Ggplot2 Draw Individual Ellipses But Color by Group

ggplot2 draw individual ellipses but color by group

Explicitly define the groups:

ggplot(ggdat) +
  geom_point(aes(x=X, y=Y,color=indiv_id),size=1) + # 
  stat_ellipse(aes(x=X, y=Y,color=group_id, group=indiv_id),type = "norm") +
  theme(legend.position='none')

Points in a scatterplot with individual ellipses using ggplot2 in R

Use geom_ellipse() from ggforce:

library(ggplot2)
library(ggforce)

d <- data.frame(
  x = c(10, 20),
  y = c(10, 20),
  ns = c(5, 8),
  ew = c(4, 4)
)

ggplot(d, aes(x0 = x, y0 = y, a = ew/2, b = ns/2, angle = 0)) + 
  geom_ellipse() +
  coord_fixed()

^{Created on 2019-06-01 by the reprex package (v0.2.1)}

how to add ellipses in ggplot when using cut in geom_point

You create the groups on the fly inside geom_point (that's what cut does) but you need to use the groups again for the ellipses and their colors. So it is better the define the groups first.

library("tidyverse")
library("vegan")

data("iris")
T <- metaMDS(iris[1:4]) #perform an ordination with the package vegan

# For reproducibility
set.seed(1234)

ScIris <- as.data.frame(scores(T)) %>%
  mutate(test_stat = runif(150, 1, 100)) %>%
  mutate(group = cut(test_stat, c(0, 25, 50, 75, 85, 95, 99, 100)))

ggplot(ScIris, aes(NMDS1, NMDS2)) +
  geom_point(aes(colour = group), size = 5) +
  stat_ellipse(aes(color = group, group = group)) +
  scale_color_manual(name = "proportions",
                     values = c("(0,25]" = "black",
                                "(25,50]" = "dark green",
                                "(50,75]" = "green",
                                "(75,85]" = "yellow",
                                "(85,95]" = "orange",
                                "(95,99]" = "purple",
                                "(99,100]" = "blue",
                                labels = c("0", "25", "50", "75", "85", "95","<100")))
#> Too few points to calculate an ellipse
#> Warning: Removed 1 rows containing missing values (geom_path).

^{Created on 2019-03-31 by the reprex package (v0.2.1)}

How to use a fixed color for the stat_ellipse while having grouped data?

Use a group aesthetic and set it to the type:

ggplot(d, aes(value1, value2, color=type)) +
    geom_point() +
    stat_ellipse(aes(group = type), colour = "black") +
    facet_grid(feature1 ~ feature2)

Draw only 2 ellipses in PCA plot (instead of 20)

The problem with using autoplot is that, although it is great for producing nice visualizations of common data structures and models with little effort, it doesn't give you the full freedom to customize the plot. However, it is pretty straightforward to do the whole thing within ggplot. The following is a full reprex:

library(ggplot2)

pc <- prcomp(iris[1:4])
df <- cbind(pc$x[,1:2], iris)

ggplot(df, aes(PC1, PC2, color = Species)) + 
  geom_point() +
  stat_ellipse(geom = "polygon", aes(fill = after_scale(alpha(colour, 0.3))),
               data = df[df$Species != "versicolor",])

^{Created on 2022-02-21 by the reprex package (v2.0.1)}

---

Related Topics