The Dendrogram is useful for finding clusters in the plotted data. The dendrogram available in Evince uses Kruskal's algorithm in order to create a minimum spanning tree with clusters. Similar clusters are linked together in the dendrogram. Due to memory limitations, the dendrogram cannot be created from matrices consisting of more than 500 objects.

Dendrogram settings:

Data range: Specifies which vectors should be plotted. The Data range for the dendrogram cannot be changed since this plot type uses all available vectors of the plotted matrix.

Color: Changes the color in the plot. Available options are [FIXED]; the entire plot has the same color, [INDEX]; the indices of the plotted vector are used for coloring, [VALUE]; the values of the plotted vector are used for the coloring, [CATEGORY]; depending on the type of plotted vector either an observation category or a variable category is used for the coloring.

Label: Changes the labels in the plot. Available options are [NONE]; no labels are used, [INDEX]; the indices of plotted vector are used as labels, [IDENTIFIER]; An identifiers of the the plotted vector is used as labels, [VALUE]; the values of the selected vector are used as labels, [CATEGORY]; depending on the type of the plotted vector, either an observation category or a variable category can be used for labeling.

Comment: Available options are [NONE], [CUSTOM] and [CATEGORY]. With [CUSTOM] the user can enter a custom comment using the advanced settings. The custom comment will be shown in the tooltip box that appears when the mouse pointer hovers above a data point. The comment is also shown in any text box created from the table toolbar. When [CATEGORY] is selected the comment will contain either an observation category or a variable category depending on the type of plotted vector.