gganimate with changing scales (axis limits)

Question

I'd like to create a gif using gganimate, but my axis ranges vary wildly in one frame. This is causing all subsequent frames to be squeezed.

In ggplot2's facets, there's an option to have scales="free". Is there a way to have free scales in each frame of gganimate?

Here's an example:

library(gapminder)
library(ggplot2)
library(gganimate)
theme_set(theme_bw())

p <- ggplot(gapminder, aes(gdpPercap, lifeExp, size = pop, color = continent,
                           frame = year)) +
  geom_point() +
  scale_x_log10()

gganimate(p)

Now we move one of the data points to some extreme value. This squeezes the points in all subsequent unaffected frames.

gapminder[1, "lifeExp"] <- 1000
gapminder[1, "gdpPercap"] <- 1e60

p <- ggplot(gapminder, aes(gdpPercap, lifeExp, size = pop, color = continent, 
                           frame = year)) +
  geom_point() +
  scale_x_log10()

gganimate(p)  # smooshed

So you think it would be useful to have the scales calculated individually at every timestep? I don't see how that would be helpful. — Mike Wise, Dec 20 '16 at 23:19
I disagree. It would be useful for my application. Much like the facet_wrap() parameter scale="free" in ggplot2 — Megatron, Dec 21 '16 at 02:12
Only if the scale changed dramatically every frame. In my application only a single frame has a vastly different scale similar to the example here. — Megatron, Dec 21 '16 at 13:44
It's also useful for timeseries, for example here: https://www.youtube.com/watch?v=UatUDnFmNTY — Lennert, Aug 30 '17 at 11:43

score 10 · Answer 1 · answered Nov 24 '18 at 22:52

10

You can try experimenting with view_follow().

Code

p <- ggplot(gapminder, aes(gdpPercap, lifeExp, size = pop, color = continent)) +
    geom_point() +
    labs(title = 'Year: {frame_time}', x = 'GDP per capita', y = 'life expectancy') +
    transition_time(year) +
    view_follow()

animate(p)

answered Nov 24 '18 at 22:52

Roman

4,744
2
16
58

3

I would suggest to replace view_follow() with view_zoom(). It produces nicer results for time series data. – pallevillesen Mar 05 '20 at 08:36

score 2 · Answer 2 · answered Dec 31 '19 at 08:20

2

To manually define the scales see view_step and view_step_manual (also view_zoom and view_zoom_manual).

answered Dec 31 '19 at 08:20

qwr

9,525
5
58
102

gganimate with changing scales (axis limits)

2 Answers2

Code

Linked