Animated Scatter Plots

Scatter plots are useful for conveying information about a number of points. By animating, we can add a new dimension to the graph.

For all of the basic animations, we'll use the txhousing dataset within the ggplot2 library, for simplicity. This dataset contains "information about the housing market in Texas provided by the TAMU real estate center".

At the end of this module, we'll have made this graph:

---

As we did in the previous section, we'll load all the libraries.

library(ggplot2) #plotting and getting the txhousing dataset
library(dplyr) #sorting the data to fit our graph
library(gganimate) #animating the ggplot

Again, we'll put the dataset into a data frame so that it's easier to see changes.

txhousing_data <- txhousing

For this plot, we need to have a data point (dot) for each city for all time frames. This point needs to have both the median sale price and the sales. Lucky for us, the dataset is already in this format:

We can go straight to plotting. The syntax is very similar to the line graph, but note that we now use geom_point instead of geom_line. We're also going to add a new parameter: group. All this does is tell R that we'd like each point to represent one city

ggplot(txhousing_data)+
  geom_point(aes(x=sales, y=median))

That's a lot of data, but for now, let's just focus on the general patterns. There's somewhat of a logarithmic curve to the data, so we'll apply a logarithmic scale to the x-axis. Let's also get rid of the scientific notation on the y-axis by adding commas.

ggplot(txhousing_data)+
  geom_point(aes(x=sales, y=median))+
  scale_x_continuous(label=scales::comma, transform = "log10")+ #requires 'scales' package is installed
  scale_y_continuous(label=scales::comma) #requires 'scales' package is installed

Once we've done that, we can make the graph look a bit nicer before we get into the animation stages. <br> is added for spacing.

ggplot(txhousing_data)+
  geom_point(aes(x=sales, y=median))+
  scale_x_continuous(label=scales::comma, transform = "log10")+
  scale_y_continuous(label=scales::comma)+
  theme_minimal()+
  labs(title="Texas Housing Market from 2000-2015",
       subtitle="House prices and sales for each dot/city have increased since 2000, <br>especially post-2008.",
       x="Sales",
       y="Median Sale Price")+
  theme(plot.title = ggtext::element_markdown(size = 22, hjust =0.5, face = "bold"), 
        plot.subtitle = ggtext::element_markdown(size = 15, hjust =0.5, face = "bold"))

We're now ready to animate. The graph might look like a mess right now, but remember that each point corresponds to a city and a month. Our goal for this animation is to show all the points for a given month, cycling through the months.

This can easily be accomplished with transition_time, which does exactly this: cycle through a time parameter, showing the data at each point in time (smoothing through transition periods). Our time parameter in this case is date, so we call transition_time(date).

There's one more thing we have to do before we plot, which is adding a group parameter to the geom_point call. This step is very important (and often forgotten) - we're essentially telling R that we'd like to have the points grouped based on an aesthetic. In this case, we want each point to represent a city, so we use group=city:

Lastly, it would be nice if the date was displayed somewhere on the graph. To do so, let's modify the subtitle parameter with the keyword {frame_time} which gives the current value of the variable we're animating over (date in this case). For simplicity, let's just round down date (which has a decimal value for months) with the as.integer function, to show each year rather than each year/month.

ggplot(txhousing_data)+
  geom_point(aes(x=sales, y=median, group=city))+ #added 'group=city'
  scale_x_continuous(label=scales::comma, transform = "log10")+
  scale_y_continuous(label=scales::comma)+
  theme_minimal()+
  labs(title="Texas Housing Market from 2000-2015",
       subtitle="House prices and sales for each dot/city have increased since 2000, <br>especially post-2008. <br><br>Year: {as.integer(frame_time)}", #changed line
       caption="Nikhil",
       x="Sales",
       y="Median Sale Price")+
  theme(plot.title = ggtext::element_markdown(size = 22, hjust =0.5, face = "bold"), 
        plot.subtitle = ggtext::element_markdown(size = 15, hjust =0.5, face = "bold"))+
  transition_time(date) #added line

Our animation is ready to go. To finalize its dimensions, frames, and resolution, let's put it into an object and animate it.

animation <- ggplot(txhousing_data)+
    #continues from above code snippet

animate(animation, fps=30, duration=60, end_pause=150, height = 7,
        width = 9, units = "in", res = 200)

Note that the increase in FPS is just to make the animation a bit smoother. Anything more than 10 fps should be good to see a smooth animation.

Congrats! You've just made an animated scatter plot.

Click here to view the raw file used to make this animation.