Ipython Notebook with code for regionalization algorithm in 'Spatial regionalization as optimal data compression', along with metropolitan area census tract networks used in the analysis.
Ethnoracial distributions for each decade are ordered: ['Non-Hispanic White', 'Non-Hispanic Black', 'Asian', 'Hispanic', and 'Other'], and all census metadata are obtained from the Longitudinal Tract Database here
If you use this algorithm or the accompanying data, please cite:
A. Kirkley, Spatial regionalization based on optimal information compression. Communications Physics 5, 249 (2022).