Appointment Scheduler

This repository contains a MILP for an appointment scheduler developed for the 'seminar applied optimization' at University of Bern.

Note that in the source code, the entity for a service is 'Queue'. This is because I want to add some properties of queues to make the time-slot assignments more fair.

You can find more details in the presentation file.

this project uses python 3.11. If you don't have python 3.11 installed, you may want to run it in docker (see at the bottom of this file)

How to use

First, make sure you have a (virtual) environment up and running. Pipenv the preferred way. In this case, just run

pipenv install

For any other case, there is also a generated requirements.txt file

Run the optimizer with random data

python main.py

Reproduce my experiments shown in presentation

python scenario_concentrated_on_500.py
python scenario_no_preferences.py
python scenario_random_concentration.py
python scenario_random_intervals.py

You will get the solution as console output and plots (see the out/ directory!)

Misc

Run within docker

It is also possible to run within docker. However, if you don't map a directory to out/, you will not be able to view the plots.

First: Build image

docker build . -t appointment-scheduler

Then, run either completely isolated

docker run -it --entrypoint bash appointment-scheduler

Or with a directory (here, your current working directory) to the out directory to get the plots

docker run -it --entrypoint bash -v ${PWD}:/app/out/ appointment-scheduler

Create/Update requirements.txt

Pipenv is used for managing dependencies. However, we can extract a requirements.txt file for users who would like to use another environment.

pipenv requirements > requirements.txt