Skip to content
Linear programming model for class schedule generation
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
sample_output
sample_preference_files
sample_schedule_templates
sample_students
sample_teachers
.gitignore
LICENSE.txt
README.md
check_schedule.py
course_selection.py
format_html_schedule.py
jam_in_course.py
solve_schedule.py

README.md

Course Scheduling Utility

Python script to generate course schedules for a small private school. Given a schedule template (listing teachers, timeslots, and exclusions, as in sample_schedule_templates/empty_sm.txt) and a set of preferences (that is, lists of courses offered by each teacher, as well as classlists listing the students that would prefer to take each course, as in sample_preference_files/test.txt), the script generates an Integer Linear Programming model and solves it using CoinMP in order to minimize the number of schedule conflicts students experience (in aggregate).

(Other utilities are included, such as jam_in_course.py, which was developed to assist in determining how to split a class into two sections, or where to add a new course without having to recalculate the entire schedule.

For advanced users, lines 182-183 in solve_schedule.py can be edited in order to identify a course to de-prioritize. The logic behind de-prioritization is that certain courses may be essential to a student's graduation, whereas others are optional extras; conflicts involving the optional courses could be weighted slightly less in order to ensure the scheduler prioritizes working on the essential courses.)

Requires PuLP and CoinMP to be installed. (PuLP has other backends besides CoinMP, but only CoinMP has been verified to work with this program.) Run

./solve_schedule.py --help
./check_schedule.py --help

for an explanation of the options.

How to run, e.g. on the sample data provided:

./solve_schedule.py -H output.html -p sample_preference_files/test.txt sample_schedule_templates/empty_sm.txt -t 5

Or, alternately:

./solve_schedule.py -H output.html sample_students sample_teachers sample_schedule_templates/empty_sm.txt -t 5

This produces a file similar to sample_output/output.html. Using a longer duration (-t option) can produce a schedule with fewer conflicts.

Provided under the terms of the MIT License, as stated in the file LICENSE.txt.

Setting things up

Get the Pulp-OR LP modeler library. Download and extract the 1.5.4 tarball and run python setup.py install (possibly as administrator).

Get CoinMP. For instance, download a Linux tarball and follow the installation instructions, or get the Brew package on OS X using brew install coinmp.

Get the progressbar Python module, e.g. sudo easy_install -U progressbar.

You can’t perform that action at this time.