The Covid-19 Open Visualization (COViz) project is a modular open-source web toolkit for hybrid map/graph visualization of COVID-19 risk burden, resource and testing sufficiency and trajectories over time. In contrast to many dashboards that are data-driven and/or built on proprietary software, COViz is an efficient, lightweight open platform that focuses on use of sound cartographic principles to construct targeted visualizations that intuitively communicate information relevant to specific epidemiological questions. These include:

• What is the current risk of transmission?
• What is the recent trajectory of disease burden and resource availability?
• How great is the cumulative burden of disease?
• Are resources and testing sufficient?

To visually communicate answers to these questions, COViz aims to combine adherence to traditional cartographic principles with innovative map forms and easy-to-use animation/interaction components.

Traditional Cartographic Principles

• All maps communicate rates per population. Raw case counts, hospitalizations and deaths are meaningless without the context of the population within which they are occurring.
• Rates are expressed using sequential color schemes, following the principles of choropleth mapping.
• Raw counts, when shown, are expressed using proportional symbols, as symbol sizes are intuitively associated with amounts.
• District polygons are highly generalized so that the focus is on the thematic information, not the intricacies of the coastline.

Innovative Map Forms

COViz features the use of cartograms. Also known as “population maps” and “density equalizing map projections”, cartograms have been used in epidemiology since at least the 1920s to display districts at a size proportional to their underlying population. This gives a better sense of the relative distribution of disease within the population. While they take a little getting used to, cartograms enable certain advantages for visual inference:

• When symbols (e.g. circles) are used to show raw case counts, the case rate (cases per population) can be inferred visually by the proportion of each district occupied by the circle.
• When sequential colors (e.g. light to dark) are used to show case rates, the total number of cases is proportional to the darkness value multiplied by the size of the state.
• Symbols on a cartogram will have less overlap than on a regular map.

Animation and Interaction

Some cartographic techniques that work well on static maps break down in animated sequences. We are currently experimenting with ways to enhance visual clarity on the animated timelines.
This is an open-source project. You are free to download, customize, modify data sources and incorporate into your own website, with the caveat that no warranty is provided.

Keyboard shortcuts (these work inconsistently at the moment):

T	Place focus on the theme selector
← →	Move along timeline (when focus off var. sel.)
L,P	Toggle between land and population map types

(click on any state for data values)

This website is being developed by the EIU GIScience Center. Data comes from covidtracking.com, the COVID-19 Canadian Open Data Working Group, and the Center for Systems Science and Engineering at John Hopkins University. Coding performed by:

	Barry Kronenfeld Director, GIScience Center, Eastern Illinois University
	Kwang-il Yoo (Jason) PSM in GIScience, Eastern Illinois University
	Sushma Saragadam PSM in GIScience, Eastern Illinois University
	Sarah Kronenfeld University of Toronto

Latest project updates can be found on our github repository. This is a work in progress. Feel free to send comments or feedback.