another exercise

Author

…

Published

April 8, 2026

Languages data

The Foreign Service Institute (FSI) has created a list to show the approximate time you need to learn a specific language as an English speaker.

library(ggplot2)
library(tidyverse)

tribble(~language, ~level, ~weeks,
        "Spanish",    1,   23.5,
        "Portuguese", 1,   23.5,
        "German",     2,   30,
        "Mandarin",   5,   88) ->
  language_levels


language_levels |> 
  ggplot() + 
  # order this variable as it appears in data
  # else we'll be alphabetical
  aes(x = language |> fct_inorder()) +
  aes(y = weeks) + 
  geom_col()

Esperanto?

Esperanto is the most successful constructed international auxiliary language, and the only such language with a sizable population of native speakers. https://en.wikipedia.org/wiki/Esperanto

1/4 of the time as non-constructed languages!?!

Is ggplot2 like Esperanto?

–

Consistency and discipline will keep the language

  • predictable
  • easy to learn
  • easy to use

… Though sometimes there’s room for improvement…

library(ggplot2)
library(tidyverse)

p <- language_levels |> 
  ggplot() + 
  aes(x = language |> fct_inorder(), 
      y = weeks) + 
  geom_col(width = .3)

p

p + coord_polar() # Adds interest? Harder to interpret?  

p + coord_equal() # 1:1 aspect ratio (Not great choice for this data)

In what way is ggplot2 like English? (even though English is not totally predictable.

English today’s ‘Lingua Franca’ - most spoken language in the world and bridge language

Zero-shot translation: https://youtu.be/xH19VRSOG7g?si=O4zOqGCdnQSOauWM&t=1007

ggplot2…

  • very successful
  • ggplot2 syntax a bridge language? between:
    • visual ‘language’ (visual channels (color, position, shape), tried and true chart types)
    • computers (programmed to interpret syntax, and render viz)
    • human language (syntax is close to how we might describe things)

“Visualization as Lingua Franca in Machine Learning”…- Fernanda Viegas

Varga Paraphrase: “data visualization is a window into statistical/ML understanding (and the phenomenon we study with statistics/ML)”

ggplot: ‘The “Layered” grammar of graphics’

What does ‘layer’ refer too?

Exercises

  1. point, tile pile
  2. lollipop

Text and annotation

  1. geom_text, geom_label
  2. annotate(“text”, label = “hello”)

Open ended exercise, use the cheat sheet or ‘grammar guide’ and create two more chart with a geom that we haven’t used. Style with labs(), theme_*(), and more!

https://evamaerey.github.io/ggplot2_grammar_guide/about

Michel Thomas?

https://www.youtube.com/watch?v=U9Xh-by50pI

https://www.youtube.com/watch?v=U9Xh-by50pI