Clustering of Companies in an Inhomogeneous Space
Ilona Berková, Tomáš Mrkvička

Language: en
Last modified: 2018-04-10


Locations of companies are modelled by inhomogeneous point process where inhomogeneity was solved by local scaling. As the scaling factor population was chosen because the hypothesis that the population in unevenly distributed and companies’ choose their locations according to the number of the local population. To study relationships of companies the locally scaled L-function and global envelope test were used. The companies are located in three regions in the Czech Republic (Plzensky, Jihocesky and Vysocina) and contains information from the year 2013. The main aim of this paper is revealing whether positions of companies are spatially dependent or independent. It was found out that the companies tend to cluster even when the population is taken into account.


Local scaling; Inhomogeneous point process; L-function; Global envelope test; Spatial statistics; Spatial clusters; Agglomeration

Full Text: PDF