Making beautiful and customizable heatmaps just got way easier… Introducing the superheat R package!
Using superheat, it is now extremely easy to produce plots like the example below describing 10 randomly selected cars from the famous mtcars dataset.
library(superheat)
set.seed(1347983)
selected.rows <- sample(1:nrow(mtcars), 10)
X.col <- matrix("black", ncol = ncol(mtcars), nrow = 10)
X.col[scale(mtcars[selected.rows, ]) < 0] <- "white"
superheat(mtcars[selected.rows,],
# add text
X.text = round(as.matrix(mtcars[selected.rows, ])),
X.text.col = X.col,
# scale columns
scale = T,
# label aesthetics
left.label.size = 0.5,
bottom.label.size = 0.15,
bottom.label.text.angle = 90,
bottom.label.text.alignment = "right",
bottom.label.col = "white",
# dendrogram
row.dendrogram = T,
# top plot
yt = sapply(mtcars, function(x) cor(x, mtcars$hp)),
yt.plot.type = "bar",
yt.axis.name = "correlation\nwith\nhorsepower",
# column order
order.cols = order(sapply(mtcars, function(x) cor(x, mtcars$hp))),
# grid lines
grid.vline.col = "white",
grid.hline.col = "white")
To see more details on what you can do with superheat, see the vignette, as well as our paper outlining case studies using Superheat currently available on arXiv.
