A dendrogram is an upside-down tree that displays the hierarchy formed by clustering similar observations. Hierarchical clustering begins with one cluster per observation, repeatedly merges clusters, and ends with one overall cluster. The diagram makes those nested relationships visible without requiring an outcome variable. section 3.9
NBA draft picks and career success
A dendrogram can group first-round NBA draft selections according to pick number and average career win shares. In the resulting plot, the leaves identify selection numbers, while branch height indicates Euclidean distance. Cutting the tree into two groups with rect.hclust(..., k = 2) highlights a broad segmentation of picks based on those two measures. The analysis is useful for discovering structure, but hierarchical clustering can become slow on large data sets. section 3.9
