File:Heatmap-gaps.png: Difference between revisions

Summary

# 1. Prepare the data
set.seed(42)
mat = matrix(rnorm(30, mean = 0, sd = 1.2), nrow = 5, ncol = 6)
mat[2, 1] = -2.5
mat[2, 6] = 2.5
mat[1, 2] = -1.8
mat[4, 4] = 1.8

# 2. Define the color palette
library(circlize)
col_fun = colorRamp2(c(-2.5, 0, 2.5), c("steelblue", "white", "darkred"))

# 3. Create the heatmap object
library(ComplexHeatmap)
ht = Heatmap(
  mat,
  col = col_fun,
  cluster_rows = FALSE,
  cluster_columns = FALSE,
  show_row_names = FALSE,
  show_column_names = FALSE,
  border = "black",
  rect_gp = gpar(
    col = "white",
    lwd = 5 # Increased lwd to create a larger gap
  ),
  heatmap_legend_param = list(
    title = "",
    direction = "horizontal",
    at = c(-2.5, 0, 2.5)
  )
)

# 4. Draw the heatmap
draw(ht, heatmap_legend_side = "bottom")

The RdBu (Red-Blue) palette from RColorBrewer is a "diverging" palette, which is ideal for data with a central neutral value (like 0) and extremes on both ends.

library(RColorBrewer)
library(ComplexHeatmap)
library(circlize)

# Set up the data matrix
set.seed(42)
mat = matrix(rnorm(30, mean = 0, sd = 1.2), nrow = 5, ncol = 6)
mat[2, 1] = -2.5
mat[2, 6] = 2.5
mat[1, 2] = -1.8
mat[4, 4] = 1.8

# Get the RdBu color palette and select the key colors
# The RdBu palette is a red-to-blue gradient, so we reverse it to get blue-to-red
rd_bu_palette = rev(brewer.pal(9, "RdBu"))
three_colors = rd_bu_palette[c(1, 5, 9)] # Select the blue, white, and red shades

# Use colorRamp2 to map the data range to these three colors
col_fun = colorRamp2(c(-2.5, 0, 2.5), three_colors)

# Create the heatmap object
ht = Heatmap(
  mat,
  col = col_fun,
  cluster_rows = FALSE,
  cluster_columns = FALSE,
  show_row_names = FALSE,
  show_column_names = FALSE,
  border = "black",
  rect_gp = gpar(
    col = "white",
    lwd = 5
  ),
  heatmap_legend_param = list(
    title = "",
    direction = "horizontal",
    at = c(-2.5, 0, 2.5)
  )
)

# Draw the heatmap
draw(ht, heatmap_legend_side = "bottom")