Difference between revisions of "Visualization"

From TedYunWiki
Jump to navigation Jump to search
 
(10 intermediate revisions by the same user not shown)
Line 11: Line 11:
 
# Visual Analytics
 
# Visual Analytics
  
−
'''''The human visual system is the highest bandwidth channel to the human brain!'''''
+
'''''The human visual system is the highest bandwidth channel to the human brain.'''''
 +
 
 +
==== Graphs Reveal Data that Statistics May Not ====
 +
 
 +
e.g. Anscombe's Quartet (four data sets with the identical linear model but looks very different in visualization)
 +
 
 +
=== Data Visualization Process ===
 +
# Classify data types
 +
# Determine which visual attributes represent data types most effectively
 +
 
 +
==== Data Types ====
 +
* Nominal (labels, names) e.g. fruits: apples, oranges.
 +
*; Operations: =, !=
 +
* Ordinal (ordered): e.g. quality of meat: Grade A, AA.
 +
*; Operations: =, !=, <, >, <=, >=
 +
* Quantitative
 +
** Interval (location of zero arbitrary): e.g. date, location
 +
*** Like a geometric point, cannot compare directly.
 +
*** Only differences (i.e. intervals) may be compared.
 +
**; Operations: =, !=, <, >, <=, >=, - (can measure distances or spans)
 +
** Ratio (zero fixed): physical measurement e.g. length, mass.
 +
*** Counts and amounts.
 +
*** Like a geometric vector, origin is meaningful.
 +
**; Operations: =, !=, <, >, <=, >=, -, / (can measure ratios or proportions)
 +
 
 +
=== Visual Attributes ===
 +
==== Bertin's 7 Visual Attributes ====
 +
# Position
 +
# Size
 +
# Value (lightness)
 +
# Texture
 +
# Color
 +
# Orientation
 +
# Shape
 +
# (3D)
 +
# (Time)
 +
(Card and Mackinlay's extension)
 +
 
 +
==== Perceptual Properties ====
 +
In the order of accurate perception,
 +
; Position > Length > Angle = Shape > Area > Volume > Color = Density
 +
(Mackinlay)
 +
 
 +
=== Effective Visual Encoding ===
 +
* Importance Ordering: Encode the most important information in the most perceptually accurate way.
 +
* Expressiveness: Depict all the data, and only the data.
 +
* Consistency: The properties of the image (visual attributes) should match the properties of the data. e.g. Do not map one dimensional data to two or three dimensional representations.
 +
(Mackinlay)
 +
 
 +
=== Spatial Position ===
 +
How to increase the amount of information encoded by spatial position?
 +
# Composition
 +
# Alignment
 +
# Folding
 +
# Recursion
 +
# Overloading

Latest revision as of 22:07, 13 November 2013

Goals

  • Effective communication of information
  • Clarity
  • Integrity
  • Stimulate viewer engagement

Data Visualization

Three categories:

  1. Information Visualization
  2. Scientific Visualization
  3. Visual Analytics

The human visual system is the highest bandwidth channel to the human brain.

Graphs Reveal Data that Statistics May Not

e.g. Anscombe's Quartet (four data sets with the identical linear model but looks very different in visualization)

Data Visualization Process

  1. Classify data types
  2. Determine which visual attributes represent data types most effectively

Data Types

  • Nominal (labels, names) e.g. fruits: apples, oranges.
    Operations
    =, !=
  • Ordinal (ordered): e.g. quality of meat: Grade A, AA.
    Operations
    =, !=, <, >, <=, >=
  • Quantitative
    • Interval (location of zero arbitrary): e.g. date, location
      • Like a geometric point, cannot compare directly.
      • Only differences (i.e. intervals) may be compared.
      Operations
      =, !=, <, >, <=, >=, - (can measure distances or spans)
    • Ratio (zero fixed): physical measurement e.g. length, mass.
      • Counts and amounts.
      • Like a geometric vector, origin is meaningful.
      Operations
      =, !=, <, >, <=, >=, -, / (can measure ratios or proportions)

Visual Attributes

Bertin's 7 Visual Attributes

  1. Position
  2. Size
  3. Value (lightness)
  4. Texture
  5. Color
  6. Orientation
  7. Shape
  8. (3D)
  9. (Time)

(Card and Mackinlay's extension)

Perceptual Properties

In the order of accurate perception,

Position > Length > Angle = Shape > Area > Volume > Color = Density

(Mackinlay)

Effective Visual Encoding

  • Importance Ordering: Encode the most important information in the most perceptually accurate way.
  • Expressiveness: Depict all the data, and only the data.
  • Consistency: The properties of the image (visual attributes) should match the properties of the data. e.g. Do not map one dimensional data to two or three dimensional representations.

(Mackinlay)

Spatial Position

How to increase the amount of information encoded by spatial position?

  1. Composition
  2. Alignment
  3. Folding
  4. Recursion
  5. Overloading
Retrieved from "https://wiki.tedyun.com/index.php?title=Visualization&oldid=148"