This method is based on maximum distance and can be thought of as the opposite of Single linkage. To calculate the distance between two clusters, each possible combination of two rows (or columns) between the two clusters is compared. The distance between the two clusters is the same as the distance between the two rows (or columns) in the clusters that are most distant.
Assume that there are three clusters called C1, C2 and C3. Clusters C2 and C3 are aggregated to form a new single cluster called C4. The distance between cluster C1 and the new cluster C4 is calculated as:
See also: