Multidimensional Digital Searching -- Alternative Data Structures
Peter Kirschenhofery And, Peter Kirschenhofery, Helmut Prodingery
. We analyze the average cost of partial match queries in M--dimensional Digital Search Trees and Patricia Tries. Our results allow a precise comparison of the average behaviour of these data structures with the M--dimensional Tries studied by Flajolet and Puech [1]. It turns out that Patricia is superior to Digital Search Trees, the latter ones being superior to Tries. 1. Introduction and Main Results Random trees under the Bernoulli model occur frequently in connection with data structures in Computer Science. In this paper we concentrate on some aspects of multidimensional search trees. The manipulation of multidimensional data is of vital importance with respect to the construction of database systems and applications, where we mention e.g. graphical algorithms. See Flajolet/Puech [1] for an exhaustive list of references on the subject. One of the important data structures in this…