Data Visualization Foundation
The Utah Teapot.
About
Data visualization is the process of mapping quantitative data to visuals (shapes, color, position, etc) to create a graph made of geometric object.
Information visualization:
- is defined as “visual representations of abstract data to amplify cognition”.
- is not useful when the information is specific (for a single instance)
The greatest value of a picture is when it forces us to notice what we never expected to see.
A picture is worth a thousand words
The Purpose of computing is insight, not numbers.
See What are images? (also known as pictures)
Giving shapes to data !
Document structure
A scene graph is generally the document containing all graphic information.
Data Type
Characters
Numeric
Graphical methods class:
- diagram techniques,
- chart techniques,
- plot techniques,
Visual Encoding + Data Type + Relationship = Chart Types
| Type | Description |
|---|---|
| Data Visualization - Map | Geographical data |
| Data Visualisation - Voronoi | Map |
| Data Visualisation - Heat Map | For a lot of data |
| Data Visualisation - Scatterplot | For a lot of data |
| Tree - Treemap Chart | hierarchy data |
| Data Visualization - Stream graph | |
| Data Visualisation - Histogram (Frequency distribution) | distribution |
| Data Visualization - Box Plot | summary of distribution |
| Data Visualization - Bar Chart | ranking, comparison |
| Data Visualization - Line Chart | deviation, trend |
| Data Visualisation - Area Chart | ratio |
| Table | Table |
| Ring Charts | hierarchy data |
| Flow Charts | Data or Business Processing |
Choosing
Software
Real-time
In realtime chart, you try to repaint only a part of the chart and not to repaint it completely.
Generally updating completely once per second is fine, but updating multiple times per second results in high CPU load.
Best practices / Fact
- Use common scales to be able to compare across the graphs
- Proportions are difficult to interpret
- Avoid pie charts – Angular and curvature comparisons are hard to interpret.
- Let it simple. Do not use 3-D charts, shading. Limit border, …
Quality
Foundation Vis Papers and Books
- wiki/Jacques Bertin started ranking visual features (Semiology of Graphics (1983) )
- William S. Cleveland validated this ranking with a scientific method. Graphical perception: Theory, Experimentation, and Application to the Development of Graphical Methods . He wrote two books: The Elements of Graphing Data and Visualizing Data, based on the research carried out in this paper.
- Automating the Design of Graphical Presentations of Relational Information. Jock Mackinlay. one of the minds behind Tableau Software. Automatically design effective visualizations automatically by matching data features with visual features through the use of logic rules. Knowing how to match data features to visual features is one of the most important skills of knowledgeable data visualization experts.
- How NOT to Lie with Visualization. Bernice E. Rogowitz, Lloyd A. Treinish This work is fully focussed on color use and perception. How to build effective color scales that take into account data distribution.
- The Eyes Have It: A Task by Data Type Taxonomy for Information Visualizations Ben Shneiderman A classification of information visualization techniques according to data type and more importantly the explanation of the famous infovis mantra. : “overview first, zoom and filter, details on demand”.
