• Create a wide variety of interactive graphics using the plotly and ggiraph packages
• Learn the basics of shiny
• Create some animations (not really interactive but sometimes useful)

• A very rich package, but has the key advantage of one amazing function: ggplotly

library(plotly)
library(palmerpenguins)
p <- ggplot(data = penguins, 
       aes(x = bill_length_mm, y = flipper_length_mm)) + 
  geom_point()
ggplotly(p)

• You can use ggplotly with more or less any ggplot that you have created
• However it doesn’t work for every type of plot, and doesn’t always show exactly what you want in the interactive elements
• There are extra options that will allow you to adjust what is shown in the tooltip which is the highlighted information when you hover your mouse over the plot

• plotly also has its own syntax for creating individual plots

plot_ly(data = penguins, x = ~bill_length_mm, 
        y = ~bill_depth_mm, color = ~sex, 
        type = 'scatter', mode = 'markers')

plot_ly(data = penguins, x = ~species, y = ~bill_depth_mm, type = 'box')

plot_ly(data = penguins, x = ~bill_length_mm, 
        y = ~bill_depth_mm, z = ~flipper_length_mm, 
        type = 'scatter3d', mode = 'markers')

• A very comprehensive support website at https://plotly.com/
• Versions for all the major data analytics languages
• A lot of overlap with ggplot2; though has never been quite as popular

• A more direct and controllable way of producing interactive graphs is using ggiraph
• Just replace geom_X with geom_X_interactive
• Can then control hover behaviour using tooltip, clicking behaviour using onclick, and the data behaviour using data_id

library(ggiraph)
p <- iris %>% mutate(Row.name = 1:n()) %>% 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width, colour = Species,
             tooltip = Row.name, data_id = Species)) +
           geom_point_interactive()
girafe(ggobj = p, width = 10)

• The shiny package allows you to create beautiful GUIs which open and interact with an internet browser
• A shiny application involves (at least) two R functions. The first function, called ui.R contains the details of what’s actually going to appear on the webpage. The second, server.R, contains all the details of the R code that needs to be run
• The website https://shiny.rstudio.com contains some nice examples to get started
• See http://https://shiny.rstudio.com/gallery/ for some really impressive examples - bear in mind these are not particularly complicated to create

• Click File > New File > Shiny Web App
• Choose a name
• Click Run App

The animation package contains functions which will turn static R plots into animations: gifs, html, swf, etc The general framework is:

1. Set the animation options using ani.options
2. Put the animation frames for each plot in a loop
3. Call the ani.pause function to stop the animation running too fast
4. Wrap the above in a saveGIF/saveSWF/saveHTM function call to output to a file

(It’s a bit like using the pdf and similar commands in base R plotting)

library(animation)
oopt = ani.options(interval = 0.2, nmax = 100)
for (i in 1:ani.options("nmax")) {
    plot(density(rnorm(10*i)),
         main = 10*i)
    ani.pause()
}

Has special function called transition_state which works a little bit like facet_wrap but overlays the facets

library(gganimate)
library(gifski)
p <- ggplot(data = penguins, aes(x = flipper_length_mm, 
                                 y = body_mass_g)) +
  geom_point() +
  # Add an animation of the penguins' species over time
  transition_states(species, transition_length = 2, 
                    state_length = 1) +
  # Add a label for the current species
  labs(title = "Penguin species: {closest_state}")
animate(p, renderer = gifski_renderer())

library(gapminder)
p <- ggplot(gapminder, aes(gdpPercap, lifeExp, 
                           label = country,
                           size = pop, 
                           colour = country)) +
  geom_point(alpha = 0.7, show.legend = FALSE) +
  #geom_text() + 
  scale_colour_manual(values = country_colors) +
  scale_size(range = c(2, 12)) +
  scale_x_log10() +
  facet_wrap(~continent) +
  labs(title = 'Year: {frame_time}', x = 'GDP per capita', y = 'life expectancy') +
  # Here comes the gganimate specific bits
  transition_time(year) +
  ease_aes('linear')
animate(p, renderer = gifski_renderer())

• Great manual/website at https://gganimate.com
• Lots of different transition functions for all kinds of different functionality
• Loads of other different functions for behaviour of background data, movements between frames, scale changes, aesthetic changes, etc
• Can output the files in a gif or other movie file using the animate function

• plotly is the way to start if you just want to create a simple animation from a ggplot. If you want more control use ggiraph
• Use shiny for full dashboard control of plots and any other R feature. A whole other course required to understand this fully
• gganimate a great package for creating animations (though also see plotly which can add a play button)