Why are colours appearing in the labels of my gganimate sketch?-CodePudding

I have a gganimate sketch in R and I would like to have the percentages of my bar chart appear as labels.

But for some bizarre reason, I am getting seemingly random colours in place of the labels that I'm requesting.

If I run the ggplot part without animating then it's a mess (as it should be), but it's obvious that the percentages are appearing correctly.

Any ideas? The colour codes don't correspond to the colours of the bars which I have chosen separately. The codes displayed also cycle through about half a dozen different codes, at a rate different to the frame rate that I selected. And while the bars are the same height (they grow until they reach the chosen height displayed in the animation) then they display the same code until they stop and it gets frozen.

Code snippet:

df_new <- data.frame(index, rate, year, colour)
df_new$rate_label <- ifelse(round(df_new$rate, 1) %% 1 == 0, 
                            paste0(round(df_new$rate, 1), ".0%"), paste0(round(df_new$rate, 1), "%"))

p <- ggplot(df_new, aes(x = year, y = rate, fill = year))  
  geom_bar(stat = "identity", position = "dodge")   
  scale_fill_manual(values = colour)   
  #geom_text(aes(y = rate, label = paste0(rate, "%")), vjust = -0.7)   
  geom_shadowtext(aes(y = rate, label = rate_label),
                  bg.colour='white',
                  colour = 'black',
                  size = 9,
                  fontface = "bold",
                  vjust = -0.7,
                  alpha = 1
  )  
  coord_cartesian(clip = 'off')  
  ggtitle("% population belonging to 'No religion', England and Wales census")  
  theme_minimal()   
  xlab("")   ylab("")  
  theme(legend.position = "none")   
  theme(plot.title = element_text(size = 18, face = "bold"))  
  theme(axis.text = element_text(size = 14))  
  scale_y_continuous(limits = c(0, 45), breaks = 10*(0:4))
p

p <- p   transition_reveal(index)   view_follow(fixed_y = T)

animate(p, renderer = gifski_renderer(), nframes = 300, fps = frame_rate, height = 500, width = 800,
        end_pause = 0)

anim_save("atheism.gif")

CodePudding user response：

So it was answered

CodePudding user response：

I think you have missed some delicate points about ggplot2. I will try my best to describe them to you. First of all, you need to enter the discrete values as factor or integer. So you can use as.factor() before plotting or just factor() in the aesthetic. Also, you should consider rounding the percentages as you wish. Here is an example:

set.seed(2023)
df_new <- data.frame(index=1:10, rate=runif(10), year=2001:2010, colour=1:10)
df_new$rate_label <- ifelse(round(df_new$rate, 1) %% 1 == 0, 
                            paste0(round(df_new$rate, 1), ".0%"), 
                            paste0(round(df_new$rate, 1), "%"))

The ggplot for this data is:

library(ggplot2)
p <- ggplot(df_new, aes(x = factor(year), y = rate, fill = factor(colour)))  
  geom_bar(stat = "identity", position = "dodge")   
  geom_text(aes(y = rate, label = paste0(round(rate,2), "%")), vjust = -0.7)  
  coord_cartesian(clip = 'off')  
  ggtitle("% population belonging to 'No religion', England and Wales census")  
  theme_minimal()   
  xlab("")   ylab("")  
  theme(legend.position = "none",
        plot.title = element_text(size = 18, face = "bold"),
        axis.text = element_text(size = 14)) 
p

And you can combine all theme element in one theme() function (as did I). The output is:

And you can easily animate the plot using the following code:

library(gganimate)
p   transition_reveal(index)

And the output is as below:

Hope it helps.