Basic Dendrogram — ggplot2

A dendrogram visualizes hierarchical clustering by showing how data points or clusters merge at different distance levels. The tree-like structure reveals relationships and similarity between items, with branch heights indicating the distance at which clusters merge. This visualization is essential for understanding the hierarchical structure in data and identifying natural groupings.
R source (ggplot2)

#' anyplot.ai
#' dendrogram-basic: Basic Dendrogram
#' Library: ggplot2 3.5.1 | R 4.4.1
#' Quality: 88/100 | Created: 2026-06-18

library(ggplot2)
library(dplyr)
library(ragg)
library(gapminder)

# Theme tokens (Imprint palette)
THEME       <- Sys.getenv("ANYPLOT_THEME", "light")
PAGE_BG     <- if (THEME == "light") "#FAF8F1" else "#1A1A17"
ELEVATED_BG <- if (THEME == "light") "#FFFDF6" else "#242420"
INK         <- if (THEME == "light") "#1A1A17" else "#F0EFE8"
INK_SOFT    <- if (THEME == "light") "#4A4A44" else "#B8B7B0"
IMPRINT_PALETTE <- c(
  "#009E73", "#C475FD", "#4467A3", "#BD8233",
  "#AE3030", "#2ABCCD", "#954477", "#99B314"
)

# Data: first 20 European countries (2007) clustered by development indicators
europe_2007 <- gapminder |>
  filter(continent == "Europe", year == 2007) |>
  arrange(country) |>
  slice(1:20) |>
  as.data.frame()

feat_mat <- scale(as.matrix(europe_2007[, c("lifeExp", "gdpPercap", "pop")]))
rownames(feat_mat) <- as.character(europe_2007$country)

# Hierarchical clustering with Ward's linkage
hc       <- hclust(dist(feat_mat), method = "ward.D2")
n_leaves <- nrow(feat_mat)
n_merges <- n_leaves - 1

# Leaf x positions in dendrogram display order
x_pos <- integer(n_leaves)
for (pos in seq_along(hc$order)) {
  x_pos[hc$order[pos]] <- pos
}

# Internal node x (midpoint of children) and y (merge height)
node_x <- numeric(n_merges)
node_y <- hc$height

for (k in seq_len(n_merges)) {
  lft <- hc$merge[k, 1]
  rgt <- hc$merge[k, 2]
  xl  <- if (lft < 0) x_pos[-lft] else node_x[lft]
  xr  <- if (rgt < 0) x_pos[-rgt] else node_x[rgt]
  node_x[k] <- (xl + xr) / 2
}

# Build segment data frame: 3 segments per merge (left vert, horiz, right vert)
n_segs   <- 3L * n_merges
seg_x    <- numeric(n_segs)
seg_xend <- numeric(n_segs)
seg_y    <- numeric(n_segs)
seg_yend <- numeric(n_segs)

for (k in seq_len(n_merges)) {
  lft <- hc$merge[k, 1]
  rgt <- hc$merge[k, 2]
  xl  <- if (lft < 0) x_pos[-lft] else node_x[lft]
  yl  <- if (lft < 0) 0            else node_y[lft]
  xr  <- if (rgt < 0) x_pos[-rgt] else node_x[rgt]
  yr  <- if (rgt < 0) 0            else node_y[rgt]
  i   <- (k - 1L) * 3L + 1L
  seg_x[i]       <- xl; seg_xend[i]       <- xl; seg_y[i]       <- yl;         seg_yend[i]       <- node_y[k]
  seg_x[i + 1L]  <- xl; seg_xend[i + 1L]  <- xr; seg_y[i + 1L]  <- node_y[k]; seg_yend[i + 1L]  <- node_y[k]
  seg_x[i + 2L]  <- xr; seg_xend[i + 2L]  <- xr; seg_y[i + 2L]  <- yr;        seg_yend[i + 2L]  <- node_y[k]
}

seg_df <- data.frame(x = seg_x, xend = seg_xend, y = seg_y, yend = seg_yend)

# Cut tree into 3 clusters; assign Imprint palette colors to leaves
clust_assign <- cutree(hc, k = 3)
leaf_clust   <- clust_assign[hc$order]
max_h        <- max(node_y)
label_y      <- -max_h * 0.02

label_df <- data.frame(
  x       = seq_len(n_leaves),
  y       = label_y,
  label   = hc$labels[hc$order],
  cluster = factor(leaf_clust)
)

plot_title    <- "dendrogram-basic · r · ggplot2 · anyplot.ai"
plot_subtitle <- "European countries clustered by life expectancy, GDP per capita & population · gapminder 2007 · Ward's linkage"

# Plot — horizontal dendrogram via coord_flip; leaves on the right, root on the left
p <- ggplot() +
  geom_segment(
    data = seg_df,
    aes(x = x, xend = xend, y = y, yend = yend),
    color = INK_SOFT, linewidth = 0.7
  ) +
  geom_text(
    data  = label_df,
    aes(x = x, y = y, label = label, color = cluster),
    hjust = 1, size = 3.2
  ) +
  scale_color_manual(
    values = setNames(IMPRINT_PALETTE[1:3], c("1", "2", "3")),
    name   = "Cluster"
  ) +
  guides(color = guide_legend(override.aes = list(label = "■", size = 4))) +
  scale_x_continuous(breaks = NULL, expand = c(0.03, 0.03)) +
  scale_y_continuous(
    limits = c(-max_h * 0.35, max_h * 1.05),
    breaks = pretty(c(0, max_h), n = 5)
  ) +
  coord_flip() +
  labs(
    title    = plot_title,
    subtitle = plot_subtitle,
    x        = NULL,
    y        = "Ward's Distance"
  ) +
  theme_minimal(base_size = 8) +
  theme(
    plot.background    = element_rect(fill = PAGE_BG, color = PAGE_BG),
    panel.background   = element_rect(fill = PAGE_BG, color = NA),
    panel.grid.major   = element_blank(),
    panel.grid.minor   = element_blank(),
    axis.text.x        = element_text(color = INK_SOFT, size = 8),
    axis.text.y        = element_blank(),
    axis.ticks.y       = element_blank(),
    axis.title.x       = element_text(color = INK, size = 10),
    axis.title.y       = element_blank(),
    plot.title         = element_text(color = INK, size = 12, face = "bold"),
    plot.subtitle      = element_text(color = INK_SOFT, size = 8),
    legend.background  = element_rect(fill = ELEVATED_BG, color = INK_SOFT,
                                      linewidth = 0.3),
    legend.text        = element_text(color = INK_SOFT, size = 8),
    legend.title       = element_text(color = INK, size = 10),
    legend.position    = "top",
    plot.margin        = margin(20, 20, 20, 20)
  )

ggsave(
  filename = sprintf("plot-%s.png", THEME),
  plot     = p,
  device   = ragg::agg_png,
  width    = 8,
  height   = 4.5,
  units    = "in",
  dpi      = 400
)
Part of Basic Dendrogram on anyplot.ai.
Basic Dendrogram — ggplot2

R source (ggplot2)

Other implementations
