The data used in the forthcoming “The modifiable areal unit problem in geospatial least-cost electrification modelling” publication. The work describes how different methods of aggregation of population data effects the results produced by the Open Source Spatial Electrification Tool (OnSSET, https://github.com/OnSSET). In the initial study three countries have been assessed: Benin, Malawi and Namibia. The choice of countries is due to their different national population densities and starting electrification rates. The following repository includes three zipped files, one for each country, containing the 26 input files used in the study. These input files are generated with the QGIS tools published in the OnSSET repository (https://github.com/onsset). This data repository also contains a file describing the naming conventions for the results used and the summary files generated with OnSSET. For more information on how to generate these datasets, please refer to the following GitHub repository https://github.com/babakkhavari/MAUP and the corresponding publication (To Be Added)
Steps to reproduce
Three methods have been used to aggregate the population data: 1. DBSCAN clustering. See the following github repository for the code used: https://github.com/babakkhavari/MAUP 2. Custom clustering. See the following publication for more info: https://www.nature.com/articles/s41597-021-00897-9 3. Resampling of the original population raster The files in Results_used_in_SALib.zip are generated using OnSSET. The folder consists of 6,240 scenarios. See http://www.onsset.org/ for more information on OnSSET and https://github.com/babakkhavari/MAUP for the code used in this analysis.