PowerPoint Presentation
1
Data Visualization Framework
2
Data
Source
Data Layer
Mapper
Data
Collection
Data
Source
Data
Source
Mapping Layer Graphics Layer
Data Layer
• Locating and
obtaining data
• Importing data in
proper format
• Relating data for
proper
correspondence
• Data analysis and
aggregation
Mapping Layer
• Associating
appropriate
geometry with
corresponding
data channels
• Data analysis and
algorithms
(e.g. contouring)
Graphics Layer
• Conversion of
geometry into
displayable image
• Decorations
• Managing interaction
3
Data Types
4
Ordered
(values are
comparable)
Unordered
(values not
comparable)
Discrete
(no between values)
Continuous
(values between)
Cyclic values,
e.g. directions, hues
Ordinal,
e.g. size: S,M,L,XL,…
Quantitative,
e.g. counts: 1,2,3,…
Fields,
e.g. altitude,
temperature
Nominal,
e.g. shape:
Categories,
e.g. nationality
Mapping Quantitative Values
• Position
• Length
• Angle/Slope
• Area
• Volume
• Color/Density
2
CLEVELAND, W. S., AND MCGILL, R. Graphical perception: Theory, experimentation and application to
the development of graphical methods. Journal of the American Statistical Association, 79(387) 1984
Mapping Quantitative Values
• Position
• Length
• Angle/Slope
• Area
• Volume
• Color/Density
3
CLEVELAND, W. S., AND MCGILL, R. Graphical perception: Theory, experimentation and application to
the development of graphical methods. Journal of the American Statistical Association, 79(387) 1984
Mapping Quantitative Values
• Position
• Length
• Angle/Slope
• Area
• Volume
• Color/Density
4
Perceptual
Accuracy
CLEVELAND, W. S., AND MCGILL, R. Graphical perception: Theory, experimentation and application to
the development of graphical methods. Journal of the American Statistical Association, 79(387) 1984
Mapping Quantitative Values
• Position
• Length
• Angle/Slope
• Area
• Volume
• Color/Density
5
Perceptual
Accuracy
1-D
2-D
3-D
CLEVELAND, W. S., AND MCGILL, R. Graphical perception: Theory, experimentation and application to
the development of graphical methods. Journal of the American Statistical Association, 79(387) 1984
6
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
7
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
8
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
9
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
10
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
11
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
12
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
13
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Length
Angle
Slope
Area
Volume
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
14
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
15
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
16
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
17
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
18
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
19
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Hue
Texture
Connection
Containment
Density
Saturation
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
20
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Hue
Texture
Connection
Containment
Density
Saturation
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
21
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Hue
Texture
Connection
Containment
Density
Saturation
Length
Angle
Slope
Area
Volume
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
22
Position
Length
Angle
Slope
Area
Volume
Density
Saturation
Hue
Quantitative
Position
Density
Saturation
Hue
Texture
Connection
Containment
Length
Angle
Slope
Area
Volume
Ordinal
Position
Hue
Texture
Connection
Containment
Density
Saturation
Shape
Length
Angle
Slope
Area
Volume
Nominal
J. Mackinlay, Automating the Design of Graphical Presentations of
Relational Information, ACM Transactions on Graphics 5(2), 1986
Bar Chart
2
Quantitative
dependent
variable
Discrete/nominal
independent variable
Benefits from both
position (top of bar)
and length (size of bar)
Line Chart
3
Quantitative
continuous
dependent
variable
Quantitative continuous
independent variable
Benefits from
position but
not length
Scatter Plot
4
Quantitative
independent
variable
Quantitative
independent variable
Relies mostly
on position,
but clusters
also yield
density
Table
6
Discrete/nominal
independent
variable
Discrete/nominal
independent variable
Benefits from
position only
(notice the lateral
inhibition flashing?)
What to Use?
7
Nominal or
Q. Discrete
Nominal or
Q. Discrete
Quantitative
Discrete
Table
Bar
Quantitative
Continuous
Quantitative
Continuous
Line
Quantitative
Continuous
Scatter Gantt
Gantt
Bar
Bar
Ind.
Dep.
Independent