An Intro to Animating Charts with gganimate

Install the gganimate package

First things first - ensure that you have ggplot2 (included in tidyverse) and gganimate packages installed, then call them via the library command:

library(tidyverse)
library(gganimate)

Plot the static visual

Before we add the gganimate code lines, it’s good practice to first plan and plot out our static visual through ggplot.

For this example we’re using the ChickWeight database built within R - which tracks the weight of 50 chicks on four different diets over time.

If we want to see the growth in weight of each chick depending on their diet, we can use the following ggplot code:

ChickWeight %>% 
  ggplot(aes(weight, Chick, color=Diet)) + 
  #plot weight on x-axis
  #plot chick reference number on y-axis
  #then differentiate the colors based on which Diet they are on
  geom_point() #use scatterplot to show weights

This chart is still too ‘noisy’, and could benefit on having an animation built in to show progress over the 21-day experiment period.

Add in the gganimate lines

In the code below, we introduce new functions from gganimate to bring our static chart to life:

• labs - labels same as with base ggplot, but adding functionality to have the timeframe label (labelling Time as “Days” for this example)
• transition_time - declares which variable the animation is based on (using Time as the variable for this example)
• ease_aes - controls easing of aesthetics (basically making the animation smoother)

ChickWeight %>% 
  ggplot(aes(weight, Chick, color=Diet)) +
  geom_point() +
  #here comes the gganimate code
  labs(title = 'Days: {frame_time}', x = 'Chick weight (grams)', y = 'Chick number') +
  transition_time(Time) +
  ease_aes('linear')

Notice that when you run this code, it takes a while to render the visual. This is because RStudio is actually making a gif of your animated chart!

With the animation, it’s more engaging to see how the experiment tracked across the days and which chick grew in weight more than the rest.

Saving our animated chart

To save your new animation, you can call on the anim_save function. The code below saves the last animation you did using a file name of your choice:

anim_save("Chick_Weights_over_time.gif", animation = last_animation(), path = )

Plot the static visual

Same as the first example, it’s best practice to plan out our static visual first. If we want to look at Oceania countries (Australia vs New Zealand) specifically, we can have a ggplot code as follows:

gapminder %>% 
  filter(continent=="Oceania") %>% #filter to just AU & NZ
  ggplot(aes(gdpPercap, lifeExp, size=pop, colour = country)) + #plot x & y axes, size, and color
  geom_point(alpha = 0.5, show.legend = TRUE) + #use scatter plot
  scale_size(range = c(2,12)) #make sizing more prominent

The static chart above is already quite compelling, but the story is incomplete since we don’t have a view of the time period data.

Animate the scatterplot

Now we introduce the gganimate lines right after ggplot. For this example, we introduce a new function from gganimate:

• shadow_wake - defines how previous data is treated, making it trail like a shadow

gapminder %>% 
  filter(continent=="Oceania") %>% 
  ggplot(aes(gdpPercap, lifeExp, size=pop, colour = country)) +
  geom_point(alpha = 0.5, show.legend = TRUE) +
  scale_size(range = c(2,12)) +
  # Here comes the gganimate specific bits
  labs(title = 'Year: {frame_time}', x = 'GDP per capita in USD', y = 'Life Expectancy') +
  transition_time(year) +
  shadow_wake(.3)+
  ease_aes('linear')

With our animated chart, it becomes evident that it was in the 70’s when Australia was able to overtake New Zealand in terms of Life Expectancy and GDP per capita!

Save your animation

Remember to save your animation as a file:

anim_save("AU_NZ_GDPpercapita_Lifeexpectancy.gif", animation = last_animation(), path = )