In many massive dataset applications the data must be stored in space and query efficient data structures on external storage devices. Often the data needs to be changed dynamically. In this chapter we discuss recent advances in the development of provably worst-case efficient external memory dynamic data structures. We also briefly discuss some of the most popular external data structures used in practice.

) Pankaj K. Agarwal Lars Arge y Jeff Erickson z Paolo G. Franciosa x Jeffrey Scott Vitter -- Abstract We show how to preprocess a set S of points in R d to get an external memory data structure that efficiently supports linear-constraint queries. Each query is in the form of a linear constraint a \Delta x b; the data structure must report all the points of S that satisfy the query. Our goal is to minimize the number of disk blocks required to store the data structure and the number of disk accesses (I/Os) required to answer a query. For d = 2, we present the first near-linear size data structures that can answer linear-constraint queries using an optimal number of I/Os. We also present a linear-size data structure that can answer queries efficiently in the worst case. We combine these two approaches to obtain tradeoffs between space and query time. Finally, we show that some of our techniques extend to higher dimensions d. Center for Geometric Computing, Computer...

We study the question of finding a deepest point in an arrangement of regions, and provide a fast algorithm for this problem using random sampling, showing it sufficient to solve this problem when the deepest point is shallow. This implies, among other results, a fast algorithm for solving linear programming with violations approximately. We also use this technique to approximate the disk covering the largest number of red points, while avoiding all the blue points, given two such sets in the plane. Using similar techniques imply that approximate range counting queries have roughly the same time

Two decades ago, Megiddo and Dyer showed that linear programming in 2 and 3 dimensions (and subsequently, any constant number of dimensions) can be solved in linear time. In this paper, we consider linear programming with at most k violations: finding a point inside all but at most k of n given halfspaces. We give a simple algorithm in 2-d that runs in O((n + k ) log n) expected time; this is faster than earlier algorithms by Everett, Robert, and van Kreveld (1993) and Matousek (1994) and is probably nearoptimal for all k n=2. A (theoretical) extension of our algorithm in 3-d runs in near O(n + k ) expected time. Interestingly, the idea is based on concave-chain decompositions (or covers) of the ( k)-level, previously used in proving combinatorial k-level bounds.

We give a data structure that allows arbitrary insertions and deletions on a planar point set P and supports basic queries on the convex hull of P , such as membership and tangent-finding. Updates take O(log 1+" n) amortized time and queries take O(log n) time each, where n is the maximum size of P and " is any fixed positive constant. For some advanced queries such as bridge-finding, both our bounds increase to O(log 3=2 n). The only previous fully dynamic solution was by Overmars and van Leeuwen from 1981 and required O(log 2 n) time per update. 1 Introduction Although the algorithmic study of convex hulls is as old as computational geometry itself, the basic problem of optimally maintaining the planar convex hull under insertions and deletions of points [30, 34] remains unsolved and has been regarded by some as one of the foremost open problems in the area [14, 26]. Besides its natural appeal, such a dynamic data structure has a wide range of applications, as it is often us...

We describe the following data structures. For halfspace range reporting, in 3-space using expected preprocessing time O(n log n), worst-case storage O(n log log n) and worst-case reporting time O(log n + k), where n is the number of data points and k the number of points reported; in d-space, with d even, using worst-case preprocessing time O(n log n), storage O(n) and reporting time O(n 1 1=bd=2c log c n + k), where c is a constant. For ray shooting in a convex polytope in d-space determined by n facets, using deterministic preprocessing time O((n= log n) bd=2c log c n) and storage O((n= log n) bd=2c 2 c log n ) and with query time O(log n). For ray shooting in arbitrary direction among n hyperplanes using preprocessing O(n d = log bd=2c n) and query time O(log n). We also describe a randomized algorithm for constructing the k-level of n planes in 3-space. In the case of planes dual to points in convex position, in which the size of the k-level is O(nk), the a...

Abstract We present new algorithms for approximate range counting,where, for a specified "> 0, we want to count the number of data points in a query range, up to relative error of". We first describe a general framework, adapted from Cohen [10], for this task, and then specialize it to two important instances of range counting: halfspaces in R3 anddisks in the plane. The technique reduces the approximate

In light of recent developments, this paper re-examines the fundamental geometric problem of how to construct the k-level in an arrangement of n lines in the plane. ffl The author's recent dynamic data structure for planar convex hulls improves a decade-old sweep-line algorithm by Edelsbrunner and Welzl, which now runs in O(n log m+m log 1+" n) deterministic time and O(n) space, where m is the output size and " is any positive constant. We discuss simplification of the data structure in this particular application, by viewing the problem kinetically. ffl Har-Peled recently announced a randomized algorithm with an expected running time of O((n + m)ff(n) log n). We observe that a version of an earlier randomized incremental algorithm by Agarwal, de Berg, Matousek, and Schwarzkopf yields almost the same result. ffl The current combinatorial bound by Dey shows that m = O(nk 1=3 ) in the worst case. We give an algorithm that guarantees O(n log n + nk 1=3 ) expected time. 1 Introd...

Given an n-point set, the problems of enumerating the k closest pairs and selecting the k-th smallest distance are revisited. For the enumeration problem, we give simpler randomized and deterministic algorithms with O(n log n + k) running time in any fixed-dimensional Euclidean space. For the selection problem, we give a randomized algorithm with running time O(n log n + n 2=3 k 1=3 log 5=3 n). We also describe output-sensitive results for halfspace range counting that are of use in more general distance selection problems. None of our algorithms requires parametric search. Keywords: distance enumeration, distance selection, closest pairs, range counting, randomized algorithms. 1 Introduction Finding the closest pair of an n-point set has a long history in computational geometry (see [34] for a nice survey). In the plane, the problem can be solved in O(n log n) time using the Delaunay triangulation. In an arbitrary fixed dimension d, the first O(n log n) algorithm, based on di...

We give the first optimal solution to a standard problem in computational geometry: three-dimensional halfspace range reporting. We show that n points in 3-d can be stored in a linear-space data structure so that all k points inside a query halfspace can be reported in O(log n + k) time. The data structure can be built in O(n log n) expected time. The previous methods with optimal query time required superlinear (O(n log log n)) space. We also mention consequences, for example, to higher dimensions and to external-memory data structures. As an aside, we partially answer another open question concerning the crossing number in Matouˇsek’s shallow partition theorem in the 3-d case (a tool used in many known halfspace range reporting methods). 1