I want to use optimized hot spot analysis in Arc GIS 10.3 for grouping of buildings in a particular area, say group of 32 or 64 buildings in each group, because of available splitting ratio of Power splitters in FTTB access networks.
I have two small questions:-
- From where i can get the geo database .gdb of these buildings in the selected area whom i want to cluster. Can i export the database from OSM (Open Street Map) or Google earth or parse the database in Matlab and import in ArcGIS.
- If i use Kmeans in Matlab, it does the clustering BUT i can't control the size of clusters (because of its inherent algorithm of cluster assignment and moving centroids steps) only thing which i can define is the total number of clusters required and the data set on which it should work.
I have seen the available geo database of buildings in Munich, Germany from online resources of ArcGIS. So it is polygonal incident data and doesn't have any analysis field associated with it. How to fix this? one solution can be to assign each building with nearest street node and change polygon data to point data. What you all think!! Any suggestions are welcome.
Anybody expert of using ArcGIS 10.3 desktop/ Pro or know some one!!
Looking forward for your candid advice/ suggestions
http://www.esri.com/pr…/arcgis-capabilities/spatial-analysis
