Getting started with data visualization

What is a data visualization?

Why do we need data visualizations?

A good data visualization provides insights that are not immediately obvious from the data alone.

A good data visualization asks and answers a question or tells a story.

A good data visualization provides appropriate context for an audience.

Tools of the trade

Charting

Chartbuilder

https://quartz.github.io/Chartbuilder/

Fusion Tables

https://fusiontables.google.com

Tableau

https://tableau.com

D3

https://d3js.org/

Mapping

Google Maps

https://google.com/mymaps

QGIS

http://www.qgis.org/en/site/

Map Box

https://www.mapbox.com/

Leaflet

http://leafletjs.com/

How to make a good data visualization

Kara Dapena

Visual story-telling tool

A good data visualization is a story-telling tool to deliver accurate information in a format that requires the least effort possible from the audience to gain insight

Creation process

1. Be curious
2. Define the punchline
3. Sketch
4. Test and revamp
5. Publish

Principles of design

• Variety
• Unity
• Heirarchy

Variety

Unity

Hierarchy

Principles of design

Data-ink ratio

Be mindful of what to show and not to show

Parts that make the whole

Descriptive headline

Introduction

Graphic component

Byline

Note (if needed)

Source

Create a style guide

• Streamline your workflow
• Give your graphics a consistent look and feel, to create a sense of familiarity for your audience
• Pick one or two typefaces and establish typography for common elements
• Make a baseline color pallet

FiveThirtyEight

http://projects.fivethirtyeight.com/2016-election-forecast/

Decide on your format

• What platform are you using and what constraints or possibilities does it have?
• What structure is best suited to tell the story? Should it be a database, a map, a chart, a table, a timeline, etc...

Comparison precision scale

Most to least precise:

1. Position along a common scale
2. Positions along nonaligned scales
3. Length, direction, angle
4. Area
5. Volume, curvature
6. Shading, color saturation

Do's and dont's

3D charts

Do not distort your charts by making them look 3D

Start at zero

In a bar chart, the length of the bar indicates the value, so (almost) always start at 0

Apples to apples

Be sure that in addition to being accurate, your data is painting a fair picture

Don't fight your audience

Use the cultural conventions of your audience to make your visualization easier to read

Annotate

Annotate to give your audience clues and insights

It's also OK to break the "rules"