# Lecture

- best channel to convey information = vision
	- highest bandwidth of all the senses
	- extends memory and cognition – we use drawings to “think more clearly”
	- people think visually
- pre-attentive perception
	- the time to find the blue dot is constant – no matter the total number of dots
		- assuming we can distinguish the colors
- Anscombe's Quartet – four datasets with the same statistical properties
- when computing average, we assume that it represents the underlying distribution well
- information visualization = use of computer-supported, interactive visual representations of data to amplify cognition
- scientific visualization – the datasets have a given spatialisation (the coordinates already exist), continuous
- data visualization – web-based, communication
- challenges: diversity, scale
- visual information-seeking mantra
	- overview first
	- zoom and filter
	- details on demand
- visual perception is a two stage process
	- parallel extraction of low-level properties
	- sequential goal-directed processing
- retina performs parallel processing of different attributes
- brain performs sequential processing (object segmentation, identification)
- readings: [Graphs in Statistical Analysis](http://iihm.imag.fr/blanch/teaching/infovis/readings/1973-Anscombe-Graphs_in_Stats.pdf) & [The Eyes Have It](http://iihm.imag.fr/blanch/teaching/infovis/readings/1996-Shneiderman-Mantra.pdf); viewing: [Hans Rosling’s TED talks](https://www.ted.com/speakers/hans_rosling)
- density of receptors decreases with the distance from the center of vision (so the peripheral vision is not that sharp)
- peripheral vision is mostly colorblind
- http://xkcd.com/1080/
- Gestalt psychology
	- one stimulus, two perceptions
	- there is a difference between stimulus and perception
	- emergence – we need to see the whole picture, not just parts
	- reification – perception contains more spatial information than the stimulus
	- multistability – ambiguous stimuli can generate different perceptions but they cannot coexist simultaneously
	- invariance – objects are recognized independently of various variations (transformations, lightning, …)
	- laws of grouping – law level perceptions are grouped into higher-level objects
		- good Gestalt
- information visualization pipeline
	- source data → data tables
		- data transformations
	- data tables → visual abstraction
		- visual mappings
		- transition from data form to visual form
		- we use data attributes to create (visual) marks
	- visual abstraction → views
		- view transformations
		- result … actual pixels
- example: gapminder (data attributes → visual channels)
	- income → x
	- life expectancy → y
	- country → details on demand
	- population → size
	- region → color
- interactions
	- the user can manipulate 1) data transformations, 2) visual mappings, 3) view transformations
	- zoom … view transformation
	- filtering … data transformation
	- in gapminder, we can also edit visual mappings
- taxonomies of data types
	- “what comparisons can I make?”
	- “how can I aggregate the data?”
	- nominal … only equality (=)
		- aggregation … mode (we show the most frequent one) or top-k
			- if we have a taxonomy (hierarchy), we can group the data
	- ordered … ordering and equality (<, =)
		- aggregation … median, quantiles, histogram (count per bucket)
	- quantitative … “how much smaller is it?”
		- → intervals … $v-v'$
		- → ratios … $v/v'$
			- only if there is a meaningful zero on the scale
		- aggregation … mean, std dev, skew