need to refer to the pip-installer associated to the python installation, that will be used to run the tool.

Input

Input through Web service - (Recommended)

Download all submissions from ILIAS (Please make sure your language setting of ILIAS is English or Deutsch). An xlsx file will also be created. Specify the tasks and their points for each assignment in your config (see example: example/config_example.txt), and run the script like this to start the web server:

> python3 assignment_feedback.py -m websever -l <lecture-marker> -o <directorypath> -c <filepath>

This will create a database file named -l <lecture-marker>.sqlite3 in -o <directorypath>. Then click Add Assignment button on the web server and upload Assignment ?.xlsx, which you obtained from ILIAS.

NOTE: If the assignment already exists in the database, the system will prompt you whether to overwrite it.

You may merge the gradings from your colleagues. Click Merge Gradings button on the webpage, then upload the sqlite3 database file. This will (only) merge (=> overwrite) the incoming gradings for the currently selected assignment! It is thereby also only possible for existing tables, i.e. it is not advised to attempt to merge assignment tables that don't exist in the active webserver instance yet. In this case one may either use the other database as basis for the webserver and merge what's needed of the current database into it, or create a new empty table in the current database as a foundation using a Assignment ?.xslsx-file.

After all the gradings for one assignment have been finished, click the Generate Feedbacks button to download the feedback files. You will receive a zip file containing transcribed feedback Markdown-files for each team submission. Now you can stop the web server and upload the feedback files to ILIAS. See Automatic upload.

Manual Input through legacy mode - (Optional)

If you wish to use the tool without the web server, you require multiple CSV-files, each associated with an assignment. The CSV-files should each contain the reached points as well as the associated feedback of each task (see example: example/grading_example.txt).

> python3 assignment_feedback.py -m legacy -l <lecture-marker> -o <directorypath> -c <filepath>

Input CSV Formatting

The provided information on points and feedback should be split by a comma and each surrounded by quotation-marks (to allow for the use of regular comma without breaking the format):

• List of all group members:
  "<member_1>,<member_2>,<member_X>"
• Feedbacks for each task (3 options)
  • simply returning points reached:
    "<total_task_points_reached>"
  • returning points reached with single-line comment:
    "<total_task_points_reached>:<points_substracted> <comment>"
  • returning points reached with multiple lines of comments (by splitting with a '|'-symbol):
    "<total_task_points_reached>:<first line pts_substr> <first line comment>|<second line pts_substr> <second line comment>"

The enumeration, the paths, and the maximum reachable points of each task need to be specified and given in a separate configuration file (see example: example/config_example.txt).

Output

This tool returns all created feedbacks for the given assignments in Markdown format (see example (for legacy mode): example/ass1).

Automatic upload

Finally, this tool allows to automatically upload the resulting feedbacks onto Ilias, by running:

> python3 assignment_feedback.py -m feedback -u <feedback directorypath>

For this to work, the feedback files in the specified directory should contain the team's id as the prefix of each file's name (webserver outputs should have those automatically).

Example

The webserver pipeline of this tool could look like this:

# Start webserver and setup database in outputdirectory (here: 'outs')
> python3 assignment_feedback.py -l lecture -o outs -c lecture_config.txt
# TODO: Use GUI to load xlsx-files downloaded from Ilias, input points and comments, then create and unpack feedback archive.

# Upload feedback to Ilias by specifying the unpacked directory
> python3 assignment_feedback.py -m feedback -u 'Feedbacks_Assignment X'

The legacy pipeline for this tool could look like this (requires specification of csv-files (see format specifications) in config.txt):

# For legacy usage
> python3 assignment_feedback.py -m legacy -l lecture -o outs -c lecture_config.txt
# TODO: Manually add team id to created output feedback file's names as prefix!

# Upload feedback to Ilias by specifying the renamed feedbacks' directory
> python3 assignment_feedback.py -m feedback -u 'outs/assX'

Authors

• Alexander Röhl
• Haoran Sun