Mapping Wide, Sparsely Populated Tables – The Case of Retail Market Shares Estimated by the Huff Model

The Huff Model, a widely-used gravity model in retail location analysis, predicts customer behavior based on proximity, store attractiveness, and competition. In this paper, we address the cartographic challenge of mapping wide, sparsely populated tables resulting from Huff model estimations. Using a case study of Toronto's grocery market, we visualized the market shares of 63 major stores across 572 Census tracts, condensing results into a dominant store and probability per tract. Our solution employs a value-by-alpha technique in QGIS, enabling simultaneous representation of categorical and probabilistic data. This approach enhances legibility while maintaining key information, allowing for clearer insights into retail catchment areas.