gradingmagic is a suite of dumb little scripts that help to automate the grading &
feedback process on D2L.
D2L has a bulk-feedback submission mechanism, but to be truly useful it seems to require that the files containing feedback from the instructor must have the same name as the corresponding student's original submission.
Run gradingmagic once on a directory full of unzipped student submissions, and
it will create a .md file for each one. Edit each .md file with your
feedback, and when you're done, run gradingmagic again to generate a properly
formatted .zip that you can submit back to D2L using the Add feedback files button in the dropbox.
To make these easiest to use, it is recommended you add them to your $PATH.
pandoczip
Usage: gradingmagic [path] [output_tag]
path is optional -- if no path is given, the script will default to the
current directory.
output_tag will appear in the name of the output files, and is always appended
with _feedback. It will default to whatever the basename of path is, if no
value is given. Note that this is bash, and I haven't included support for
position-independent parameters, so to use output_tag, you need to specify a
path.
Creates empty Markdown templates, and converts them to PDF after you've edited
them, and then zips those PDFs up into a zip file that you can upload to D2L.
The empty templates are created in the path directory, and in order for them
to be compiled, the must be moved to [path]/feedback. The feedback tool can
automate the process of editing and moving them to [path]/feedback, or you can
just run mv *md feedback/.
Usage: feedback
No arguments. Must be run from whatever your equivalent is of the local Lab1
folder in the example workflow below. It will open a students submission
automatically (using xdg-open, although this may need to be replaced with
gnome-open or something depending on your system) so you can grade it, as well
as opening the corresponding Markdown template in VIM. When you close a file
with VIM, it is immediately moved to [path]/feedback. You can quit using
feedback and any point and pick up where you left off, since the Markdown
files are moved to [path]/feedback as you finish editing each one.
.md to [path]/feedback
Usage: gengrades [feedback_path] [exported_grades_csv] [max_points]
gengrades generates a CSV file with grades that are extracted from the
feedback Markdown files that you create. exported_grades_csv must be obtained
from D2L -- it contains unique identifiers for each student that must be used in
any CSV you want to upload to D2L.
gengrades uses the name of feedback_path's parent directory as the name
of the grade item.
As of now, it supports numerical grades only.
See example_feedback.md for an example feedback Markdown with grading syntax for automatic final score calculation.
Imagine you have a dropbox folder called Lab1, with a maximum point value of
30 points.
- Download zipped submissions for dropbox titled
Lab1. - Unzip contents of zip file to local directory called
Lab1 cd Lab1/gradingmagicfeedbackcompilefeedbackgengrades ./feedback [path to CSV grades exported from D2L] 30<-30is the max points for this lab.
To avoid duplicate feedback items, it's best if you set your D2L dropbox to only allow one student submission (i.e. new submission overwrites older submission).