Spatial Pattern Analysis
In this assignment you are trying to see if spatial clustering is present by looking at confidence intervals and probability, using different statistical methods, and then checking the validity of the clustering.
Using one scenario as an example that involves 'calls for service,' I determine if spatial clustering exists using the following statistical methods: Average Nearest Neighbor, Getis-Ord General G, Ripley's K Function and Moran's I. After clustering is determined, I examine further if it is clustering in a specific location.
The Average Nearest Neighbor - This method calculates real data against generated data and measures the difference, which results in a degree of physical clustering.
Getis-Ord General G - This method looks at if similar values are clustered near each other and those values that are similar will show heavy clustering.
Ripley's K Function - Looks at nearest multiple features and distances to determine clustering.
Moran's I - This determines the validity of the clustering with spatial autocorrelation.
Diagram of analyses performed
Maps showing the methods used and validity of results.
Proposed new stations using Nearest Neighbor Call priority clustering using General G
Spatial cluster analysis using K Function Library patron analysis with spatial autocorrelation
Cluster and outlier analysis using Moran's I Cluster and outlier analysis with 900 distance band
Hot Spot analysis of clustering of income levels
A local resort wants to find out if the amount of money that hotel guests spend is related to how far away they live.
A shapefile that would meet the location requirements of the resort, possibly limiting the analyis to a certain geographic area. Customer data with address and amount of money spent at the resort.
Tabular data of customer may need to be formatted and joined with the shapefile. A spatial autocorrelation would see if there was a relation between distance and money spent and Getis-Ord G to check for clustering of values.