| hclust | Hierarchical clustering |
| -sl | single link clustering (MERGE) |
| -al | average link clustering (MERGE) |
| -cl | complete link clustering (MERGE) |
| -ce | centroid (group average) (MERGE) |
| -me | median clustering (MERGE) |
| -mv | min variance clustering (MERGE) |
| -2m | 2-means clustering (SPLIT) |
| -ms | Macnaughton-Smith clustering (SPLIT) |
| -d <datain> | the name of the input data frame |
| -dout <dataout> | the name of the output data frame |
| -cout <cldata> | data frame for clusters |
| -clvis <clvis> | data frame for tree structures in visualization |
| [-cut <layer>] | choose in which layer to cut the tree |
| [-verbose ] | request additional information about the clustering process |
This operation divides the data set, <data>, into clusters by hierarchical clustering algorithms. The centroids of the clusters are stored into the frame <dataout> and classification of data records into <cldata>. The flag -clvis is used for visualization. The frame <clvis> contains the tree structure of the network. Please note, that the visualization tree structure does not visualize the locations of the actual clusters of data!
Example (ex5.14): This example shows the basic use of the command.
NDA> load boston.dat
NDA> select cluflds -f boston.indus boston.crim boston.zn
boston.rate
NDA> hclust -d cluflds -dout hcludat -cout cld -clvis hhier -sl
NDA> mkgrp win1 -d hcludat
NDA> setgdat win1 -d hcludat
NDA> topo win1 -c hhier -on
...
