Up: Component summary Component


Clusters data using the original implementation of the X-shift algorithm.

Version 1.0
Bundle tools
Categories Clustering
Authors Antti Häkkinen (antti.e.hakkinen@helsinki.fi)
Issue tracker View/Report issues
Requires installer (bash) ; xshift-GlassComponents.jar (jar) ; xshift-Sandbox.jar (jar) ; xshift-VorteX.jar (jar)
Source files component.xml
Usage Example with default values


Name Type Mandatory Description
in CSV Mandatory Data to cluster. Columns correspond to variables and rows observations.


Name Type Description
out CSV A tab-separated CSV file with a single column "label" containing a cluster label (1-based) for each sample.


Name Type Default Description
columns string "" Comma separated list of column headers of the input file to use in the clustering, or "" to select all the columns.
distance string "euclidean" Distance measure for similarity: either "euclidean" or "angular".
kFrom int 150 Neighbor samples to use in density estimation, left bound of the range.
kStep int -5 Neighbor samples to use in density estimation, step size.
kTo int 5 Neighbor samples to use in density estimation, right bound.
n int -1 Neighbor clusters to use in mode finding, or "-1" to compute it from the input dimensions.

Test cases

Test case Parameters IN
case1 properties in out

columns = "x(:,1)","x(:,2)",
distance = euclidean,
kFrom = 150,
kTo = 5,
kStep = -5,
n = -1

Generated 2019-02-07 07:42:33 by Anduril 2.0.0