Skip to content

Visualization

“The greatest value of a picture is when it forces us to notice what we never expected to see.” -- John W. Tukey

Tools for data visualization:

scene Package
basics data visualization ggplot2, lattice, grid
interactive plots plotly, Shiny
more advanced challenges D3.js

7. Visualizing data distributions

  • Variable types: categorical (ordinal or not) and numeric (discrete or continuous)
    • discrete numeric data can be considered ordinal
  • Distributions:
    1. empirical cumulative distribution function (eCDF)
    2. histograms
    3. smooth density plots: make it easier to compare two distributions
    4. normal distribution: use mean and standard deviation to describe the distribution.
  • Boxplots: provides a five-number summary composed of the range (the minimum and maximum) along with the quartiles (the 25th, 50th, and 75th percentiles)
  • stratification: Dividing observations into groups based on the values of one or more variables associated with those observations.
    • the resulting groups is strata.
  • quantile: are cutoff points that divide a dataset into intervals with set probabilities quantile(data, q)
  • pnorm(a, avg, s) gives the value of the cumulative distribution function  for the normal distribution defined by average avg and standard deviation s.
  • Quantile-quantile plots, or QQ-plots, are used to check whether distributions are well-approximated by a normal distribution.

8. ggplot2

  • benifit: easy for beginners; more intuitive for beginners (it uses a grammar of graphics)
  • limitation: work exclusively with data tables in tidy format.
  • ggplot2 consists of some components
    1. Data
    2. Geometry
    3. Aesthetic mapping

DATA |> ggplot() + LAYER 1 + LAYER 2 + … + LAYER N

  • Packages: ggplot2, ggthemes, ggrepel
  • some functions in ggplot2:
    • barplots: geom_bar
    • histograms: geo_histogram
    • density plots: geo_density
    • boxplots: geo_boxplot
    • images: geom_tile and geom_raster

ggplot2 cheatsheet pdf ggplot2-1 ggplot2-2

S
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
library(ggthemes)
library(ggrepel)

r <- murders |> 
  summarize(rate = sum(total) /  sum(population) * 10^6) |>
  pull(rate)

murders |> 
  ggplot(aes(population/10^6, total)) +   
  geom_abline(intercept = log10(r), lty = 2, color = "darkgrey") +
  geom_point(aes(col = region), size = 3) +
  geom_text_repel(aes(label = abb)) + 
  scale_x_log10() +
  scale_y_log10() +
  labs(title = "US Gun Murders in 2010",
                x = "Populations in millions (log scale)", 
                y = "Total number of murders (log scale)",
                color = "Region") +
  theme_economist()

9. Data visualization principles

  • Encoding data using visual cues: position, aligned lengths, angles, area, brightness, color hue
    • In general, when displaying quantities, position and length are preferred over angles and/or area. Brightness and color are even harder to quantify than angles.
    • But they are sometimes useful when more than two dimensions must be displayed at once.
  • Know when to include 0
    • When using length as a visual cue, it is misinformative not to start the bars at 0.
    • When using position rather than length, it is then not necessary to include 0
  • Do not distort quantities
  • Order categories by a meaningful value
  • Show the data
S
1
2
3
4
# dot plot showing the data
heights %>% ggplot(aes(sex, height)) + geom_point() 
# jittered, alpha blended point plot 
heights %>% ggplot(aes(sex, height)) + geom_jitter(width = 0.1, alpha = 0.2)
  • Ease comparisons
    • Use common axes
    • Aligning plots for comparisons
  • Transformations
    • log transformation
    • logistic transformation
    • square root transformation
  • Visual cues to be compared should be adjacent
  • Think of the color blind
S
1
2
3
4
5
color_blind_friendly_cols <- c("#999999", "#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", "#D55E00", "#CC79A7") 
p1 <- data.frame(x = 1:8, y = 1:8, col = as.character(1:8)) %>%  
    ggplot(aes(x, y, color = col)) +  
    geom_point(size = 5)  
p1 + scale_color_manual(values = color_blind_friendly_cols)
  • Plots for two variables
    • Slope charts
    • Bland-Altman plot
  • Encoding a third variable
    • We encode categorical variables with color and shape.
    • For continuous variables, we can use color, intensity, or size.
  • Avoid pseudo-three-dimensional plots
  • Avoid too many significant digits: options(digits = 3)
  • Know your audience
    1. for our own exploratory data analysis
    2. to convey a message to experts
    3. to help convey a message to a general audience

10  Data visualization in practice

Case Study 1

Case Study 2

  • Logistic transformation: \(f(p)=\log(\frac{p}{1-p})\): This scale is useful when we want to highlight differences near 0 or 1