Conventional methods for cartographic classification are often solely based on underlying attributive values. There are numerous algorithms to determine the resulting classes, such as Jenks Optimal classification, but they still do not account for the spatial patterns that are inherent to spatial data. This can cause a visual disruption of areas that would normally be considered a cluster, thus making the overall message of a map harder to grasp. With a method called “Autocorrelation-Based Regioclassification” TRAUN & LOIDL (2012) introduced an alternative approach that takes spatial properties into account and classifies data values in respect to their statistical and spatial properties. This approach has been implemented in ArcMap as an Add-In using ArcObjects in C#.
TRAUN, C. & LOIDL, M. (2012): Autocorrelation-Based Regioclassification – A self-calibrating classification approach for choropleth maps explicitly considering spatial autocorrelation. International Journal of Geographical Information Science: iFirst 1-17.
Mayrhofer C. (2012): The Implementation of Autocorrelation-Based Regioclassification in ArcMap Using ArcObjects, In: Car, A., Griesebner, G. & Strobl, J. eds. GI Forum. Berlin: Wichman Verlag, 140-150
Categories: GIS Development
Start: Aug, 2011
End: Feb, 2012
Updates in Mar-Jun, 2015