Spatial Agglomeration of Firms in a Traditional Business Center. A K-function Analysis

We compare two of the most robust spatial statistics reported in the contemporary literature to analyze locational patterns of firms: the K-planar and the K-network functions. These functions generate values that, under certain conditions, provide insights into the basic spatial structure of the location of the variables under analysis. Both functions are able to incorporate measures of statistical significance on the agglomeration/dispersion/randomness of locational patterns. The K-planar function does it in Euclidian spaces while the K-network function does it in network spaces. Both functions consider space as continuous territories, not as the arbitrary addition of discrete geographical units (i.e. municipalities, states). To our knowledge, there are no published studies in the international literature that explore the differences and similarities of the K-planar and K-network functions in characterizing basic spatial patterns of tertiary firms in the intra-urban space. This paper seeks to fill this gap in the literature.