Table Of Contents

Next topic

Computational experiments

This Page

clusterPy: Library of spatially constrained clustering algorithms

Analytical regionalization (also known as spatially constrained clustering) is a scientific way to decide how to group a large number of geographic areas or points into a smaller number of regions based on similarities in one or more variables (i.e., income, ethnicity, environmental condition, etc.) that the researcher believes are important for the topic at hand. Conventional conceptions of how areas should be grouped into regions may either not be relevant to the information one is trying to illustrate (i.e., using political regions to map air pollution) or may actually be designed in ways to bias aggregated results. For a literature review on spatially constrained algorithms see [Murtagh1985], [Gordon1996], [Duque_Ramos_Surinach2007].

Working with arbitrary spatial units may lead to aggregation problems such as the modifiable areal unit problem, the small numbers problem, spurious spatial autocorrelation, aggregation bias, aggregation error (in location allocation problems). Analytical regions arise as a way to minimize this type of problems.

Clustering a regular lattice:

import clusterpy
n100 = clusterpy.importArcData("clusterpy/data_examples/n100")
n100.cluster('arisel', ['SAR1'], 6, wType='rook', inits=10, dissolve=1)
n100.results[0].exportArcData('testOutput/demo')
_images/example.png

Clustering California:

import clusterpy
calif = clusterpy.importArcData("clusterpy/data_examples/CA_Polygons")
calif.cluster('arisel', ['PCR2002'], 9, wType='rook', inits=10, dissolve=1)
calif.results[0].exportArcData('testOutput/demo')
_images/example2.png

Special features

Running times and sample data

In this section we report the results of running the algorithms on simulated data. The purpose of these experiments is to provide to the user some information about running times and solution quality of each algorithm. This information will be used as a benchmark in future releases of clusterPy.

Help and documentation

Online documentation

Python shell help system

After importing clusterPy you can use the CPhelp command for more information about a function:

To see the help of a class, in this case ‘’Layer’‘, type:

import clusterpy
clusterpy.CPhelp("Layer")

For a specific function, just type the name of the function

Example 1:

import clusterpy
clusterpy.CPhelp("importArcData")

Example 2:

import clusterpy
clusterpy.CPhelp("new")

Citing ClusterPy library

Please cite ClusterPy when using the software in your work

  • Duque, J.C.; Dev, Boris; Betancourt, A.; Franco, J.L. (2011).ClusterPy: Library of spatially constrained clustering algorithms, Version 0.9.9. RiSE-group (Research in Spatial Economics). EAFIT University. http://www.rise-group.org.*

A BibTeX entry for LaTeX users is:

@Manual{ClusterPy,
title = {ClusterPy: {Library} of spatially constrained clustering algorithms, {Version} 0.9.9.},
author = {Juan C. Duque and Boris Dev and Alejandro Betancourt and Jose L. Franco},
organization = {RiSE-group (Research in Spatial Economics). EAFIT University.},
address = {Colombia},
year = {2011},
url = {http://www.rise-group.org},
}

Indices and tables

Bibliography

[Openshaw_Rao1995](1, 2, 3, 4) Openshaw, S. and Rao, L. (1995). Algorithms for reengineering 1991 census geography. Environment and Planning A, 27(3):425-446.
[Duque_Anselin_Rey2010]Duque, J.C.; Anselin L., Rey S. (2010). The max-p region problem. Working Paper. GeoDa Center for Geospatial Analysis and Computation.
[Kohonen1990]Kohonen, T. (1990). The self-organizing map. Proceedings of the IEEE.
[Kohonen2001]Khonen, T. (2001). Self-Organizing Maps. (Springer, Eds.) (3rd ed.). Berlin: Springer.
[Alstadt_Getis2006]Alstadt, J. and Getis, A. (2006). Using AMOEBA to create a spatial weights matrix and identify spatial clusters. Geographical Analysis, 38(4):327-343
[Bacao_Lobo_Painho2004]Bação, F.; Lobo, V. and Painho, M. (2004). Geo-self-organizing map (Geo-SOM) for building and exploring homogeneous regions. Geographic Information Science, 22-37.
[Schimidt_Rey_Skupin2010]Schmidt C.; Rey S.; Skupin A. (2010). Effects of Irregular Topology in Spherical Self-Organizing Maps. International Regional Science Review, 34(2)215-229.
[Duque_Church2004]Duque, J.C. and Church, R.L. (2004). A new heuristic model for designing analytical regions. In North American Meeting of the Regional Science Association International, Seattle, WA. November.
[Duque_Ramos_Surinach2007]Duque, J, Ramos, R, and Surinach, J (2007). Supervised regionalization methods: A survey. International Regional Science Review, 30:195-220.
[Duque_Alstadt_Velasquez_Franco_Betancourt2010]Duque, J. C., Aldstadt, J., Velasquez, E., Franco, J., & Betancourt, A. (2010). A computationally efficient method for delineating irregularly shaped spatial clusters. Journal of Geographical Systems, 1–18. Springer.
[Duque_Church_Middleton2011]Duque, J.C., Church, R.L., and Middleton, R.S. (2011) The p-regions problem. Geographical Analysis, 43(1):104-126.
[Gordon1996]Gordon, A D (1996). A survey of constrained classification. Computational Statistics & Data Analysis, 21:17-29.
[Murtagh1985]Murtagh, F (1985). A survey of algorithms for contiguity-constrained clustering and related problems. The Computer Journal, 28(1):82-88.
[Ord_Getis1995]Ord J, Getis A (1995) Local spatial autocorrelation statistics: Distributional issues and application. Geographical Analysis 27(4):286-306
[Getis_Ord1992]Getis A, Ord J (1992) The analysis of spatial association by use of distance statistics. Geographical Analysis, 24(3):189-206
[Glover1977]Glover F (1977). Heuristic for integer programming using surrogate constraints. Decision Science 8:156-166.
[Battiti_Tecchiolli1994]Battiti R and Tecchiolli G (1994). The reactive tabu search. ORSA J Comput 6(2): 126-140.
[Kirkpatrick_Gelatt_Vecchi1983]Kirkpatrick S, Gelatt CD, Vecchi MP (1983). Optimization by simulated annealing. Science 220:671-680.