LECTURE 1 - INTRO

Barrie Robison

2024-01-10

WHO AM I?

Barrie Robison

Department of Biological Sciences

Institute for Interdisicplinary Data Sciences

Polymorphic Games

University of Idaho

WHO ARE YOU?

“…The course is designed to be”discipline agnostic” - each student is encouraged to use data sets that they deem important / interesting. The goal is to have students learn how to develop visualizations that are relevant to their own disciplinary interests…”

Briefly:

Your name
Your discipline
Your degree progress
Your technical proficiency with data visualization

COURSE SUMMARY

Students completing this course will be able to:

Describe and manipulate tabular, network, and spatial data; transform these data into a form suitable for visualization.
Make effective data visualization design choices related to marks and channels, spatial arrangement, and components of color.
Design effective data visualizations for tabular, network, and spatial data with quantitative and categorical attributes.
Implement their data visualization designs using existing tools in R (or other toolkits preferred by the student).
Explain whether a visual encoding is perceptually appropriate for a specific combination of task and data.
Demonstrate their skills with at least two novel visualizations suitable suitable for inclusion in an online Data Science Portfolio.

The course materials are located on Canvas and the course website.

VISUALIZATION

Computers provide visual representations of datasets designed to help people carry out tasks more effectively.

Tamara Munzner

Department of Computer Science

InfoVis Group

University of British Columbia

THE HUMAN

Why have a human in the loop?

Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively.

We don’t need visualization when a trusted fully automatic solution exists.
Visualization is suitable when there is a need to augment human capabilities rather than replace people with computational decision-making methods.

WHEN TO VISUALIZE

Visualization is useful when:

The analysis problem is ill-specified and we don’t know exactly what questions to ask in advance.
We are interested in long-term use for end users (ex: exploratory analysis of scientific data).
We are presenting known results (ex: DATA JOURNALISM - New York Times Upshot).
We need a stepping stone to assess requirements before developing models.
Developers of an automatic solution want to refine & determine parameters.
We need to help end users of automatic solutions verify and build trust.

THE REPRESENTATION

Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively.

EXTERNAL REPRESENTATIONS: Replace cognition with perception.

WHY DEPEND ON VISION?

Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively.

The human visual system is a high-bandwidth channel to the brain.
Overview is possible due to background processing, providing the subjective experience of seeing everything simultaneously.
Significant processing occurs in parallel and pre-attentively.
What about sound? lower bandwidth and different semantics, overview not supported, subjective experience of sequential stream.
What about touch/haptics? impoverished record/replay capacity, only very low-bandwidth communication thus far.
What about taste, smell? no viable record/replay devices.

WHY REPRESENT (ALL THE) DATA?

Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively.

summaries lose information
details matter
confirm expected and find unexpected patterns
assess validity of statistical model
ANSCOMBE’S QUARTET is a fun example that we shall use to illustrate these points!

ANSCOMBE’S QUARTET

Code

library(ggplot2)
library(grid)
library(gridExtra)
library(datasets)
library(tidyverse)
library(dplyr)
datasets::anscombe

   x1 x2 x3 x4    y1   y2    y3    y4
1  10 10 10  8  8.04 9.14  7.46  6.58
2   8  8  8  8  6.95 8.14  6.77  5.76
3  13 13 13  8  7.58 8.74 12.74  7.71
4   9  9  9  8  8.81 8.77  7.11  8.84
5  11 11 11  8  8.33 9.26  7.81  8.47
6  14 14 14  8  9.96 8.10  8.84  7.04
7   6  6  6  8  7.24 6.13  6.08  5.25
8   4  4  4 19  4.26 3.10  5.39 12.50
9  12 12 12  8 10.84 9.13  8.15  5.56
10  7  7  7  8  4.82 7.26  6.42  7.91
11  5  5  5  8  5.68 4.74  5.73  6.89

Anscombe’s Quartet

The four x-y pairs have identical summary statistics.

Code

tidy_anscombe <- anscombe %>%
 pivot_longer(cols = everything(),
              names_to = c(".value", "set"),
              names_pattern = "(.)(.)")
tidy_anscombe_summary <- tidy_anscombe %>%
  group_by(set) %>%
  summarise(across(.cols = everything(),
                   .fns = lst(min,max,median,mean,sd,var),
                   .names = "{col}_{fn}"))
#> `summarise()` ungrouping output (override with `.groups` argument)

vars<-c("set", "x_mean", "x_var",  "y_mean", "y_var")
thing<- as.data.frame(tidy_anscombe_summary[vars])
knitr::kable(thing)

set	x_mean	x_var	y_mean	y_var
1	9	11	7.500909	4.127269
2	9	11	7.500909	4.127629
3	9	11	7.500000	4.122620
4	9	11	7.500909	4.123249

VIZ MATTERS

Code

ggplot(tidy_anscombe,
       aes(x = x,
           y = y)) +
  geom_point() +
  geom_point(data = tidy_anscombe_summary, aes(x=x_mean, y = y_mean, color = "red", size = 5),
             show.legend = FALSE)+
  facet_wrap(~set) +
  geom_smooth(method = "lm", se = FALSE)

RESOURCE LIMITATIONS

Visualization designers must take into account three very different kinds of resource limitations:

Limitations of computers.
Limitations of humans.
Limitations of displays.

COMPUTATIONAL LIMITS

CPU time

System Memory

DISPLAY LIMITS

Pixels are precious and are the most constrained resource.

Information Density: ratio of space used to encode information vs unused whitespace.
There is a tradeoff between clutter and wasting space.
Designer must find the sweet spot between dense and sparse.

HUMAN LIMITS

Time
Memory
Attention