Summary

In this chapter, we extended our prior insights on temporal and spatial data to tackle analytics problems with spatiotemporal datasets, including a store sales dataset and a Burkina Faso market millet price dataset. We sliced our two datasets by time, calculated spatial statistics, and analyzed network centrality metrics to identify changepoints and periods of volatility in our data. Periods of volatility correspond to network vulnerability to crashes and spikes in prices, sales, and other metrics, as the interconnectedness of the system allows for effects on one part of the network to spill into other parts of the network. We saw this at play during COVID-19 and the Ukraine conflict for Burkina Faso’s markets, as well as the impacts of COVID-19 on the sales data. In Chapter 7, we'll look at sales and goods pricing across both time and geography to see how network science can solve problems in spatiotemporal data. In Chapter 8, we'll consider dynamic networks...