Information Visualization Quiz 2

Definition of Information Visualization

The activity of transforming abstract data into visual representations to improve human cognition.

Why Visualization?

Tells a story about data

Discovers stories from data

Amplifies Cognition

-Expand Working Memory

-Reduce searching time

-Pattern detection

-Inference

Challenges in Visualization

1) Principles for designing effective visualizations

2)Tools to produce them

3) Large data

4) Common language

Visualization

Input: data

output: visual elements

Can be seen as a map between quantifiable features to aesthetics

Aesthetics

Describe aspects of graphical elements

position

shape

size

color

line width

line type

Quantitate data

Data associated with mathematical models and statistical techniques used to analyze spatial location and association.

Qualitive Data

information scientists first collect, describes anything physical (color, odor, shape)

Scales

Map data values onto aesthetics

Important: keep scales to the min required - for human cognition

Typically:

-two positions

-one color

-but many other possibilities

Coordinate Systems and axes

first step in constructing a visual representation

Cartesian

Most used coordinate system

axes can be different units

can be transformed

Log Scale

Large skewness in data

Showing multiplicative factors/ ratios

Square root scale

Need the use of zero, log cannot use 0

Polar coordinates

The location of a point as given by an angle and a distance.

Good for periodic patterns

Color in visualizations

Distinguish groups

represent values

highlight aspects

Color for grouping

Distinguish discrete items or groups without intrinsic ordering

Qualitative Color Scale

-Finite set of colors

-Clearly distinct from eachother

-No one color should stand out

-No impression of ordering

Color to represent values

Using color as an additional dimension in visualization

Sequential Color Scale

-Usually infinite set of colors

-Smoothly-varying

-Perceptually uniform scale

-Clear impression of ordering

Diverging Color Scale

-Same as sequential but there is a clear impression of order but with a clear definition for a mid-point

Color to Highlight

Accent Color Scale

-Finite or infinite set of colors with a subset that clearly stands out

-can be a modification of qualitive, sequential or divergent scales

Visualizations of Amounts

Bars/ dot plots/ Stacked bars/ heatmap

Visualization of Distributions

Histograms/ Density Plot/ Cumulative Density/ Quantile-Quantile Plot/ Boxplots/ Violins/ Strip Charts/ Sina Plots/ Stacked Histograms/ Overlapping Densities/ Ridgeline Plot

Visualization of Proportions

Pie/ Bars/ Stacked Bars/ Muli Pie/ Grouped Bars/ Stacked Bars/ Stacked Densities/Mosaic Plot/ Treemap/ Parallel Sets

Visualizations of Relationships

Scatterplots/ Bubble Charts/ Paired Scatterplot/ Slopegraph/ Density Contours/ 2D bins/ Hex Bins/ Correlogram/ Line graph/ Connected Scatterplot/ Smooth line Graph

Visualization of Geospatial data

Maps/ Choropleth/ Cartogram/ Cartogram Heatmap

Visualization of Uncertainty

Direct interpretation: Error bars

More Detailed: Confidence Strips/ Eyes/ Half-Eyes/ Quantile Dot Plot

Trend Lines: Confidence Band/ Graded Confidence Bands/ Fitted Draws

Visualization of Distributions Strenghts/Weakness

Useful to visualize how a set of numbers is distributed

Data has at lease one set of values

Histogram

a bar graph depicting a frequency distribution

Density Plot

similar to a histogram except that rather than having a summary bar representing the frequency of scores, it shows each individual score as a dot. They can be useful for looking at the shape of a distribution of scores

Cumulative Densities

No Arbitrary Parameters choices

All data displayed

but less intuitive

Skewed Distributions

Existence of heavy tails make visualization challenging

Decay is slower than exponential

Q-Q plot

Quantile-quantile plot -- a plot comparing the quantiles of two data sets, or one data set and a distribution, to see whether they might have a common distribution.

Box Plot

A graph that displays the highest and lowest quarters of data as whiskers, the middle two-quarters of the data as a box, and the median

Invented between 50s-70s by Mary Eleanor Spear and John Tukey

Simple to draw

Hides some information

Violin Plot

a graph that shows an approximation of the frequency distribution of a numerical variable in each group and its mirror image

Invented in 1998 by Hintze and Nelson

Very simple to read

Require computers to produce correctly

Can still be misleading

Strip Charts

uses strips of paper to record M mode or spectral doppler information

User jittering to avoid overplotting

Sina Plot

Combination of

Ridgeline

Multiple density plots on a y-axis

Very effective to represent trends along time

Evokes a more intuitive understanding