Thesis

code for my PhD thesis “State Space Models for Epidemiological Surveillance of COVID-19 - Computational Methods and Application to German data”

Introduction

This repository contains the code and latex files to build my PhD thesis. The code heavily relies on my package isssm, extending it to non-standard use-cases.

You can find the most recent version of my thesis here.

The code is developed in Jupyter notebooks and organized by the chapters and sections of the thesis. Simulations and analyses are run with Python, figures are created with R, using ggplot and tikzDevice to create tikz graphics.

Reproducibility

Data

The required raw data to run the notebooks is available through two means:

  • smaller files (< 100MB) are directly checked into this repository and
  • larger files are available through Zenodo.

Additionally, all downstream data (e.g. model results used for figures) are also available on Zenodo.

Code

This project uses the Python package manager uv. To run the code in Python Jupter notebooks, run the following commands

# install uv
curl -LsSf https://astral.sh/uv/install.sh | sh

uv sync

This creates a Python virtual environment under .venv. Use this virtual environment in all Jupyter notebooks.

For reproducibility of R code, this package uses renv. Install renv and run

renv::restore()

to install packages necessary to re-create the figures.