clusterPointsDBScan

Cluster Points DBScan

Clusters the points from the input point layer using Density-based spatial clustering of applications with noise (DBScan) algorithm. The clusters are defined as continuous areas of high density points. Two parameters are used to control the cluster creation - Cluster Tolerance and Minimum Number Points to be used for cluster creation.

Inputs:

A Point feature layer.
Cluster Tolerance - in the units of the spatial reference of the input dataset. Data points will be valid neighbors if their mutual distance is less than or equal to the specified Cluster Tolerance (often referred as epsilon).
Minimum Number Points - The minimum number of data points within the radius of a neighborhood (Cluster Tolerance) for the neighborhood to be considered a cluster.
Output options:

Point: The result will be a point layer. A new field [ClusterID] will be added to the attribute table. The value in this field will identify the cluster to which each point belongs. Points that do not participate in a cluster (Noise) will have ClusterID = -1
Polygon: The result will be a polygon layer. Each polygon will represent a cluster of points and is built as the concave hull of the points belonging to this cluster. New fields will be added to the attribute table:
- [ClusterID] - identifies the cluster represented by the polygon.
- [ET_Count] - the number of points in this cluster.

Outputs:

New point or polygon layer depending on the user selected option (see above).

Notes:

The algorithm starts from a random point and if finds 2 points closer than the Cluster Tolerance to the start point creates an initial cluster polygon. As long as there is another point closer to the initial cluster polygon, it is joined to it. The function is suitable for creating elongated clusters for points representing linear spatial phenomenon.
See also the ClusterPolygonsC function which uses a different algorithm and creates Center based clusters.

Examples:

Source Points

Clustered Points Classified using the values in the ClusterID field

Cluster Polygons overlaid with the source points Minimum Number Points = 4

Cluster Polygons overlaid with the source points Minimum Number Points = 28

Running Programmatically

(Go to TOP)

Parameters

Expression	Explanation
Function Name	ClusterPointsDBScan
<input dataset>	A String representing the input layer. Must be of Point type.
<output dataset>	A String - the full name of the output layer.
<Cluster Tolearance>	A Double representing the Cluster Tolerance( in the units of the spatial reference of the input dataset).
<Min Number Points>	An Integer representing the Minimum number points (see above).
<Output Type>	A String ("Point" or "Polygon") indicating the desired output (see above).

Running the function

ETGWPath used in the table below is the full path to ETGWRun.exe (E.G. "C:\Program Files\ETSpatial Techniques\ETGeo Wizards\ETGWRun.exe")

Language	Syntax
Python	subprocess.call([ETGWPath, "ClusterPointsDBScan", "input dataset", "output dataset", "Cluster Tolearance", "Min Number Points", "Output Type"])
.NET using ETGWRun.exe	StartInfo.FileName = ETGWPath StartInfo.Arguments = "ClusterPointsDBScan" "input dataset" "output dataset" "Cluster Tolearance" "Min Number Points" "Output Type"
.NET using ETGWOutX.dll	ClusterPointsDBScan(input dataset, output dataset, Cluster Tolearance, Min Number Points, Output Type)
ArcPy	arcpy.ClusterPointsDBScan(input dataset, output dataset, "Cluster Tolearance", "Min Number Points", "Output Type")

Notes:

<> - required parameter
{} - optional parameter
See examples for Python , .NET or ArcPy
The argument separator for StartInfo.Arguments is space. If a string might contain a space, you need to double quote it.

(Go to TOP)