Skip to content

Dendrogram

Constructs a dendrogram from the hierarchical clustering of data.

Each element becomes a leaf; clusters are merged bottom-up and every merge is drawn as an elbow connector at the height of the cluster dissimilarity.

$ wo 'Head[Dendrogram[{1, 2, 5, 6, 12}]]'
Graphics

Numeric vectors are clustered by Euclidean distance.

$ wo 'Head[Dendrogram[{{1, 1}, {1, 2}, {10, 10}, {10, 11}}]]'
Graphics

Strings are clustered by edit distance.

$ wo 'Head[Dendrogram[{"apple", "aple", "banana", "bananna"}]]'
Graphics

Leaves can be labeled with element -> label rules, a {elements} -> {labels} rule, or an association keyed by label.

$ wo 'Head[Dendrogram[{1 -> "one", 2 -> "two", 10 -> "ten"}]]'
Graphics

$ wo 'Head[Dendrogram[<|"x" -> 1, "y" -> 2, "z" -> 10|>]]'
Graphics

A second argument sets the orientation of the root: Top (default), Bottom, Left or Right.

$ wo 'Head[Dendrogram[{1, 2, 5, 6, 12}, Left]]'
Graphics

Options

  • DistanceFunction — how to measure the distance between two data elements (e.g. EuclideanDistance, ManhattanDistance or a pure function).
  • ClusterDissimilarityFunction — the linkage used to measure the dissimilarity between two clusters: "Single", "Complete", "Average", "WeightedAverage", "Centroid", "Median" or "Ward".
  • ImageSize — overall size in pixels.
$ wo 'Head[Dendrogram[{1, 2, 5, 6, 12}, ClusterDissimilarityFunction -> "Single", DistanceFunction -> ManhattanDistance]]'
Graphics