Skip to main content
main-content
Top

About this book

Art, or Science? Which of these is the right way to think of the field of visualization? This is not an easy question to answer, even for those who have many years experience in making graphical depictions of data with a view to help people understand it and take action. In this book, Graham Wills bridges the gap between the art and the science of visually representing data. He does not simply give rules and advice, but bases these on general principles and provide a clear path between them

This book is concerned with the graphical representation of time data and is written to cover a range of different users. A visualization expert designing tools for displaying time will find it valuable, but so also should a financier assembling a report in a spreadsheet, or a medical researcher trying to display gene sequences using a commercial statistical package.

Table of Contents

Frontmatter

Chapter 1. History

Abstract
Measuring and understanding time has been a concern for the human race since before recorded history. Whether it is on the scale of years – knowing when to plant and when to reap, when the rains or floods should arrive – or on a smaller scale, being able to distinguish when events occur during the day and relate them to an accepted means of measurement, the ability to record and measure the passing of time, has always been important. With measurement came visualization, communicating the measurements from one person to another. This chapter looks at how time has been visualized over the centuries, selecting representative or important examples and discussing them not primarily for their historical content but for what we can learn from them in our overall goal: to display data in a way that leads to action. The great advantage of studying historical visualizations of time is that the goal of actionability was always at the forefront of the designers’ minds. Unlike now, when graphs and charts can be created with ease, previously the creation of a visualization was a time-consuming proposition, and so the charts we see had a solid purpose to them. Another advantage of studying historical visualizations of time is more Darwinian in nature: Bad visualizations have not generally survived the centuries, and so we are left with the cream of the crop from which to learn.
Graham Wills

Chapter 2. Framework

Abstract
This chapter lays the groundwork for the chapters to come. In it is described the basics of how a visualization is constructed, giving a rough overview of the data pipeline and describing the components of a general visualization. I define a visualization in terms of a set of orthogonal features that can be composed to form a complete visualization. These “building blocks” include the coordinates system; the type of element (line, point, bar); color, shape, and other mappings from data to appearance; statistics; and paneling. These definitions allow us to talk about visualizations sensibly and precisely, as well as to classify visualizations by their features. In later chapters we will see how different choices of features serve different graphical goals, and we will then be able to compose the features we want to serve our goals, resulting in an effective visualization.
Graham Wills

Chapter 3. Designing Visualizations

Abstract
Study without reflection is a waste of time; reflection without study is dangerous. — Confucius , Analects (551–479 bce)
Graham Wills

Chapter 4. Types of Data

Abstract
The most important part of a visualization of data is the data themselves. If it does not show the data clearly, a visualization can be pretty, interesting, technically well executed, and all kinds of other good things, but it fails in its basic task. In this chapter we will look at data, and time-based data in particular, with a view to understanding how they can best be represented. Different forms of data require different visualization techniques. Just as a photographer will use different equipment to photograph a Formula-1 race from that used in their still-life studio, a good visualization designer will adapt his tools to the data, producing a portrayal of the data that faithfully and naturally represents the data.
Graham Wills

Chapter 5. Time as a Coordinate

Abstract
This chapter deals with the most common method of presenting time in a plot of data – using it as a coordinate to locate elements of the graphic. The time plot, with time on the horizontal dimension and a continuous value plotted as a moving line on the vertical axis, epitomizes the power of this principle. Most books on time series analysis will lead off with time series displays, inviting the reader to observe them and then using the plots to motivate the following analysis. In this chapter we will start at a more basic level – the 1-D chart with time as a single dimension. Variants and decorations for this very simple chart will be discussed; aesthetics that map other variables play a key role, as do positional modifiers that stack elements and the use of additional elements to indicate patterns. Sometimes a second dimension will be used, straying into the territory of the next chapter, but the thrust of this chapter is to explore how time can be mapped to a single coordinate so as to show patterns and enable action.
Graham Wills

Chapter 6. Coordinate Systems, Transformations, Faceting, and Axes

Abstract
The previous chapter looked at the use of time as a single, 1-D coordinate, presented as a simple line. This chapter goes beyond that use to present a range of additional options and techniques for the use of time as a coordinate. Beyond one dimension, we will look at using time as part of a multidimensional coordinate system, introduce transformations of such systems, and consider time displays within a faceting and using time data to facet other charts. We will consider details of the presentation such as aspect ratio, axes and tick marks, and scales. In short, this chapter explores how to make the most of an already powerful technique, time as a coordinate.
Graham Wills

Chapter 7. Aesthetics

Abstract
Time is often one of the more important variables to be displayed in a chart, and it can be placed as a positional dimension, giving it a high visual impact and affording the user the strongest ability to make comparisons. However, in other situations time is not the focus of a visualization – the goal is to show some other features of the data instead. In such a situation time may be of secondary importance; we do not want to make large changes to the chart to display time, but we do want to show it as an additional feature that may help viewers understand the data better. In such a case, we can portray time using an aesthetic, a visual property of a graphical element that represents a variable. In this chapter the focus of the discussion is on enhancing a visualization that is nontemporal in nature by adding an aesthetic that shows time.
Graham Wills

Chapter 8. Transformations

Abstract
Mathematically, time is typically treated as being continuous. It can be divided as finely as we desire, and we usually consider each section of time as being of the same importance as any other section of the same length; we do not think of certain times as being intrinsically more important than others. For many purposes this concept of time – as a linear, continuous dimension – is all that is needed to visualize time effectively. This chapter deals with other purposes: when we want to distort time to give more importance to some regions, when we want to analyze time in terms of frequencies, and when we want to divide time into discrete chunks. The transformations suggested in this chapter should be used to modify the mappings between variables and coordinates, aesthetics and facetings, and since the transformations are often parameterized, those parameters can be put under interactive control to allow a range of transformations to be explored under user control. Transformation is an optional step in the pipeline between data and visualization, but it is a common need and can open up a broader set of possibilities for effective visualization of time data.
Graham Wills

Chapter 9. Interactivity

Abstract
Interactivity is a fundamental property of items in the real world. We expect to be able to turn things on, move them around, change locations, or at the very least shift our viewpoint so as to see them in a different way. We color objects, resize and reshape them, and interact with them in many ways. Exploration via interactivity is a basic children’s activity, leading to understanding. It is a key way of learning we encourage in children and is no less valuable for adults. An old Chinese adage, often attributed to Confucius, states “I hear and I forget. I see and I remember. I do and I understand.” That is the basic premise of interactivity. It leverages the way we achieve understanding in the physical world, building on the visual (“I see and I remember”) and enforcing it with actions that allow us to manipulate displays and so improve our knowledge of the data that are being visualized (“I do and I understand”). Interactive data visualizations allow users to see pop-ups and tool tips; rotate, pan, and zoom in on plots; selectively show details in some areas while maintaining an overall view of the context; drill down into subareas of the data space; manipulate parameters such as degree of smoothing; select parts of one view and see corresponding items highlighted in another view; and dynamically filter data with controls or via direct manipulation.
Graham Wills

Chapter 10. Topics In Time

Abstract
While most of this book has been aimed at providing general advice, this chapter looks at a set of specific topics. The principles explained in previous chapters are used throughout this one, but the purpose of this chapter is not to give general insights, but instead to provide details on how to visualize time in two specific contexts. The first context is that of large data sets. If large volumes of data are being collected, it is almost certain that there is a temporal component – after all, there is typically a limit on how much data can be collected all at once, and even if you are only collecting a few rows of data every minute, then those intervals add up over the decades. The second topic concerns time events that have strong structural relationships, and where the main goal is to visualize the relationships between the data items rather than the data items themselves. In most of the book our goal has been to discover patterns in the data; this chapter focuses on presenting known patterns so as to learn from them.
Graham Wills

Chapter 11. Gallery of Figures

Abstract
This section is concerned with simplicity and its opposite – complexity. To make an effective and useful chart, a visualization should be the least complex one that faithfully portrays the data. In this section an approach to identifying complexity is explored and the results are used to provide an ordered list of all the figures in this book. Rather than the traditional combinations of titles and page references, this chapter provides a graphical thumbnail list of the figures. The figures are not ordered by their sequence in the book (which would be the natural default time-based ordering), but instead by an index of the grammatical complexity of the figure – a value relating to the size and complexity of the coordinate system, number of aesthetics, and presence of faceting. The simpler figures appear first, and the more complex ones later. This chapter is motivated by the belief that, to a great extent, the complexity of a visualization can be defined by the grammatical features of that visualization.
Graham Wills

Backmatter

Additional information

Premium Partner

    Image Credits