’s Visual Variables
• Location variables (position, relative to a coordinate frame)
– e.g. horizonal and vertical axes on a scatterplot; longitude and latitude on a map
– (so fundamental to presenting map information that these variables are often ignored in cartography)
Copyright By PowCoder代写 加微信 powcoder
• Retinal variables (perceptual properties)
– ways of representing differences between objects
– size, shape, colour (hue), colour (value), texture, orientation
This separation makes clear the difference between the spatial relationships between symbols and the perceptual properties of the symbols themselves
• Location variables
– fix a ‘graphic mark’ (symbol, visual object) on to a position on the plane
• Retinal variables
– ‘elevate’ that mark with a different ‘pattern of light’
, Mapping It Out (1993)
Bertin, Sémiologie Graphique (1967)
Deux Dimensions du Plan Taille
Couleur Orientation Forme
Two Dimensions on the Plane Size
Colour (value)
Colour (hue) Orientation Shape
Representing multi-dimensional data with Glyphs
• A glyph is a visual object representing a single multi-dimensional data point
• The values of the different (retinal) visual variables in the glyph represent the values of the different dimensions
• The values of the two (location) variables add spatial dimensions to the visualization – by placing glyphs at different positions
Four dimensions:
retinal visual variable (orientation) The “body” is always vertical
There are four “limbs” that can be drawn at different angles
Any of these 12 glyphs can be chosen to represent data of four dimensions
The angle of each limb represents a dimension value
(If the body is angled too, then five dimensions can be represented)
, Grinstein G. Iconographic displays for visualizing multidimensional data. Proc. IEEE Conference on Systems, Man and Cybernetics 1988; 164 – 170.
Four hill runners
The angle of each line – at the point of connection – shows the distance they ran on each of four days [1..36km] (working from left clockwise)
Colour indicates age group
Adding the Location visual variable Using glyph number 12, with
body orientation
“Five channels of data from [a weather] satellite”
The western end of Lake Ontario and part of the eastern tip of Lake Erie
, Grinstein G. Iconographic displays for visualizing multidimensional data. Proc. IEEE Conference on Systems, Man and Cybernetics 1988; 164 – 170.
Ward, M.O. A taxonomy of glyph placement strategies for multidimensional data visualization. Information Visualization (2002) 1, 194-210
Circles are drawn at 50%
1932: shape of face 1936: length of nose 1940: curvature of mouth 1960: width of mouth 1964: slant of eyes
1968: length of eyebrows
Percentage of Republican votes in six Presidential Elections (1932-1968) in six Southern states
Kleiner B, Hartigan J. Representing points in many dimension by trees and castles. Journal of the American Statistical Association 1981; 76: 260 – 269
Chernoff faces
Six measurements on 87 fossils:
– Z1 inner diameter of embryonic chamber (in microns)
– Z2 total number of whorls
– Z3 number of chambers in first whorl
– Z4 number of chambers in last whorl
– Z5 maximum height of chambers in first whorl (in microns) – Z6 maximum height of chambers in last whorl (in microns)
Chernoff H. The use of faces to represent points in k-dimensional space graphically. Journal of the American Statistical Association 1973; 68: 361 – 368.
“People grow up studying and reacting to faces all of the time.
Small and barely measurable differences are easily detected and evoke emotional reactions from a long catalogue buried in the memory.
Relatively large differences go unnoticed in circumstances where they are not important.
This implies that the human mind sub-consciously operates as a high-speed computer, filtering out insignificant visual phenomena and focusing on the potentially important.
Particularly valuable is this flexibility in disregarding non-informative data and searching for useful
information.”
Chernoff H. The use of faces to represent points in k-dimensional space graphically. Journal of the American Statistical Association 1973; 68: 361 – 368.
Visual clustering shows seven different groups of fossils:
• I:(1,2,3,9,22,29)
• II:(4,5,6,7,8)
• III: (10, 11, 14, 23, 25, 26, 27)
• IVa: (13, 15, 16, 17, 18, 19, 20)
• IVb: (12, 24)
• V: (21, 28, 30, 31, 37, 38, 39, 40, 41)
• VI: (32, 33, 35, 36)
Chernoff H. The use of faces to represent points in k-dimensional space graphically. Journal of the American Statistical Association 1973; 68: 361 – 368.
Lung cancer incidence (males 1968-1972)
Rahu M. Graphical representation of cancer incidence data: Chernoff faces. International Journal of Epidemiology, 1989; 18: 763-767.
Placement of glyphs
• A single glyph only represents one data point
• A data set is therefore represented by several glyphs
• They need to be placed on the plane, where their position in relation to each other:
– does not matter
– relates to an ‘information’ dimension – relates to a ‘spatial’ dimension
Information location:
using two (represented) dimensions – star glyphs
Iris data set:
petal length, petal width, sepal length, sepal width
sepal length
sepal width
(five other sctterplot pairwise configurations possible)
Ward, M.O. A taxonomy of glyph placement strategies for multidimensional data visualization. Information Visualization (2002) 1, 194-210
Information location:
using two (unrepresented) dimensions – profile glyphs
2011 2012 2013
MSc awards
Spatial location:
using geographic co-ordinates – arrow glyphs
https://confluence.ecmwf.int/display/UDOC/How+to+clip+wind+arrows+to+the+plotting+frame+ -+Metview+FAQ
Members of a running club
Best race finishing position
Members of a running club
children, women, men
(nominal, shape)
Scotland, England, Wales, Ireland
(nominal, hue)
number of runners: 10, 50, 100
(ordered & quantitative, area)
best race finishing position
(ordered & quantitative, horizontal position)
50 40 30 20 10
Members of a running club
children, women, men
(nominal, shape)
Scotland, England, Wales, Ireland
(nominal, hue)
number of runners: 10, 50, 100
(ordered & quantitative, area)
best race finishing position
(ordered & quantitative, horizontal position)
Interaction between shape and size?
50 40 30 20 10
Swant, A.P. (2022) A Visualization Approach towards Analyzing and Correlating FORTUNE 100 Stocks (unpublished, from ResearchGate)
temperature → hue
wind speed → density pressure → size
precipitation → orientation cloud coverage → luminance
Sawant, A. P. and Healey, C. G. (2008). “Visualizing Multi- dimensional Query Results Using Animation,” Proceedings Visualization and Data Analysis (VDA 2008), San Jose, California, Vol. 6809, paper 04, pp. 1-12
Ranks (location)
High rank in the middle Lower rank near the periphery
Predicted user rating (size)
Large: high ratings Small: low ratings
Year (colour hue)
Old: blue Recent: red
Length (colour value)
Dark: short Bright: long
Genre (shape – corner triangles)
Northeast: comedy Northwest: action Southwest: romance Southeast: drama
Movie recommendations
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com