REDCAP: Regionalization with Constrained Clustering and Partitioning

Regionalization is to divide a large set of spatial objects into a number of spatially contiguous regions while optimizing an objective function, which is normally a homogeneity (or heterogeneity) measure of the derived regions. This software package includes a family of four hierarchical regionalization methods, which are based on three agglomerative clustering approaches, including the single linkage, average linkage (ALK), complete linkage (CLK), and the WARD's method. The Ward's, Full-Order-ALK, and the Full-Orde-CLK methods all can produce significantly better results than existing approaches. For a specific application context, it is recommended that these three methods are applied and compared.

Related Publication: 
  • Guo, D. (2008). "Regionalization with Dynamically Constrained Agglomerative Clustering and Partitioning (REDCAP)".International Journal of Geographical Information Science. 22(7), pp. 801-823.
  • Guo, D., M. Gahegan, A.M. MacEachren, and B. Zhou. "Multivariate Analysis and Geovisualization with an Integrated Geographic Knowledge Discovery Approach".Cartography and Geographic Information Science,Vol. 32, No. 2, 2005, pp. 113-132.
  • Guo, D., J. Chen, A. M. MacEachren, and K. Liao (2006), "A Visualization System for Spatio-Temporal and Multivariate Patterns (VIS-STAMP)", IEEE Transactions on Visualization and Computer Graphics, 12(6), pp. 1461-1474.

Java is needed to run the following software. You can verify if Java is already installed on your computer at this link: http://java.com/en/download/installed.jsp.

AttachmentSize
redcap.jar7.3 MB
redcapdata.zip (sample data)3.11 MB
redcap_manual.pdf (manual)2.1 MB

Comments

New Version of REDCAP Available

by guod - 07/15/2010

The REDCAP.jar file has been updated with a slightly new version. The old version assumed that each object had no more than 10 spatial neighbors. Now each object can have any number of neighbors. 

Important Tips for Using REDCAP

by guod - 07/07/2010

The construction of contiguity matrix and aggregation of sets of shapes to regions may take a long time if the shapes (e.g., polygons) have very detailed boundaries, which are not necessary for REDCAP since it focuses on topology (i.e., contiguity). It is highly recommended that you simplify (generalize) the shape boundaries of your data with ArcMap (version 9.3.1):

  • Open ArcMap and then open ArcToolBox
  • Data Management Tools --> Generalization --> Simply Polygon
  • Load your shape file and set a simplification tolerance value (which is to control how much you want to simply), which depends on the data coverage and details of your data.
  • Select "Resolve Errors" for handling topological errors.

Then you will have a new shape file. Compile your data set for REDCAP based on this new shape file. This will significantly improve the performance of and your experience with REDCAP.