It does what it says on the tin.
$ python3 src
Usage:
cli.py <file> [--num-weeks=<n>] [--max-in-hours-shifts=<n>] [--max-on-call-shifts=<n>]
cli.py (-h | --help)
$ time python3 src demo.csv
week,primary,secondary,shadow,primary_oncall,secondary_oncall
1,Neil Hockenberry,Aubrey Staiger,Sharleen Woltz,Oswaldo Bonham,Jarrett Hord
2,Deloris Baldon,Rocco Morra,,Martin Ashby,Robin Hoose
3,Emanuel Leinen,Vernon Minelli,,Lou Meidinger,Renae Paton
4,Floyd Olsson,Irwin Capehart,,Dewey Burgett,Jerold Bayes
5,Pierre Paulhus,Sammie Shew,,Lou Meidinger,Wilson Friesen
6,Jerome Silveria,Benita Kunz,Glynda Laubscher,Oswaldo Bonham,Temeka Lowy
7,Jame Truss,Nyla Drozd,,Annalisa Harrow,Renae Paton
8,Ryan Averett,Jerald Vangundy,,Jeannine Demos,Jerold Bayes
9,Theodore Calvery,Temeka Lowy,Chas Stucky,Dave Allred,Grant Kornfeld
10,Martin Ashby,Lacy Auyeung,Ramon Haddock,Annalisa Harrow,Oswaldo Bonham
11,Bessie Engebretson,Wilson Friesen,,Jeannine Demos,Jerold Bayes
12,Dave Allred,Lou Meidinger,,Annalisa Harrow,Santiago Raine
real 0m24.531s
user 0m24.408s
sys 0m0.121s
See the docs
directory for explanations of how the rotas are defined
and generated.
You need Cbc (Coin-or branch and cut) installed and in
your $PATH
. Other dependencies are listed in
requirements-freeze.txt
and can be installed with pip
.
Rather than install dependencies to your host machine, you can generate a rota in a Docker container:
$ docker run -it --rm -v $(pwd):/src -w /src python:3.9 bash
$ apt-get update && apt-get install -y coinor-cbc
$ pip install -r requirements-freeze.txt
$ python3 src demo.csv
This uses an approach called integer linear programming (ILP), via the PuLP library. A reasonable introduction to ILP for solving scheduling problems like this is the PyCon conference-scheduler docs.
I wrote a memo going into some detail about how it all works.