Using images as points in ggplot2

library(ggalign)
#> Loading required package: ggplot2
#> 
#> Attaching package: 'ggalign'
#> The following object is masked from 'package:ggplot2':
#> 
#>     element_polygon

geom_magick() Reads an image with magick, applies optional processing, and uses the result as the graphical shape for points in a plot.

This is useful when you want to replace the usual point symbols with arbitrary images while keeping full control over their placement, size, and interpolation.

set.seed(123)
d <- data.frame(
    x = rnorm(10),
    y = rnorm(10),
    image = "https://jeroenooms.github.io/images/frink.png",
    fill = sample(c("A", "B", "C", "D"), 10, replace = TRUE),
    alpha = rnorm(10, mean = 0.5, sd = 0.1)
)
d$alpha <- pmax(pmin(d$alpha, 1), 0)

You can combine the image with a fill color and transparency mapping. The fill is applied on top of the image, allowing for color-coded categories without losing the image shape. The size aesthetic controls the image width in "mm".

ggplot(d, aes(x, y)) +
    geom_magick(
        aes(image = image, fill = fill, alpha = alpha),
        size = 12, show.legend = FALSE
    ) + 
    coord_cartesian(clip = "off")