Named graphs and maps: scatterplot, scatterplot matrix (grid of scatter plots used to visualize bivariate relationships between combinations of variables), stripchart (1d scatterplot), bubble chart, bar chart, lollipop chart, coxcomb chart (stacked bar chart with radial layout), Marimekko chart (bar chart where the width encodes relative size, also called Mekko chart), waterfall chart, pie chart (stacked bar charts in polar coordinates, angle encodes proportion), donut chart, line chart, sparkline, slopegraph, parallel coordinates, radar chart (graphical method of displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point, also called web, spider, star, cobweb, polar, or Kiviat chart), area chart, streamgraph (tacked area graph displaced around a central axis), dendrogram, Reingold–Tilford tree (hierarchical data as linked tree layout implemented using the Reingold-Tilford algorithm), treemap (hierarchical data as nested rectangles, area proportional to value), sunburst (hierarchical data as rings. center is root node. angles are equal or proportional to value), alluvial diagram (relations between multivariate data), sankey diagram (magnitude of flow between
nodes in a network), network graph (relationships as lines between entities as nodes), heat map (matrix values as colors), chord diagram (shows directed relationships among a group of entities in a matrix), word cloud, bubble cloud, circle packing (bubble cloud technique with hierarchical information as enclosing circles), time series plot (values ordered in time as a line chart), index chart (interactive line chart that shows percentage changes for a collection of time-series based on a selected index point), gantt chart (schedule with tasks layed out on time axis), timeline (events layed out on time axis), thematic maps (choropleth, proportional symbol map, dot map (can be used to locate each occurrence of a phenomenon, one-to-one or one-to-many), cartogram, isopleth), choropleth (areas are shaded or patterned in proportion to variable), proportional symbol map (scaled symbols show data for areas/locations, also called graduated symbol map), dot map, cartogram (area used to display value, distortion used to show continuous variables), isopleth or isarithmic (use contours to show continuous variables), topographic map (detailed quantitative representation of land relief using contour lines), nautical map (charts of maritime/coastal area), image based map (using satellite or aerial imagery), combo or combination chart (combination of multiple charts, can have multiple series and y-axes).
Statistical graphics: boxplot (shows quartiles, distribution skewness, tails, outliers in unimodal distributions) also named box-and-whisker plot), violin plot (mirrored probability density), dot plot (statistical chart consisting of data points plotted on a fairly simple scale, typically using filled in circles), frequency distribution table (displays the frequency of various outcomes in a sample, also called frequency table), population pyramid or age-sex pyramid (graphical illustration of the distribution of a population by age groups and sex), stem-and-leaf plot (device for presenting quantitative data in a graphical format, similar to a histogram, to assist in visualizing the shape of a distribution), histogram (approximate representation of the distribution of numerical data), frequency polygon (histogram as a line chart), Q–Q or quantile-quantile plot (graphical method for comparing two probability distributions by plotting their quantiles against each other), Heatmap of Pearson’s correlation coefficients (graphical tool to assess correlations in multivariate data), scree plot (line plot of the eigenvalues of factors or principal components in an analysis), biplot (generalization of the simple two-variable scatterplot with variables displayed either as vectors, linear axes or nonlinear trajectories)
Graphical elements in charts: title, legend, axes, axes labels, labels grid lines, tick marks, tick labels
Infographic: graphic visual representations of information, data or knowledge intended to present information quickly and clearly, Receiver Operator or ROC Curve(diagnostic tool for binary classifiers with decision threshold),
Dashboard: graphical user interface which often provides at-a-glance views of key performance indicators (KPIs) relevant to a particular objective or business process.
Pop-out effect: universal property, independent of practice, familiarity with the features and number of distractors
The greater the distance between target and distractors the greater the pop-out effect
Visual pathways (two stream hypothesis): Works without visual input,
Where / Dorsal visual pathway: relative object location for motor tasks
What / Ventral visual pathway: object identification and recognition
Pattern recognition: process that matches information from a stimulus with information retrieved from memory
Priming: Effect in which exposure to one stimulus influences a response to a later stimulus.
Apophenia: Perception of images or sounds in random stimuli, priming increases likelihood of seeing the pattern
Convergence: group of cells form a receptive field for a cell in the brain
Neuronal tuning: Simpler tuning in earlier visual areas (V1 & V2), Complex tuning in higher visual areas (V4 & IT)
Lower visual cortex: low information, high localization, universal experience, pop-out, developed early, e.g., V1 neurons may fire to any vertical stimulus.
Higher visual cortex: high information, low localization, individual experience, no pop-out effect, e.g., IT neurons may fire only to a specific face.
Some neurons of V1 are tuned to vertical lines, others to diagonal lines
Memory types: Iconic, VSTM, VLTM, from visual persistance to information persistance
Iconic memory: Unlimited capacity, Retention: ≤ 1s, high bandwidth, works unconsciously, provides temporal integration
ensures continuity during saccades
VSTM memory: Limited capacity, Retention: ≤ 30s, Buffer that stores temporary information, Constructs and manipulate visual images
VLTM memory: Large capacity, Retention: indefinite, Capacity increases over childhood, declines with old age, Encodes information semantically for long term storage, Subject to fading, recalls help preserve it
Color and perception
Chromatic aberration effect: red/blue perceived at different distance
Types of color scales: sequential, divergent, qualitative
Primary and secondary colors
After-images and predicting colors using additive RGB color model
Color properties distinguishable: Hue, Saturation, Brightness
Color vision theories working together: trichromatic (R, G, B cones) and opponent process (visual system responds to opponent channels R-G, B-Y, B-W)
Trichromatic theory problems: no R-G or Y-B, R-G overlap, small B response, afterimages)
Subtractive color model: print, Primaries: CMY, secondary: RGB
Additive color model: computer screens, Primaries: CMY, secondary: RGB
Additive color displays: addition of illumination (projectors), partitive mixing (LCD), time mixing (OLED), binocular mixing (stereovision)
Color model: abstract mathematical model describing the way colors can be represented as tuples
RGB, HSV color models
Color blindness: about 9% of the population, R-G
Color and perception
False colors techniques: choropleth, density slicing (divides the image into few colors)
Pseudocolor, e.g., colored IR image
Simultaneous color contrast: colors of different objects affect each other
Color constancy: ensures color remain relatively constant under varying illumination
Mach bands: illusion due to simultaneous contrast - can appear in color scales
Sharpening: More sensitive to dark than light differences, affected by background
Composite projections: same projection optimized for different areas, e.g., Albers USA
Mercator projection: standard for Web mapping applications
Universal Transverse Mercator (UTM) projection is a projection over over $61 \times 6^o$ zones in cartesian coordinates, no trigonometry is needed to compute distances, units in meters as Easting and Northing
Maps in the Browser can be raster (images) or vector (SVG)
Tile Map Service (TMS): e.g., Google maps serve map tiles also called slippy maps
GeoJSON: JSON map format defining Geometry (Point, LineString, Polygon, MultiPoint, MultiLineString, MultiPolygon, GeometryCollection), Feature and FeatureCollection
TopoJSON: GeoJSON extension where to encode topology (Geometry is indexed with arcs) leading to smaller files
GeoJSON and TopoJSON coordinates can be in geographical or projected coordinates
Cairo's suggestion to deal with novel forms (use redundancy)
Minimalistic visualizations
Tufte principles: 1. Above all else show data, 2. Maximize the data-ink ratio, 3. Erase non-data-ink, 4. Erase redundant data-ink, 5. Revise and edit
Nigel Holmes design principle: Use humor to instill affection in readers for numbers and charts
Data-ink ratio (Data-ink / Total ink used to print the graphic, Proportion of a graphic’s ink devoted to the non-redundant display of data-information, 1.0−Proportion of a graphic that can be erased without loss of data-information)
Chartjunk
Visual query: a pattern cognitively specified, that if found in the display will contribute to the solution of a problem
Ware design principles: Carefully craft visualizations to optimize visual queries, leverage information that the brain processes efficiently, (e.g., pre-attentive features) to prioritize most important information
Avoid attentional overload and change blindness
Uses of colors (Tufte): Label (identify, highlight, group), measure, represent or imitate reality, enliven or decorate
Color design guidelines:
blue text hard to read (fewer blue cones)
achromatic (BW) channel is easier to read (use all 3 cones)
colors harder to read (achromatic ~ 3 x chromatic channel info)
W on B less strain than B on W
Be considerate of colorblind people
Respect well-established color sequences
Shape from shading is recognized from the luminance channel
Observe cultural conventions
Use consistent aesthetics
Most important visual queries should have most weight
Apprehandable chunk: unlearned pattern complexity that can be apprehended in one fixation
Apprehendable chunks consist of about three components
Tableau: built on VizQL an implementation of Grammar of graphics where mappings are specified interactively and visually
Dimensions (categorical) and Measures (numerical) variables in Tableau
SVG vs. canvas elements
canvas API: used with 2D context (getContext('2d'), shape, text image rendering functions
canvas with WebGL: used with 3D context (getContext('webgl')) primitives: GL_POINTS, GL_LINES, GL_LINE_STRIP, GL_LINE_LOOP, GL_TRIANGLES, GL_TRIANGLE_STRIP, GL_TRIANGLE_FAN
In WebGL and related toolkits data is passed as indexed arrays
Graphic pipeline: process of rendering a 3D scene in WebGL, includes vertex shader, triangle assembly, rasterization, fragment shader, testing and blending, steps
Model and view matrices: used in WebGL to place the object to render and the camera in a world coordinate system
Projection matrix: used to define how 3D is projected in the camera, can be defined using the clipping planes
Three.js: High-level access to WebGL and graphical utilities, e.g., scene, camera, model loaders, lights and materials
Processing: simplified Java API for drawing and graphics
Processing.js: JS API to use Processing code
P5.js: HTML5 processing implementation
Deck.gl and Mapbox GL JS 3d rendering capabilities
Notes
What type of chart is this?
Stripline
Slopegraph
Line chart
Area chart$\leftarrow$
What choice corresponds to the CSS rule:
div, p a.foo {
color: red;
font-weight: bold;
}
div elements, hyperlinks with class foo and paragraphs
div elements, hyperlinks with class foo in paragraphs $\leftarrow$
div elements, paragraphs in hyperlinks with class foo
div elements in paragraphs or in hyperlinks with class foo
<!-- assume data.csv contains:
Color
Red
Green
Blue
-->
<p>Orange</p>
<script>
d3.csv("data.csv").then(function(data) {
d3.select("body")
.selectAll("p")
.data(data)
.enter()
.append("p")
.text(function(d) {return d.Color; })
});
</script>
Orange, Color, Red, Green, Blue on separate lines
Orange, Green, Blue on separate lines$\leftarrow$
Color, Red, Green, Blue on separate lines
Red, Green, Blue on separate lines
Which statements about the pop-out effect are correct?
Basic features that pop-out correspond to the basic visual patterns processed by the brain
Pop-out is an individual property
In feature space, the smaller the distance between target and distractors the smaller the pop-out effect
The pop-put effect depends on practice, familiarity and number of distractors
Answer: A and C
Which statements about data join with keys are correct?
By adding a key to the data-join, updates can occur anywhere in the data array.
If a key is not specified, then the first datum in values is assigned to the first element in the selection and so on.
Is usually implemented by formatting the data as an array of objects with some key-value property for the keys.
Is realized by specifying the keys in the second argument in selection.data(values, [, key])
Answer: A, B, C and D
Which plots are valid for the following data?
73 42 67 78 99 84 91 82 86 122
A. The decimal point is 1 digit(s) to the right of the | C. The decimal point is 1 digit(s) to the right of the |
4 | 2 4 | 2
6 | 738 6 | 837
8 | 24619 8 | 91642
10 | 10 |
12 | 2 12 | 2
B. The decimal point is 1 digit(s) to the right of the | D. The decimal point is 2 digit(s) to the right of the |
4 | 2 0 | 4
5 | 0 | 7788899
6 | 7 1 | 02
7 | 38
8 | 246
9 | 19
10 |
11 |
12 | 2
Sorted values: 42 67 73 78 82 84 86 91 99 122
Answer: A, B and D
Which plots are valid for the following data?
10 1 0.1 7.5 8.8 2 3.5 6.7
A. The decimal point is at the | C. The decimal point is at the |
0 | 10 0 | 1
2 | 05 1 | 0
4 | 2 | 0
6 | 75 3 | 5
8 | 8 4 |
10 | 0 5 |
6 | 7
7 | 5
8 | 8
9 |
10 | 0
B. The decimal point is at the | D. The decimal point is 1 digit(s) to the right of the |
0 | 01 0 | 0124
2 | 50 0 | 789
4 | 1 | 0
6 | 57
8 | 8
10 | 0
Sorted values: 0.1 1.0 2.0 3.5 6.7 7.5 8.8 10.0
Answer: A, C and D