Name

sxhac_clustering - Hierarchical agglomerative clustering

Usage

Usage in command lines

sxhac_clustering.py stack dendoname <maskfile> --link=kind_of_linkage --dist=kind_of_distance --dissimilar

Usage in python programming

HAC_clustering(stack, dendoname, maskname, link, dist)

Example:

sxhac_clustering.py stable_ave.hdf dendo mask2d_13.hdf --link='complete' --dist='SqEuc' --dissimilar

Input

stack: the input stack of images
maskname: optional mask file to be used
dendoname: the seed name to export the dendogram
link: Kind of linkage, single, complete or average (default single)
dist: Kind of distance, SqEuc or CCC (default SqEuc)
dissimilar: Change the distance value to the negative value, meaning not agglomerate the similar object but the dissimilar

Output

dendoname.txt: depends on the seed name chose (seed name + .txt), text file which contains the threshold according the number of groups
dendoname.dendo: depends on the seed name chose (seed name + .dendo), file which contains the dendogram

Description

This program is a wrap from a part of python-cluster to sparx. This function use different kind of distance:

SqEuc - squared Euclidean distance

CCC - cross correlation coefficient

The program returns two files. The first dendoname.txt contains the number of groups according to the cutting threshold value of the dendogram. This file is to help the user to know what number of groups is saved in the second file dendoname.dendo which contains the dendogram structure. To get the averages for a number of groups select in dendoname.txt use the function sxhac_averages.py.

Reference

Author / Maintainer

Julien Bert

Keywords

category 1: APPLICATIONS

Files

applications.py

statistics.py (python-cluster)

Maturity

beta: works for author, often works for others.

Bugs

None. It is perfect.