Supplement A of "Power-law models for infectious disease spread"

by Sebastian Meyer and Leonhard Held

University of Zurich

This bundle contains the animations referenced in Section 4 of our paper entitled "Power-law models for infectious disease spread",
published in The Annals of Applied Statistics 8 (2014), no. 3, 1612-1639, DOI:10.1214/14-AOAS743 [pdf].

• Weekly reported influenza counts from 2001 to 2008.
• Reported influenza counts for 2008/W1-W20 only (the forecast horizon used in Section 4.3). We provide a cumulative as well as a sequential version.
• Reported and mean simulated epidemics for 2008/W1-W20 using different models. Again we provide a cumulative as well as a sequential version. Furthermore, there is a population-adjusted version of the animations, where we plot incidence per 100 000 inhabitants.
• Mean PIT-histograms over all districts by week of the simulation-based long-term forecast for 2008 for the competing models.
• Reported individiual cases of invasive meningococcal disease (IMD) during 2002-2008, separately for each of the two most common meningococcal finetypes. Constant time steps of 10 days are used.

Functions to produce such animations are available in the R package surveillance. The animations have been embedded into HTML pages by the saveHTML() function of the animation package.

The full bundle of animations can also be downloaded as a zip-archive: supplementA.zip (53 MB)

Software: A software guide for the spatio-temporal analysis of epidemic phenomena using the R package surveillance is available as arXiv:1411.0416 and will appear in the Journal of Statistical Software.

Supplement B is an extra document, where we provide details on likelihood inference in both modelling frameworks, and elaborate on the integration of radially symmetric functions over polygons. It also contains additional figures and tables of the power-law models for IMD and influenza.