CL HC - Classification - Hierarchical Clustering

PURPOSE

Performs hierarchical clustering according to one of several clustering criteria, on factors produced by CORAN.

SEE ALSO

CL CLA	[Classification - Clustering]
CL HD	[Classification - Hierarchical clustering, calculate classes]
CL HE	[Classification - Hierarchical clustering, create docfiles]
CL KM	[Classification - K Means clustering ||]

USAGE

.OPERATION: CL HC

.CORAN/PCA FILE (e.g. CORAN_01_IMC~) FILE: coran_t_IMC
[Enter name of the raw image data sequential file (_SEQ), image factor coordinate file (_IMC), or pixel factor coordinate file (_PIX) file containing your data. These files were created by 'CA S.

.FACTOR NUMBERS: 1,3,4,6
[Enter the factors to be included in the hierarchical clustering algorithm.]

.FACTOR WEIGHT: 1.5
[Enter a weight for each factor selected. If the answer zero is given at any point, all weights from the current factor onwards are set to one. This question is asked as many times as the number of factors specified, or is terminated by the answer zero.]

.CLUSTERING CRITERION (1-5): 2
[Enter the number indicating clustering criterion to be used. Possible choices are:
Option 1: single linkage
Option 2: complete linkage
Option 3: average linkage
Option 4: centroid method
Option 5: Ward's method]

.DO YOU WANT DENDROGRAM POSTSCRIPT PLOT? (Y/T/N): T
[Y - The whole dendrogram will be plotted
T - A truncated dendrogram will be plotted (useful for large number of objects)
N - No dendrogram will be plotted.]

If answer is Y or T then:

.DENDROGRAM FILE: MAP001
[Enter name of file where the dendrogram will be stored.]

If truncated dendrogram was requested:

.ENTER PLOT CUTOFF: 12.3
[Enter the cutoff number at which the dendrogram will be cut. Only top portion will be produced. This number must be taken from the RESULTS file from the previous run and must be in the range of the scaling of the dendrogram tree. This range is written in the line under the tree in the RESULTS file.]

.DO YOU WANT DENDROGRAM DOC FILE? (Y/N): Y
[ Y - the document file will be produced containing information about classification tree. Using this information, one can retrieve the clustering behavior for any number of groups.]

If previous answer was Y, then:

.DOCUMENT FILE: MAP001
[Enter the document file name where the dendrogram information will be stored.]

NOTES

1. To get the classification on a given "cutting" level use command 'CL HD'.

2. To get the selection doc files corresponding to the given "cutting" use command 'CL HE'.

3. To calculate averages for the classes use command 'AS DC' and selection doc. files.

4. Every vertical line at the bottom of the drawing represents an image that you input into 'CL HC'. Each vertical line is an average of the images, or vertical lines, below it. The threshold is a scaled value from 0 to 1 that informs 'CL HC' how far "up" the dendrogram you wish it to look. A threshold set at the bottem would result in the number of classes being equal to the number of input images. A median threshold value of 0.5 results in fewer classes. And a top level threshold gives a single class containing of all the inputs.

5. Implemented by P. Penczek.

SUBROUTINES

HCLS, DIST_P, CHAVA, DENDRO, DENDRO2, ARBRE, HCLP

CALLER

UTIL1