Synthetic Commodity Sales Dataset for Spatio-Temporal Forecasting

This synthetic dataset contains monthly observations of commodity sales and related exogenous variables across 40 spatially distributed stations over the period 2000–2011 (12 months per year). Each record includes: Geographic coordinates: x-axis and y-axis values defining station location Commodity sales: monthly sales volumes for the target commodity Traffic levels: categorical indicator (high / low) of station traffic Rainfall: monthly precipitation in inches Temperature: monthly average in °C The spatial layout of all 40 stations—based on their x/y coordinates—is illustrated in Figure 1 of the associated manuscript. This dataset was specifically constructed to support the development and evaluation of spatio‐temporal forecasting models that integrate geographic and seasonal drivers of sales.