Spatial data science — Map your way into analytics

In a world awash with data and intellectual forums abuzz with predictive analytics, machine learning, and AI, somehow the focus is not on spatial data science. This is rather strange. Why? Because most of the economic, demographic or climate data is geo-referenced and hence can be analyzed, summarized and visualized as maps. Need convincing. Let me illustrate spatial analytics with two simple maps that demonstrate the east-centre-west divide in the American housing markets.

I sourced housing data from https://www.realtor.com/research/data/. The variables of interest are the median list price and the year-over-year percentage change in listings across the states in September 2022. I downloaded the data as an Excel file and imported it in a Geographic Information Sytems (GIS) software Maptitude. I spatially linked the tabular data to a GIS file depicting the States. The process takes a couple of minutes. Using the thematic mapping feature in Maptitude, I plotted a choropleth map of the median list price. What did I find? Nothing special. Just that the median list price is higher on the eastern and western coasts and lower in the centre.

I also found that the listing activity was slightly higher in September from the same time last year in the western states, and slightly lower in the eastern states. So, what did I conclude: I see western states are faring the housing market downtown better than the eastern states.

The strength of spatial data science lies in its inherent ability to expose spatial trends that one may not readily observe in tabular summaries. Predictive analytics and AI are all the rage, but do not ignore learning from data by plotting and visualizing it. For more on spatial analytics, visit this URL.