This document outlines an agenda for a data visualization workshop. It discusses why visualizing data is important for exploring patterns, communicating results, and telling stories. Examples are given of historical visualizations that helped identify cholera outbreaks and military campaigns. The main steps for visualizing data are introduced: being clear on objectives, preparing the data, building visualizations using appropriate tools, and ensuring success. Global Burden of Disease visualizations are presented as examples for research settings. The document concludes with encouraging questions and further resources.
Python Notes for mca i year students osmania university.docx
Data visualization workshop
1. UNIVERSITY OF WASHINGTON
Data visualization workshop
Peter Speyer Kyle Foreman
Director of Data Development PhD Candidate
IHME Imperial CollegeJune 18, 2013
2. Agenda
• Introduction
• Interactive visualizations
• GBD visualizations: examples in a research setting
• The main steps for visualizing data
• Practical example
• Final questions
2
3. Why do we visualize data?
Review data
• Make sense of large amounts
of data
• Explore patterns and trends
• Evaluate research results
• Find stories
Communicate results
• Make data engaging
• Cut through the clutter
• Let users explore the data
• Use for presentations
• Tell stories
3
9. “People are generally better persuaded
by the reasons
which they have themselves discovered
than by those
which have come into the mind of others”
Blaise Pascal
9
10. Agenda
• Introduction
• Interactive visualizations
• GBD visualizations: examples in a research setting
• The main steps for visualizing data
• Practical example
• Final questions
10
11. Agenda
• Introduction
• Interactive visualizations
• GBD visualizations: examples in a research setting
• The main steps for visualizing data
• Practical example
• Final questions
11
12. Global Burden of Disease 2010 - Results
291 causes/4 hierarchical levels
67 risk factors/2 levels
21 age groups (3 infant age groups, 1-4, 5-9… 75-79, 80+)
Female/male/both
187 countries
1990, 2005, 2010
4 key metrics (deaths, YLLs, YLDs, DALYs)
Uncertainty bounds
12
13. Use of visualizations for research
13
Improving the
research work flow:
Mortality Visualization
COD Visualization
Review results:
GBD Compare
Share results & tell
stories:
GBD Cause Patterns
GBD Arrow Diagrams
Evaluating policy
impact:
Benchmarking tool
14. Agenda
• Introduction
• Interactive visualizations
• GBD visualizations: examples in a research setting
• The main steps for visualizing data
• Practical example
• Final questions
14
15. Be clear about your objectives
• What do I want to do/
communicate?
• Am I telling a story or letting users
explore?
• What is my audience? How much
do they know about the topic?
About statistics? About
visualizations?
15HikingArtist via Flickr
16. Prepare the data
• Identify all relevant available data
• Become intimate with your
dataset(s): metrics, units,
dimensions, uncertainty
• Prepare data: Excel, Google Refine,
Data Wrangler, AP’s Overview
16Kikishua via Flickr
17. Build it
• Select the right type of visual
– Highlight your point
– Keep it simple
• Select the degree of interactivity
• Select the right visualization tool:
start simple
– Excel
– Public tools: Google Motion Charts,
Tableau Public, ArcGIS.com
– Custom coding: D3.js, Highcharts
– Maps: visualization vs. GIS
17Edwc via Flickr
18. Final thoughts
• Facilitate viral communication
– Permalinks
– Social media integration
– Embedding visualizations
– Download screenshot
• Working with software developers
– Requirements
– Testing
– Documentation
– Priorities
18ocean.flynn via Flickr
19. How do I know if I succeeded?
19Mr. Aktugan via Flickr
21. Agenda
• Introduction
• Interactive visualizations
• GBD visualizations: examples in a research setting
• The main steps for visualizing data
• Practical example
• Final questions
21
IntroExamplesTalk through GBD viz for usageLearnings from creating those: how to go aboutPractical examplesQ&A
Let’s start with the last points and work backwardsFew historic examples to provide some contextClear take-aways
Cholera on braod streetTheory: bad airCounting casesNo proof in waterPump handle
Pie chart with 12 slices for monthsArea (from middle) proportionate to deathsCommunicable disease bigger enemy than RussiansMilitary hospital system
Number of soldiersLocationDirectionDateTemperatureStories: river crowssing
Fast-forward over 100 yearsWeb allows interactivitySoftware allows watching movies of time trends, allows for integrating any indicatorTED visualizations
Bill Gates: big milestone after Rosling’sGapminder