I would like to create a 2d density plot where depth of color represents density.
So for the data set
x1 <- rnorm(100, mean=0)
y1 <- rnorm(100, mean=0)
x2 <- rnorm(100, mean=4)
y2 <- rnorm(100, mean=4)
df <- data.frame(x = c(x1, x2), y=c(y1, y2), cat=rep(c("A", "B"), each=100))
I could make a contour plot
ggplot(df, aes(x=x, y=y, color=cat)) +
geom_density2d()
…but instead of contours, I’d like for the image to have a strong orange color at the peak of the cat=”A” data, fading out to transparency as the density decreases, and a strong cyan color at the peak of the cat=”B” data, fading out as the density decreases.
It seems like some combination of geom_tile() and aes(alpha=cat) would do this, but I can’t quite figure it out.
Do this:
# Create density plot with custom colors for categories A and B
ggplot(df, aes(x = x, y = y)) +
stat_density_2d(
aes(fill = cat, alpha = after_stat(level)),
geom = "polygon",
color = NA
) +
scale_fill_manual(
values = c("A" = "darkorange", "B" = "cyan2"),
name = "Category", # Legend title
guide = "legend"
) +
scale_alpha(range = c(0, 0.5), guide = "none")
refer this for colors.
For completenes, me personally I would do something like this:
# Create density plot with custom colors and density outlines
ggplot(df, aes(x = x, y = y)) +
stat_density_2d(
aes(fill = cat, alpha = after_stat(level)),
geom = "polygon",
color = NA
) +
geom_density_2d(
aes(color = cat),
size = 0.5, # Line thickness for the contours
linetype = "solid"
) +
scale_fill_manual(
values = c("A" = "darkorange", "B" = "cyan2"),
name = "Category", # Legend title
guide = "none"
) +
scale_alpha(range = c(0.0, 1), guide = "none") +
scale_color_manual(
values = c("A" = "darkorange", "B" = "cyan2")
) +
theme_minimal() +
labs(title = "2D Density Plot by Category with Density Outlines",
x = "X",
y = "Y")
gives you a clearer picture!
0