Density Contour Plot

Density contour lines are two-dimensional Density Curves. So the relationship between density contour lines and Heatmaps is like that between Density Curves and Histograms.

Density contour plots use kernel density estimates to calculate the curves.

• The “number of bins” in a density contour plot is the number of lines
• When using facets, watch out for the boundaries
  • clustered points may be separated into two facets, causing wrong density estimates for two facets

Implementation

• kde2d in package MASS

library(GDAdata)
library(MASS)
f1 <- kde2d(SpeedSki$Year, SpeedSki$Speed, n = 100)
image(f1)
contour(f1, add = T)

• geom_density_2d()  and geom_density_2d_filled() in ggplot2

library(GDAdata)
ggplot(SpeedSki, aes(Year, Speed)) +
    geom_density_2d() +
    geom_density_2d_filled(alpha = 0.7)

• Calculate using kde2d, plot with geom_contour in ggplot2

ggplot(con2tr(f1, aes(x, y)) + geom_contour(aes(z = z))

• Smooth Scatterplot