Effective Visualization of Multi-Dimensional Data — A Hands-on Approach

Strategies for Effective Data Visualization

Introduction

Motivation

The Datasaurus Dozen — What is common among these diverse datasets?
Anscombe’s quartet — Different datasets with same summary statistics
Visualizing the Broad Street Cholera outbreak which helped find the root cause of the disease outbreak! (Source: https://github.com/dipanjanS/art_of_data_visualization)
Causes of Mortality in the Army of the East — Florence Nightingale (Source: https://github.com/dipanjanS/art_of_data_visualization)

A Quick Refresher on Visualization

Understanding the Grammar of Graphics

Understanding Grammar of Graphics (Source: https://github.com/dipanjanS/art_of_data_visualization)
Major components of the Grammar of Graphics ( Source: https://github.com/dipanjanS/art_of_data_visualization)

Visualizing Structured Multi-Dimensional Data

The wine quality dataset
Basic descriptive statistics by wine type

Univariate Analysis

Visualizing data in One Dimension (1-D)

Visualizing attributes as one-dimensional data
Visualizing one-dimensional continuous, numeric data
Visualizing one-dimensional discrete, categorical data

Multivariate Analysis

Visualizing data in Two Dimensions (2-D)

Visualizing two-dimensional data with a correlation heatmap
Visualizing two-dimensional data with pair-wise scatter plots
Parallel coordinates to visualize multi-dimensional data
Visualizing two-dimensional continuous, numeric data using scatter plots and joint plots
Visualizing two-dimensional discrete, categorical data using bar plots and subplots (facets)
Visualizing two-dimensional discrete, categorical data in a single bar chart
Visualizing mixed attributes in two-dimensions leveraging facets and histograms\density plots
Leveraging multiple histograms for mixed attributes in two-dimensions
Box Plots as an effective representation of two-dimensional mixed attributes
Violin Plots as an effective representation of two-dimensional mixed attributes

Visualizing data in Three Dimensions (3-D)

Visualizing three-dimensional data with scatter plots and hue (color)
Visualizing three-dimensional numeric data by introducing the notion of depth
Using the notion of size or hue for representing continuous data in 3-D
Using the notion of facets for representing continous data in 3-D
Visualizing three-dimensional categorical data by introducing the notion of hue and facets
Visualizing mixed attributes in three-dimensions leveraging scatter plots and the concept of hue
Visualizing mixed attributes in three-dimensions leveraging kernel density plots and the concept of hue
Visualizing mixed attributes in three-dimensions leveraging split violin plots and the concept of hue
Visualizing mixed attributes in three-dimensions leveraging box plots and the concept of hue

Visualizing data in Four Dimensions (4-D)

Visualizing data in four-dimensions leveraging scatter plots and the concept of hue and depth
Visualizing data in four-dimensions leveraging bubble charts and the concept of hue and size
Visualizing data in four-dimensions leveraging scatter plots and the concept of hue and facets
Visualizing data in four-dimensions leveraging scatter plots and the concept of hue and facets
Visualizing data in four-dimensions leveraging scatter plots and the concept of hue and facets

Visualizing data in Five Dimensions (5-D)

Visualizing data in five-dimensions leveraging bubble charts and the concept of hue, depth and size
Visualizing data in five-dimensions leveraging bubble charts and the concept of hue, facets and size
Visualizing data in five-dimensions leveraging scatter plots and the concept of hue and facets

Visualizing data in Six Dimensions (6-D)

Visualizing data in six-dimensions leveraging scatter charts and the concept of hue, depth, shape and size
Visualizing data in six-dimensions leveraging scatter charts and the concept of hue, facets and size

Can we go higher?

Hans Rosling’s famous visualization of global population, health and economic indicators

Visualizing Unstructured Data — Text, Image, Audio

Visualizing Text Data

Basic EDA on Text Data
Constituency & Dependency Parsing — Understanding Text Structure
An example of visualizing word2vec word embeddings on The Bible

Visualizing Image Data

Visualizing image channels
Visualizing image intensity distributions
Visualizing image edges
Visualizing HOG features

CNNs changed the world

CNNs revolutionized image representation and feature engineering

Visualizing Audio Data

Visualizing Audio Waveform Amplitudes
Visualizing Audio Mel-spectrograms

Conclusion

This story is published in The Startup, Medium’s largest entrepreneurship publication followed by +398,714 people.

Subscribe to receive our top stories here.

The Startup

Get smarter at building your thing. Join The Startup’s +741K followers.