flusim

Overview

This repository holds the files for the flu-sim Python package. This is a package used for simulating, calculating and visualizing some statistics related to a simple flu-spread problem.

For example, one can visualize a single flu spread:

or simulate a flu spread multiple times and plot the computed confidence intervals:

To use this package, run

pip install -i https://test.pypi.org/simple/ flusim

To customize upon this package, clone this repository, and run:

python setup.py bdist_wheel sdist
pip install .

Usage

For usage, please refer to this link.

For a demo, you may refer to this Jupyter notbook or preview the generated HTML file.

Problem Definition

The flu-spread problem is defined as follows. Suppose inside a closed environment, there are:

1. n + 1 people in it, labeled 0, 1, ..., n
2. no people sick before day 0, and k people sick at day 0
3. on each day, a sick person has a probability p contaminating a healthy person
4. these events are independent
5. each person once get sick, will get healthy after precisely l days

then the problem is to calculate the expectation, variance of:

1. the end date of flu
2. the number of sick people in each day

This problem can be formulated as a Markov process, and simulations, solvers can be designed to solve this question.