Application of the double kernel density approach to the multivariate analysis of attributeless event point datasets