Leland Wilkinson's "Grammar of Graphics" (GoG) is a must-read for any environmental chemist who communicates data using graphics on a regular basis. Rather than thinking about graphics using pre-conceived templates (e.g., the "scatterplot", the "bar graph", or the "boxplot"), Wilkinson decomposes graphics into several orthogonal components: data, scales, statistics, geometries, and facets. In most implementations of the GoG, data comprise a table with one row for each object (e.g., a sediment sample), and one column for each variable (e.g., from which lake the sample was collected, or the concentration of an analyte in the sample). One or more of these variables can then be mapped to a scale, which can represent a position on the plot (e.g., x-axis or y-axis) or an aesthetic (e.g., colour or shape). Before being passed to a geometry (e.g., a point, line, or polygon), the data can optionally be transformed using a statistic (e.g., a linear regression). Finally, data can be split into multiple panels using facets (e.g., one panel for each surficial unit). Using these, Wilkinson argues that almost any graphic can be constructed. Collectively, use of the GoG in our research group has led to (1) more effective and/or innovative graphics, (2) the ability for researchers to visualize more complex chemical data than was previously possible, (3) more efficient reproducibility of graphics for multiple media, and (4) better data management practices. In this presentation we describe the process of communicating water quality data generated in several recent projects, with emphasis on the GoG and its implementation in the R language. Whether or not you use a GoG-based library, knowing the GoG makes it possible to decompose a graphic into its components and improve it, piece by piece. What makes effectively organized data? What makes an effective colour or shape scale? Could this graphic be more impactful if the data were split into multiple panels? "The Grammar of Graphics" provides a conceptual framework for answering these questions.

Communicating water quality data using the grammar of graphics

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.