Effective Data Visualization

Through design principles drawn from graphic design, visual art, perceptual psychology, and cognitive science

Created byJef Heer

Last updated 4/2017

English

English [Auto],

What you'll learn

Visualization design fundamentals
Visual data analysis and exploration
Perceptual psychology
Cognitive science
Data reduction strategies
Design principle
Effective visualization of big data

Course content

1 section • 12 lectures • 2h 55m total length

Introduction to value of Visualization15:55
Explore the value of visualization for data analysis, revealing patterns beyond descriptive statistics and guiding exploration with scatterplots, networks, and adjacency matrices.
Data and Image Models18:34
Learn foundational concepts for encoding data visually, from data types (nominal, ordinal, interval, ratio) to visual variables like position, size, color, and texture, and explore how mappings create interpretable graphics.
Multidimensional Data17:34
Explore mapping data variables to the eight visual encoding channels and use trellis plots, small multiples, scatterplot matrices, parallel coordinates, and PCA for dimensionality reduction in multidimensional data.
Design Principles3:23
Explore perceptual design principles for effective data visualization, including the expressiveness and effectiveness principles, which guide truthful representations and faster, more accurate interpretation.
Graphical Perception10:13
Explore graphical perception and how visual encodings, including bars, circles, area, and color, affect perception, guided by Stevens power law and Cleveland and McGill experiments for better data visuals.
Color17:02
Design color palettes that encode nominal, ordinal, and quantitative data using categorical colors and lightness ramps. Consider perceptual discrimination, color naming, and color spaces like Lab for guidance.
Perceptual Redesigns6:29
Apply perceptual redesigns to visualizations by swapping color with position encoding for gene expression time series and arterial decay, improving diagnostic accuracy with 2D diverging palettes and 3D displays.
Interaction Techniques28:27
Explore how to specify data and visual mappings, transform data, and interact with visualizations through selection, brushing, linking, and dynamic queries.
Visualizing Big Data: Supporting Interaction11:33
Learn to enable fast, interactive exploration of big data with multivariate data tiles that support brushing and linking across 3D data cubes, using GPU-accelerated queries and rendering.
Analysis Example - Antibiotic Effectiveness14:40
Explore how visualization supports exploratory data analysis of antibiotic effectiveness, using data on bacteria, gram staining, and minimum inhibitory concentration to compare penicillin, streptomycin, and neomycin across species.
Demonstration: Trifacta Visual Profiling13:09
Explore visual profiling in Trifacta to inspect data quality, shape, and structure using interactive visualizations, histograms, maps, and transformations for downstream analysis.
Visualizing Big Data: Visual Encoding18:09
Explore perceptual visual encoding for big data, using sampling, band aggregation, and hexagonal or rectangular bins to enable scalable, real-time exploration and reveal outliers.

Requirements

Computer and Internet

Description

Learn the methods you need to bring data to life through effective visualizations. In this video course, host Jefrey Heer—co-founder of Trifacta—takes you through best practices for designing interactive visualizations, performing exploratory data analysis, and examining multidimensional data.

You’ll begin by learning the value of visualization, through design principles drawn from graphic design, visual art, perceptual psychology, and cognitive science. Using Trifacta’s data transformation tools to illustrate some of the concepts, you’ll also learn techniques for scaling visualizations to extremely large data sets.

In two parts—Effective Visualization Design and Visualizing Big Data—this course will teach you practical tips on topics including:

Visualization design fundamentals: data and image models, graphical perception, and color design
Visual data analysis and exploration: exploratory data analysis, multidimensional data, interaction techniques
Data reduction strategies: sampling, aggregation, and model fitting
Effective visualization of big data: scalable visual encoding methods and interaction techniques

Jefrey Heer is a co-founder and CXO (Chief Experience Officer) at Trifacta Inc. He spent many years as a professor of Computer Science at Stanford University, where he led the Stanford Visualization Group. His group created several popular tools, including D3.js (Data-Driven Documents) and Data Wrangler. Jeff is currently a faculty member of Computer Science & Engineering at the University of Washington.

Who this course is for:

Data visualization Analyst
A data journalist