I used the volcano data (matrix) to plot a filled contour plot.
filled.contour(volcano, color.palette = terrain.colors, asp = 0.5)
Now I want to add a dot to the plot. The dot should show the highest point of the volcano. Is it possible to add a dot without using ggplot2? If yes, how?
I thought you could just find the location of the maximum and add it using the points() function, but that doesn't work because the axis coordinates displayed in the plot are not the same as the internal coordinates the plot uses to place the plot elements (see here for details).
Instead, you can use the plot.axes argument of filled.contour to get the maximum in the right place, as explained here. We also need to use the axis function to draw the axis ticks and labels, because the plot.axes argument overrides the default axes.
# Get coordinates of maximum
max.point = (which(volcano==max(volcano), arr.ind=TRUE) - 1)/(dim(volcano) - 1)
# Use plot.axes argument to plot maximum point
filled.contour(volcano, color.palette = terrain.colors, asp = 0.5,
plot.axes={
points(max.point, col="red", pch=16)
axis(side=1)
axis(side=2)
}
)
For reference, here's what happens when you try adding the point after the plot is created:
filled.contour(volcano, color.palette = terrain.colors, asp = 0.5)
points(max.point, col="red", pch=17)
Here's a ggplot2 version:
library(tidyverse)
cols = terrain.colors(3)
as.data.frame(t(volcano)) %>%
rownames_to_column(var="row") %>%
gather(col, value, -row) %>%
mutate(col=(as.numeric(gsub("V","",col)) - 1)/(nrow(volcano) - 1),
row=(as.numeric(row) - 1)/(ncol(volcano) - 1)) %>%
ggplot(aes(col, row, z=value)) +
geom_raster(aes(fill=value)) +
geom_contour(colour="grey50", size=0.2) +
geom_point(data=. %>% filter(value==max(value)),
colour="red", shape=16, size=2) +
coord_fixed(ratio = ncol(volcano)/nrow(volcano), expand=FALSE) +
scale_fill_gradient2(low=cols[1], mid=cols[2], high=cols[3],
midpoint=mean(volcano)) +
theme_minimal()
