A conditional nearest-neighbor spatial-association measure for the analysis of conditional locational interdependence

The objective of this paper is to examine descriptively the locational phenomena that the distribution of one type of establishment is conditioned by the location of another type, such as barbers near hotels or newspaper stands around stations. First, a model is formulated in which points of one type are probabilistically distributed in relation to another type of points whose location is fixed. Second, in this context, a conditional nearest-neighbor spatial-association measure is proposed. Third, as the calculation of this measure is complex, a computational method is shown. Fourth, by the use of this method, eight types of establishments in Toshima-ward in Tokyo are examined for spatial association with railway stations.