How to plot countours around points with different alpha level of countours-CodePudding

I would like to reproduce the attached plot on the training iris dataset with Petal.Length (x) and Sepal.Length (y). I assume it is about stat_density_2d arguments, but I was not able to reach the exact same picture. Any help is highly appreciated.

enter image description here

CodePudding user response：

You need to map alpha to the calculated level within stat_density2d_filled, using alpha = after_stat(level) inside aes. Just tweak the breaks argument to define where you want the contour cutoffs to be. I also found that adjusting the h argument to set a slightly wider bandwidth than the default worked better than the default setting with this particular data set. Additionally, a call to scale_alpha_manual will give you fine control over the opacity of each band - here I set a sequence of six values between 0 and 1, since only the first 5 will be used, meaning that the most dense bands still won't be fully opaque and "overpower" the points.

library(ggplot2)

ggplot(iris, aes(Petal.Length, Sepal.Length, fill = Species))  
  stat_density2d_filled(aes(alpha = after_stat(level)),
                        h = c(1, 1), breaks = c(0, 0.03, 0.1, 0.25, 0.5, 5))  
  geom_point(shape = 21, size = 3)  
  scale_alpha_manual(values = seq(0, 1, length = 6))  
  theme_minimal(base_size = 20)  
  coord_equal(expand = FALSE)  
  theme(legend.position = "none",
        panel.border = element_rect(color = "gray80", fill = NA))