Part I: From data to visualization
Mapping data onto aesthetics
Explains the basic concept of aesthetic mapping, which lies at the heart of any data visualization.
Coordinate systems and axes
Explains coordinate systems and positional scales (including nonlinear scales).
Effective use of color
Covers basic concepts of color use, as a tool to distinguish, as a tool to represent a value, and as a tool to highlight. Introduces the three basic types of color scales: qualitative, directional, diverging.
Directory of visualizations
Provides a graphical guide to the most commonly used types of data visualizations, with pointers to the relevant chapters covering them.
Visualizing amounts using bars, areas, and colors.
Visualizing paired data
Discusses common strategies for paired data, including scatter plots and paired dot plots.
Visualizing time series
Discusses common strategies for time series, including line plots and sparklines
Visualizing distributions I: Histograms and density plots
Discusses strategies for visualizing individual distributions, including pros and cons of histograms and density plots.
Visualizing distributions II: Cumulative density functions and q-q plots
Discusses strategies for visualizing distributions that are exact and non-arbitrary (unlike histograms and density plots) but are more difficult to interpret.
Visualizing many distributions at once: boxplots, violins, and ridgeline plots
Discusses strategies for visualizing many distributions, including boxplots, violin plots, jittered points, and others.
Visualizing proportions I: Pie charts, stacked bar plots, and stacked densities
Discusses stacked bar plots, stacked density plots, and pie charts.
Visualizing proportions II: Nested or changing proportions
Discusses treemaps and parallel sets (sankeys).
Discusses various approaches to smoothing data (linear regression line, GAMs, splines), and common pitfalls (many smoothers are unreliable or misleading at the edges of the data range).
Discusses error bars, confidence bands, credible intervals, posterior distributions, etc.
Visualizing geospatial data
Provides a basic intro to making maps.
Part II: Principles of figure design
Your axis labels are too small
Discusses the widespread problem of excessively small axis labels.
Choosing the right axis settings
Covers various aspects related to axis choice, including linear vs. logarithmic axes, as well as issues of axis expansion beyond the data range. Also takes up the topic of aspect ratio once again.
The principle of proportional ink
Explains that the size of colored areas needs to be proportional to the data value they represent.
Handling overlapping points
Describes different strategies to handle the problems of overlapping points or large point clouds. These problems frequently arise in large datasets, and helpful strategies include using partially transparent points, 2d density plots, hex grids, or smoothers.
Common pitfalls of color use
Also contains a section on designing for colorblindness.
Explains how to make sure that key information in the figure is provided in multiple, reduant ways, for example through color and location or color and direct labeling.
Discusses issues that arise in multi-panel figures, including proper labeling, alignment between subfigures, shared legends, and overly complex multi-panel figures.
Figure titles and captions
Discusses when to use and not to use figure titles. For captioned figures, the titles are normally the first thing shown in the caption, and thus are not shown on top of the figure.
Avoid line drawings
Argues that filled shapes and solid colors are almost always preferable to shapes shown as outlines or with hatching or cross-hatching.
Maximizing the data signal in visualizations
Discusses the general principle of maximizing the ink used to convey information. Argues to avoid boxes and frames around figure parts. Also discusses when and how to use background grids and other guide lines in figures.
Don’t go 3D
Argues why 3d plots are generally problematic (figures are fundamentally a 2D medium, and in 3D plots data is subjected to an additional, non-obvious transformation from 3D to 2D) and suggests alternatives to visualize high-dimensional datasets, including encoding additional variables in color, size, or symbol shape, and/or using faceting.
Part III: Miscellaneous topics
Understanding the most commonly used image file formats
Provides an introduction to vector and bitmap graphics and describes the pros and cons of the various most commonly used file formats.
Choosing the right visualization software
Discusses the pros and cons of different software available to make graphs.
Telling a story with data
Discusses how to compile larger sets of figures to tell a story; e.g., always move from less processed to more processed data representations; also, avoid repeating the same type of figure many times.
Provides a list of other reading material on related topics, with a brief paragraph describing the contents of each reference.