Visualizing Innovation

Visualization

Dr. Matthias Schnetzer

October 16, 2023

Pioneers of data visualization

Milestones in data visualization

William Playfair (1759-1823)

John Snow (1813-1858)

Florence Nightingale (1820-1910)

Charles Joseph Minard (1781-1870)

W.E.B. Du Bois (1868-1963)

Otto Neurath (1882-1945)

Five guidelines for better visualization

1. Choose the adequate type of graph

Bad:

Good:

Choose a chart type with respect to your data (numeric, categorical, ranking, time series etc.)
What would you like to show: comparison, distribution, composition, relationship?
To help you find the adequate type from a myriad of alternatives, you can also take a look at https://www.data-to-viz.com and https://datavizproject.com

A simple decision tree of chart types

2. Visualize data accurately and faithfully

Bad:
Good:

Prioritize data accuracy, clarity, and integrity
Avoid misleading the reader by truncating the y-axis, using two different y-axis, cherry-picking data, not providing context, etc.
A good story based on data visualization does not involve deceptive manipulation of the data!

Example

Another example

A third example

3. Integrate graphics and text

Bad:
Good:

Don’t make people turn their head to read labels
Think about a logical order of the chart (alphabetical, values)
Add direct labels rather than a legend
Choose a meaningful title that focuses on your message

Best practice

4. Reduce the clutter

Bad:
Good:

Unnecessary visual elements distract the readers from the central data
Avoid elements that do not contain information!
Basic elements like heavy tick marks or gridlines should be removed
Think carefully which visual elements are really needed to read the chart

Best practice

5. Avoid the spaghetti chart and start with gray

Bad:

Good:

When the graph contains too much information, it looks like spaghetti
Try to break overloaded single charts into smaller parts (facets, small multiples) or highlight the relevant information
Start with gray: you are forced to be strategic in the use of color, labels, etc.

Best practice

Deceptive graphs

Misleading titles

Deceptive graphs meet their goal

The Gestalt Principles

What are the Gestalt Principles?

Gestalt Principles describe how humans group similar elements, recognize patterns and simplify complex images. “Gestalt” is German for “unified whole”.

The question is how humans typically gain meaningful perceptions from the chaotic stimuli around them. The idea is that the mind “informs” what the eye sees by perceiving a series of individual elements as a whole.

Which chart is random and which has structure in it?

Gestalt rules

Proximity Things that are spatially near to one another seem to be related.
Similarity Things that look alike seem to be related.
Connection Things that are visually tied to one another seem to be related.
Continuity Partially hidden objects are completed into familiar shapes.
Closure Incomplete shapes are perceived as complete.
Figure and Ground Visual elements are taken to be either in the foreground or the background.
Common Fate Elements sharing a direction of movement are perceived as a unit.

Our brains look for structure

Let’s start with {ggplot}

The grammar of {ggplot}

Component	Function	Explanation
Data	ggplot(data)	The raw data that you want to visualise.
Aesthetics	aes()	Aesthetic mappings between variables and visual properties.
Geometries	geom_*()	The geometric shapes representing the data.
Statistics	stat_*()	The statistical transformations applied to the data.
Scales	scale_*()	Maps between the data and the aesthetic dimensions.
Coordinate System	coord_*()	Maps data into the plane of the data rectangle.
Facets	facet_*()	The arrangement of the data into a grid of plots.
Visual Themes	theme()	The overall visual defaults of a plot.

Geometries and aesthetic mapping

data |>  
  ggplot(aes(x = year, y = value, color = category)) +
  geom_point(alpha = 0.5, shape = 21) +
  scale_color_manual(values = c("red", "blue", "yellow"))

Geometries	Aesthetics	Scales	Dimensions
points	positions (x, y)	scale_x_(), scale_y_()	position (continuous, discrete, reverse, log10, sqrt, date)
lines	colors (color, fill)	scale_color_(), scale_fill_()	colors (continuous, discrete, manual, gradient, gradient2, brewer)
polygons	shapes (shape, linetype)	scale_shape_(), scale_linetype_()	shapes (continuous, discrete, manual, ordinal)
text	size (size)	scale_size_(), scale_radius_()	sizes (continuous, discrete, manual, ordinal, area, date)
boxplot	transparency (alpha)	scale_alpha_*()	transparency (continuous, discrete, manual, ordinal, date)
…many more!	groupings (group)

First steps

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm)) +
  geom_point(size = 1.5, alpha = 0.5)

Colors

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  theme_minimal()

Scales

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  scale_color_manual(values = MetBrewer::met.brewer("Lakota")) +
  scale_x_continuous(limits = c(30,60), breaks = seq(30,60,10)) +
  scale_y_continuous(limits = c(12,21), breaks = seq(12,21,3)) +
  theme_minimal()

Labels

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  scale_color_manual(values = MetBrewer::met.brewer("Lakota")) +
  scale_x_continuous(limits = c(30,60), breaks = seq(30,60,10)) +
  scale_y_continuous(limits = c(12,21), breaks = seq(12,21,3)) +
  labs(x = "Bill length (in mm)", y = "Bill depth (in mm)",
      title = "Penguins are awesome",
      subtitle = "Depth and length of bills") +
  theme_minimal()

Annotation rather than legend

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  scale_color_manual(values = MetBrewer::met.brewer("Lakota")) +
  scale_x_continuous(limits = c(30,60), breaks = seq(30,60,10)) +
  scale_y_continuous(limits = c(12,21), breaks = seq(12,21,3)) +
  annotate("text", x = c(34.7, 55.7, 50.7), y = c(20.7, 19, 13.6), 
  color = MetBrewer::met.brewer("Lakota")[1:3], 
  label = c("Adélie","Chinstrap","Gentoo"), fontface = "bold", size = 4) +
  labs(x = "Bill length (in mm)", y = "Bill depth (in mm)",
      title = "Penguins are awesome",
      subtitle = "Depth and length of bills") +
  theme_minimal() +
  theme(legend.position = "none")

Themes

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  scale_color_manual(values = MetBrewer::met.brewer("Lakota")) +
  scale_x_continuous(limits = c(30,60), breaks = seq(30,60,10)) +
  scale_y_continuous(limits = c(12,21), breaks = seq(12,21,3)) +
  annotate("text", x = c(34.7, 55.7, 50.7), y = c(20.7, 19, 13.6), 
  color = MetBrewer::met.brewer("Lakota")[1:3], 
  label = c("Adélie","Chinstrap","Gentoo"), fontface = "bold", size = 4) +
  labs(x = "Bill length (in mm)", y = "Bill depth (in mm)",
      title = "Penguins are awesome",
      subtitle = "Depth and length of bills") +
  theme_minimal() +
  theme(legend.position = "none",
  plot.title.position = "plot",
  plot.title = element_text(size = 16, face="bold"),
  plot.subtitle = element_text(size = 13),
  panel.grid.minor = element_blank())

Calculate median and standard deviation

data_summary <- data |>
  group_by(species) |>
  summarise(across(c(bill_length_mm, bill_depth_mm),
  list(median = ~median(., na.rm = TRUE), 
       sd = ~sd(., na.rm = TRUE))))

data_summary

# A tibble: 3 × 5
  species   bill_length_mm_median bill_length_mm_sd bill_depth_mm_median
  <fct>                     <dbl>             <dbl>                <dbl>
1 Adelie                     38.8              2.66                 18.4
2 Chinstrap                  49.6              3.34                 18.4
3 Gentoo                     47.3              3.08                 15  
# ℹ 1 more variable: bill_depth_mm_sd <dbl>

Median

data |> ggplot(aes(x = bill_length_mm, 
                  y = bill_depth_mm, 
                  color = species)) +
  geom_point(size = 1.5, alpha = 0.5) +
  geom_errorbar(
    data = data_summary,
    aes(x = bill_length_mm_median,
      ymin = bill_depth_mm_median - bill_depth_mm_sd,
      ymax = bill_depth_mm_median + bill_depth_mm_sd,
      color = species,
      color = after_scale(colorspace::darken(color, .2, space = "combined"))), 
      inherit.aes = FALSE, width = .5, size = .8) +
  geom_errorbar(
    data = data_summary,
    aes(y = bill_depth_mm_median,
      xmin = bill_length_mm_median - bill_length_mm_sd,
      xmax = bill_length_mm_median + bill_length_mm_sd,
      color = species,
      color = after_scale(colorspace::darken(color, .2, space = "combined"))), 
      inherit.aes = FALSE, width = .25, size = .8) +
  scale_color_manual(values = MetBrewer::met.brewer("Lakota")) +
  scale_x_continuous(limits = c(30,60), breaks = seq(30,60,10)) +
  scale_y_continuous(limits = c(12,21), breaks = seq(12,21,3)) +
  annotate("text", x = c(34.7, 55.7, 50.7), y = c(20.7, 19, 13.6), color = MetBrewer::met.brewer("Lakota")[1:3], label = c("Adélie","Chinstrap","Gentoo"), fontface = "bold", size = 4) +
  labs(x = "Bill length (in mm)", y = "Bill depth (in mm)",
      title = "Penguins are awesome",
      subtitle = "Depth and length of bills") +
  theme_minimal() +
  theme(legend.position = "none",
  plot.title.position = "plot",
  plot.title = element_text(size = 16, face="bold"),
  plot.subtitle = element_text(size = 13),
  panel.grid.minor = element_blank())

Final plot

Bibliography

Dougherty, Jack/Ilyankou, Ilya (2021). Hands-on data visualization: Interactive storytelling from spreadsheets to code. O’Reilly Media.

Friendly, Michael (2008). A brief history of data visualization. In Chen, C./Härdle, W./Unwin, A. (Eds.), Handbook of computational statistics: Data visualization (Vol. III, 15–56). Springer, Berlin.

Healy, Kieran (2018). Data visualization: A practical introduction. Princeton University Press.

Lauer, Claire/OBrien, Shaun (2020). How people are influenced by deceptive tactics in everyday charts and graphs. IEEE Transactions on Professional Communication, 63(4), 327–340. DOI: 10.1109/tpc.2020.3032053

Scherer, Cédric (2022). Graphic design with ggplot2. https://rstudio-conf-2022.github.io/ggplot2-graphic-design/

Schwabish, Jonathan A. (2014). An economist’s guide to visualizing data. Journal of Economic Perspectives, 28(1), 209–234. DOI: 10.1257/jep.28.1.209