Events occurring in a city, such as car accidents, attacks on the security or activities locations are channelled in the urban network. The goal of this report is to provide a GIS tool able to compute densities and diversities in a network space rather than in the euclidean one. The network space created to support these calculations is a shortest path tree for a given bandwith. On the basis of this shortest path tree, three parameters are calculated: the density and the diversity of activities and the density of edges. The case study of this report is the city of Barcelona. The activities of this city were projected on the network. Indeed, activities like other events in the urban space tend to follow the network. The Human Space Lab (HSL) of the Politecnico di Milano (IT) has applied its Multiple Centrality Assessment to Barcelona's network. Thus, edges become a value of centrality. The network density of the edges presented in this report will take into account these attributes such as the value of population used in other available GIS density tools. The further idea, is to weigh the events (activities or edges) by their distances, as in a standard kernel density estimation. The network approach reveals a spatial distribution of the values sensitive to the favourite direction of the network. The diversity measured along the network make out particular area of the city as residential or commercial zones.