Data Visualization

Robert Grant is a British statistician specialising in data visualization and Bayesian models. He worked in biomedical research and taught statistics at St George's Medical School, Kingston University, the Royal College of Physicians of London, and the National Institute for Health and Care Excellence before launching his own training and coaching business, BayesCamp, in 2017.

"Excellent!" –Andrew Gelman"… Data visualization is a very useful tool for this, but only when used well. That is what this book offers: visualization techniques, a link with statistics, principles for designing a good visualization, and lots of examples. The author aims to teach a way to design graphics that have an impact. The author is a statistician and this is reflected in the way he approaches the topics: for example, there is no display of error bars without explaining what they represent …. Examples are the backbone of the book, each chapter has one or more visualizations to compare what the advantages are to present data a certain way … without forgetting the reasoning behind it, whether it be statistical or visual). This makes the book very readable and appealing."—Thibaut Cuvelier, Developpez.com, June 2019"In [this book], Professor Grant gives an overview of a wide range of techniques and challenges, while staying accessible to anyone interested in working with and understanding data. [It] includes chapters on big data, statistical and machine learning models, visual perception, high-dimensional data, and maps and geographic data; Contains more than 125 visualizations, most of which were created by Professor Grant; And is supported by a website with all code for creating the visualizations, further reading, datasets and practical advice on crafting the images. … [This book] is especially recommended reading for students considering a career in data science, analysts who wants to learn more about visualization, and managers of teams working with data. Introducing a broad range of data visualization methods, [this] is a valued and unreservedly recommended addition to college and university library collections." –Micah Andrew, Reviewer, Midwest Book Review, January 2019"This book gives a large number of interesting tips and suggestions for anybody (statistician or not) who wants to use graphics to represent and convey the information contained in their data. Before going into a more detailed analysis of the strengths and weaknesses of this book, it is worth remarking on its very broad descriptions, analysis and overview of graphics, not so much from a statistical and mathematical point of view as a cultural point of view, considering also the physiology of the brain... Comparisons of important concepts are very well presented through clear and effective examples and graphs: for instance, differences between correlation and causation, joint and marginal distributions, interpretation and use of counts and percentages, means and medians (even if some graphs compared in the examples do not have the same axis scales, neither in the paper version nor in the online one), meaning and interpretation of the effect of confounders and interactions in a statistical model. An intuitive approach is used to explain several complex concepts like artificial intelligence and convolutional neural networks. Some common mistakes (sometimes due to misunderstanding) are presented and commented on in a very clear way...this book is highly recommended to researchers of all fields who want to learn about how to start a data analysis in a proper way, how to present their data and results clearly and attractively, and how to avoid mistakes that can jeopardize their work."- Silvana Romio, University of Milano-Bicocca, Appeared in ISCB News, 2020

• Section I The basics 1. Why visualise? 2. Translating numbers to images Section II Statistical building blocks 3.Continuous and discrete numbers 4.Percentages and risks 5. Showing data or statistics 6. Differences, ratios, correlations Section III Specific tasks 7.Visual perception and the brain 8. Showing uncertainty 9. Time trends 10. Statistical predictive models 11. Machine learning techniques 12. Many variables 13. Maps and networks 14. Interactivity 15. Big data 16. Visualisation as part of a bigger package Section IV Closing remarks17. Some overarching ideas