And that is exactly what I want to try with you today: Identify distinct neighborhoods based on their price per square meter
As data we will use all transactions of single-family and duplex houses within Vienna in 2016. Thanks to Vienna’s open data policy we can download those from the land charge register at Kaufpreissammlung Liegenschaften - Wien. Unfortunately, no transaction data of apartments are freely available. Of course the data exists, but is hidden behind a massive pay-wall (a few Euros per transaction) and therefore completely out of reach for the budget of this blog.
As a technique we will use Regionalization, which is the process of dividing an area into smaller regions that are homogeneous with respect to a specified metric (in our case the square meter price) aka spatially constrained clustering. For this example we will use the R-package ClustGeo, but check out other packages like HiClimR and SPODT for your own project.
Real estate is all about location! At least this is what I always hear and as far as I have seen it seems to be true. This means prices within a neighborhood should be quite similar and vice versa adjacent properties with similar prices are probably part of the same neighborhood or Grätzel like we call it in Austria.