The Average Nearest Neighbor tool measures the distance between the centroid of each feature and the centroid of its nearest neighbor features, and then calculates the average of all these nearest neighbor distances. At the same time, the Nearest Neighbor Index will be obtained, if the Nearest Neighbor Index is less than 1, then the performance pattern is clustering; if the index is greater than 1, then the performance pattern tends to spread. If the average distance is less than average distance in the assumed Random Distribution, the analyzed feature distribution is considered a clustered feature. If the average distance is greater than average distance in the assumed Random Distribution, the feature is considered a scatter feature. Average Nearest Neighbor can get a specific index of the degree of aggregation of data, through this index, we can compare different data, which data has the greatest degree of aggregation.
The Average Nearest Neighbor method is very sensitive to the area value. Average Nearest Neighbor is best used to compare different elements in a fixed study area. If no area value is set, the default area value is used. The default area value is the area of the Dataset's minimum area Bounds.
Application case
- Assessing competition or territory: quantifying and comparing the spatial distribution of multiple plant or animal species in a fixed study area; comparing the Average Nearest Neighbor distances of different types of businesses in a city.
- Monitor Changes Over Time: Evaluate changes over time in the spatial clustering of one type of enterprise in a fixed study area.
- Comparing the observed distribution with the control distribution: In a timber analysis, given the distribution of all harvestable timber, you would best compare the harvested area pattern with the harvestable area pattern to determine if the cut area is more clustered than expected.
Function entrance
- Spatial Statistical Analysis tab-> Analysis Mode-> Average Nearest Neighbor. （iDesktopX）
- Toolbox, Spatial Statistical Analysis, Analysis Mode, Average Nearest Neighbor. （iDesktopX）
Main parameters
- Source Data: Set the Vector Dataset to be analyzed, which supports three types of Dataset: point, line and surface.
- Study Area: Set the size of the study area in square meters. The area range is ≥ 0. If the study area is 0, the minimum Bounds area of Source Dataset will be automatically calculated as the study area.
- Measure Distance Method: The Measure Distance method uses Euclidean distance and Manhattan distance. Detail Description for Euclidean Distance and Manhattan Distance. Refer to the Basic Vocabulary of Spatial Statistical Analysis .
Explanation of results
Analyst Result is a CAD Dataset and will be displayed in a Map.
Average Nearest NeighborAnalyst Result include: Nearest Neighbor Index, Expected Mean Distance, Average Observation Distance, z-score, P-value five parameters. Nearest Neighbor Index is the ratio of Average Observation Distance to Expected Mean Distance, and if Nearest Neighbor Index is less than 1, the pattern is clustering. If the Nearest Neighbor Index is greater than 1, the displayed pattern tends to spread. As shown in the following figure:
Instance
Case data: Click here to download the case data . After downloading, unzip it for use.
Comparing the aggregation degree of Residence and Villa in a certain area, the Analyst Result is shown as follows:
In this region, the Nearest Neighbor Index of residential areas is 0.2143, and the Nearest Neighbor Index of villas is 0.4708, both of which are less than 1, indicating that the distribution patterns of the two elements are clustered, but the degree of clustering of residential areas is greater.
Related topics
Incremental Spatial Autocorrelation