Application of DBSCAN algorithm to an example sample. Each panel represents one step in the process, with panel 4 showing the final result of the algorithm. The grey dots represent points that are not assigned to any cluster (noise points in panel 4), while the coloured, big dots depict points assigned to clusters (C1, C2, and C3) and each cluster is represented with a different colour. Coloured points with a black border are core points in a cluster, while those without one represent border points. Black circles show the ϵ environment of some representative points, which are also labeled. First panel: depicts the detection of a cluster within the sample. The ϵ-vicinity of the point labeled as p has more than N_min = 5 sample members within, so p is identified as a core point (represented as a black-edged, coloured point) of the first identified cluster, C1, and the points within its ϵ-neighbourhood are marked as blue, since they are directly reachable from p and, thus, belong to the cluster C1. A subsequent step is to check the ϵ-vicinities of these points belonging to C1, which leads to the discovery of new members of C1, and the classification of some of them as core points. Bottom left panel: a latter epoch in the algorithm application, where a second cluster C2 composed only of core points is identified. Last panel: final results of the algorithm, where a third cluster, C3, with a single core point, is identified. The point q is in the ϵ-vicinities of core points in C1 and C3, so it will be assigned to either cluster depending on the processing order of the points.