The first exploratory paper as part of the thesis ‘Hedonic models with machine learning’ will consider spatiotemporal variations in house prices and rents in the capital of Spain, Madrid. Ten years of property adverts in the city have been provided by online real estate company, Idealista. Divergence in house prices can provide a broader insight into geographic inequality in cities. The paper will analyse spatial house price variation over time, in order to identify clusters of high and low growth. In addition, the analysis can identify ripple effects and could be used to inform future housing policy.