Evaluate Bin Sizes for Point Aggregation (Spatial Statistics)

Summary

Evaluates multiple bin sizes and recommends an appropriate bin size when aggregating and counting point incidents in a square or hexagonal grid. The tool also allows you to assess various other bin sizes to determine how the resulting counts and patterns would change.

The inputs to the tool are the points that will be aggregated, the bin type (hexagon or square), and aggregation boundary polygons defining where points can occur (such as a city boundary for tree locations in a city). The outputs are the polygon bins along with charts to explore the results of various bin sizes.

Learn more about how Evaluate Bin Sizes for Point Aggregation works

Illustration

Usage

• The tool evaluates each tested bin size by calculating two criteria: one that generally prefers small bin sizes and one that generally prefers large bin sizes. The two criteria are then combined to produce a final evaluation score for the bin size, and the larger the evaluation score, the better the bin size balances both criteria.

  Learn more about evaluation scores

• The Aggregation Boundary parameter is used to define the area in which points will be aggregated (sometimes called a study area) and should represent the area where it is possible for points to occur and be recorded. To estimate an appropriate bin size, it is important to differentiate between whether an area has no points because it happened to have no incidents (such as a section of a city having no robberies in a particular week) or whether it is not possible for points to be observed in the area (such as whale sightings on land).

  Learn more about the aggregation boundary

• The tool creates a group layer to hold the outputs of the tool. The outputs include the polygon bins using the recommended bin size, a table of evaluation scores along with charts, and the aggregation boundary polygon. The recommended bin size is also included as a derived output and is displayed in the messages.

  Learn more about the outputs of the tool

• The tool assumes that there is a single bin size that is appropriate for aggregating the points. However, in many cases, there is no single bin size that will adequately represent the points across the entire aggregation boundary. For example, in a large county that has rural areas with low population density and urban areas with high population density, it may be difficult to aggregate emergency calls across the entire county. Bins small enough to adequately represent the urban areas will be mostly empty in rural areas, while bins large enough for rural areas will condense urban centers into only a few bins. A common sign of this problem is very wide confidence intervals around the recommended bin size, indicating high uncertainty about which bin size to use. A potential solution is to separate the points into different datasets and aggregate them separately using different bin sizes.

Parameters

Dialog

Label	Explanation	Data Type
Input Point Features	The input points that will be aggregated into bins.	Feature Layer
Output Feature Class	The output polygon bins containing the count of points within each bin.	Feature Class
Output Evaluation Scores Table for Charts	The output table that will contain the evaluation scores for all bin sizes. The table will come with charts showing the evaluation scores.	Table
Output Aggregation Boundary Polygons	The aggregation boundary polygons that will be used to create the bins.	Feature Class
Bin Type (Optional)	Specifies the shape of each bin. Square—Points will be aggregated into square bins. Hexagon—Points will be aggregated into hexagonal bins. This is the default.	String
Aggregation Boundary (Optional)	Specifies the boundary or study area in which the points will be aggregated into hexagonal or square bins, and bins will only be included in the output feature class if they intersect the aggregation boundary. The boundary should define the area where it is possible for points to occur. To estimate an appropriate bin size, it is important to differentiate between whether an area has no points because it happened to have no incidents (such as a section of a city having no robberies in a particular week) or whether it is not possible for points to occur in the area (such as whale sightings on land). Using an aggregation boundary that is too large (one that includes many areas where points are not possible or were not recorded) will often result in a bin size that is unrealistically large. Convex hull—The convex hull of the input points will be the boundary for the aggregation. Envelope—The rectangular envelope of the input points will be the boundary for the aggregation. Custom polygons—A custom polygon feature class will be the boundary for the aggregation. Concave hull—The concave hull (alpha shape) of the input points will be the boundary for the aggregation. This is the default.	String
Custom Polygons (Optional)	The custom polygons that will be used as the aggregation boundary.	Feature Layer

Derived Output

Label	Explanation	Data Type
Output Bin Size	The bin size with the largest evaluation score that is used to create the output feature class. The unit is the height of the bin (for squares, it is also the side length).	Double
Output Layer Group	The output group layer that will contain the output features, output table, and output aggregation boundary polygons.	Group Layer

Python

arcpy.stats.EvaluateBinSizes(in_point_features, out_features, out_charts_table, out_agg_bdry, {bin_type}, {aggregation_boundary}, {custom_polygons})

Name	Explanation	Data Type
in_point_features	The input points that will be aggregated into bins.	Feature Layer
out_features	The output polygon bins containing the count of points within each bin.	Feature Class
out_charts_table	The output table that will contain the evaluation scores for all bin sizes. The table will come with charts showing the evaluation scores.	Table
out_agg_bdry	The aggregation boundary polygons that will be used to create the bins.	Feature Class
bin_type (Optional)	Specifies the shape of each bin. SQUARE—Points will be aggregated into square bins. HEXAGON—Points will be aggregated into hexagonal bins. This is the default.	String
aggregation_boundary (Optional)	Specifies the boundary or study area in which the points will be aggregated into hexagonal or square bins, and bins will only be included in the output feature class if they intersect the aggregation boundary. The boundary should define the area where it is possible for points to occur. To estimate an appropriate bin size, it is important to differentiate between whether an area has no points because it happened to have no incidents (such as a section of a city having no robberies in a particular week) or whether it is not possible for points to occur in the area (such as whale sightings on land). Using an aggregation boundary that is too large (one that includes many areas where points are not possible or were not recorded) will often result in a bin size that is unrealistically large. CONVEX_HULL—The convex hull of the input points will be the boundary for the aggregation. ENVELOPE—The rectangular envelope of the input points will be the boundary for the aggregation. CUSTOM—A custom polygon feature class will be the boundary for the aggregation. CONCAVE_HULL—The concave hull (alpha shape) of the input points will be the boundary for the aggregation. This is the default.	String
custom_polygons (Optional)	The custom polygons that will be used as the aggregation boundary.	Feature Layer

Derived Output

Name	Explanation	Data Type
out_bin_size	The bin size with the largest evaluation score that is used to create the output feature class. The unit is the height of the bin (for squares, it is also the side length).	Double
output_layer_group	The output group layer that will contain the output features, output table, and output aggregation boundary polygons.	Group Layer

Code sample

EvaluateBinSizes example 1 (Python window)

The following Python window script demonstrates how to use the EvaluateBinSizes function.

# Aggregate emergency calls within a city.
import arcpy
arcpy.env.workspace = r"c:\mydata\mydata.gdb"
arcpy.stats.EvaluateBinSizes(
    in_point_features="emergency_calls",
    out_features=r"emergency_call_bins",
    out_charts_table=r"out_evaluation_table",
    out_agg_bdry=r"out_agg_boundary",
    bin_type="HEXAGON",
    aggregation_boundary="CUSTOM",
    custom_polygons="city_boundary"
)

EvaluateBinSizes example 2 (stand-alone script)

The following stand-alone script demonstrates how to use the EvaluateBinSizes function.

# Aggregate emergency calls within a city.  

import arcpy 

# Set the current workspace.
arcpy.env.workspace = r"c:\mydata\mydata.gdb" 

# Run tool

arcpy.stats.EvaluateBinSizes(
    in_point_features="emergency_calls",
    out_features=r"emergency_call_bins",
    out_charts_table=r"out_evaluation_table",
    out_agg_bdry=r"out_agg_boundary",
    bin_type="HEXAGON",
    aggregation_boundary="CUSTOM",
    custom_polygons="city_boundary"
)

# Print the messages.
print(arcpy.GetMessages())

Environments

Extent, Output Coordinate System, Parallel Processing Factor, Random number generator

Licensing information

• Basic: Yes
• Standard: Yes
• Advanced: Yes