## Methods for Achieving Fast Query Times in Point Location Data Structures (1997)

Citations: | 20 - 1 self |

### BibTeX

@MISC{Goodrich97methodsfor,

author = {Michael T. Goodrich and Mark Orletsky and Kumar Ramaiyer},

title = {Methods for Achieving Fast Query Times in Point Location Data Structures},

year = {1997}

}

### Years of Citing Articles

### OpenURL

### Abstract

Given a collection S of n line segments in the plane, the planar point location problem is to construct a data structure that can efficiently determine for a given query point p the first segment(s) in S intersected by vertical rays emanating out from p. It is well known that linear-space data structures can be constructed so as to achieve O(log n) query times. But applications, such as those common in geographic information systems, motivate a re-examination of this problem with the goal of improving query times further while also simplifying the methods needed to achieve such query times. In this paper we perform such a re-examination, focusing on the issues that arise in three different classes of point-location query sequences: ffl sequences that are reasonably uniform spatially and temporally (in which case the constant factors in the query times become critical), ffl sequences that are non-uniform spatially or temporally (in which case one desires data structures that adapt to s...

### Citations

387 | Applications of random sampling in computational geometry
- Clarkson, Shor
- 1989
(Show Context)
Citation Context ...-semi-net of order !sn=2 with size O(r), provided that f 0 is non-decreasing. Proof: The proof follows from applications of general proof techniques of Chazelle and Friedman [9] and Clarkson and Shor =-=[12]-=- for range spaces with finite VC-dimension (see also [27]). In our case, f 0 (r) is equal to the number of trapezoids defined by a trapezoidal decomposition of an r-segment subset S ` S; hence f 0 (r)... |

371 |
Self-adjusting binary search trees
- Sleator, Tarjan
- 1985
(Show Context)
Citation Context ...ntaining fast preprocessing bounds and linear space. 3.1 Splay Trees We achieve our results concerning adaptive point location queries by employing the splay tree data structure of Sleator and Tarjan =-=[37]. Splay tr-=-ees are self-adjusting binary trees, and they form a simple and very interesting class of "balanced" binary search trees. We highlight the term balanced, because, unlike other binary search ... |

260 | Epsilon-nets and simplex range queries
- Haussler, Welzl
- 1987
(Show Context)
Citation Context ...is a (1=r)-semi-net 4 of order !s0 if X 0tr N Y (tn=r; S) maxft ! ; 1g = O(f 0 (r)); where the sum ranges over all values of t from 0 to r for which N Y (tn=r; X) is non-zero. (Y is simply an ffl-net =-=[23]-=- if N Y (tn=r; S) = 0 for t ? 1.) Lemma 2.3: Let (S; R) be a segment-trapezoid range space. If Y is a subset of X defined by n mutually-independent indicator random variables, each of which is 1 with ... |

255 |
Optimal Search in Planar Subdivision
- Kirkpatrick
- 1983
(Show Context)
Citation Context ...d to construct a data structure that can efficiently determine for a given query point p the first segment(s) in S intersected by vertical rays emanating out from p. This problem is very well-studied =-=[13, 14, 17, 24, 25, 31, 36]-=-, and there are a number of solutions that asymptotically achieve query times of O(log n) using O(n) space, which is optimal. The query time in each of these solutions is bounded by the number of poin... |

246 | Making data structures persistent
- Driscoll, Sarnak, et al.
- 1989
(Show Context)
Citation Context ...e the space to be O(n). 2.1 A Simple Persistent Method Let S be a collection of n line segments in the plane. We can construct a planar point location data structure for S by applying the persistence =-=[15]-=- paradigm to a left-to-right plane sweep of S using a vertical line, L, as noted by Sarnak and Tarjan [36] (and Cole [13] as well, using a different terminology). Applying the persistence paradigm to ... |

230 |
A Dichromatic Framework for Balanced Trees
- Guibas, Sedgewick
- 1978
(Show Context)
Citation Context ...and they form a simple and very interesting class of "balanced" binary search trees. We highlight the term balanced, because, unlike other binary search trees such as AVL trees [1] or red-bl=-=ack trees [21, 39]-=-, splay trees do not enforce explicit global structural constraints. Sleator and Tarjan define a splay operation at a node v, which involves a series of rotations to take v to the root (but care is ta... |

202 |
A Guided Tour to Chernoff Bounds
- C
- 1989
(Show Context)
Citation Context ...ucture. We have already noted that Steps 1, 2, and 4 can be implemented in O(n log n) time with probability at least 1 \Gamma 1=n c for any constant cs1. Likewise, by a simple Chernoff bound analysis =-=[22]-=-, we can show that Step 3 also runs in time O(n log n) with probability at least 1 \Gamma 1=n c for any constant cs1. Therefore, we have established the following: Theorem 2.4: Given a set S of n non-... |

178 |
An algorithm for the organization of information
- Adelson-Velskii, Landis
- 1962
(Show Context)
Citation Context ...djusting binary trees, and they form a simple and very interesting class of "balanced" binary search trees. We highlight the term balanced, because, unlike other binary search trees such as =-=AVL trees [1]-=- or red-black trees [21, 39], splay trees do not enforce explicit global structural constraints. Sleator and Tarjan define a splay operation at a node v, which involves a series of rotations to take v... |

172 |
Planar point location using persistent search trees
- Sarnak, Tarjan
- 1986
(Show Context)
Citation Context ...d to construct a data structure that can efficiently determine for a given query point p the first segment(s) in S intersected by vertical rays emanating out from p. This problem is very well-studied =-=[13, 14, 17, 24, 25, 31, 36]-=-, and there are a number of solutions that asymptotically achieve query times of O(log n) using O(n) space, which is optimal. The query time in each of these solutions is bounded by the number of poin... |

159 |
Geographic Information Systems and Cartographie Modeling
- Tomlin
- 1990
(Show Context)
Citation Context ...ft of L, or to the right of L. We are interested in re-examining the point location problem, focusing on issues that arise in important application areas, such as geographic information systems (GIS) =-=[7, 6, 18, 35, 40]-=-. Our re-examination of the point-location problem focuses on a number of practical issues that have not been explicitly explored in previous work, including an examination of the exact number of poin... |

158 |
Optimal point location on a monotone subdivision
- Edelsbunner, Guibas, et al.
- 1986
(Show Context)
Citation Context ...d to construct a data structure that can efficiently determine for a given query point p the first segment(s) in S intersected by vertical rays emanating out from p. This problem is very well-studied =-=[13, 14, 17, 24, 25, 31, 36]-=-, and there are a number of solutions that asymptotically achieve query times of O(log n) using O(n) space, which is optimal. The query time in each of these solutions is bounded by the number of poin... |

106 | A deterministic view of random sampling and its use in geometry
- Chazelle, Friedman
- 1990
(Show Context)
Citation Context ... n + rffi log n + nffi) with probability 1 \Gamma 1=2 ffi , for any ffis1, and deterministically in polynomial time. Our proof of this lemma is based upon the general theory of geometric range spaces =-=[2, 9, 20, 26]-=-. In the context of point location data structures a range space is defined by a set S of line segments and R, the set of all combinatorially distinct ways of intersecting segments of S with trapezoid... |

93 |
Multidimensional searching problems
- Dobkin, Lipton
- 1976
(Show Context)
Citation Context |

76 |
An optimal convex hull algorithm in any fixed dimension
- Chazelle
- 1993
(Show Context)
Citation Context ...size, for r = n= log 2 n. This step takes O(n log n) time with probability at least 1 \Gamma 1=n c . 4 This definition of a semi-net is similar to the (1=r)-semi-cutting notion introduced by Chazelle =-=[8], as w-=-ell as proof techniques given in [9, 20, 26]. 2. Apply Lemma 2.1 to form a point-location data structure D for C of size O(r log r) = O(n= log n). 3. For each range R �� defined by a trapezoid �... |

61 |
Location of a Point in a Planar Subdivision and its Applications
- Lee, Preparata
- 1977
(Show Context)
Citation Context |

55 |
A platform for combinatorial and geometric computing
- LEDA
- 1999
(Show Context)
Citation Context ...well as the Edahiro et al. [16] bucket method, were implemented and experiments were conducted to compare the query times of the various methods. All algorithms were implemented in C++ using the LEDA =-=[28, 30]-=- library of data structures and algorithms (v3.2.1), and they were compiled on a SUN SPARC station ELC running SUN OS Release 4.1.1 with the g++ compiler (v4.2). Three different classes of input subdi... |

50 |
Biased search trees
- Bent, Sleator, et al.
- 1985
(Show Context)
Citation Context ...e we use the weights of the nodes to bias the depth of the nodes. As a result, the nodes of larger weights will be at smaller depth. Our persistent tree is therefore now a globally biased search tree =-=[5]-=- rather than a simple complete balanced binary tree, but we can still bound its height to be O(log n). This step takes O(n) time and space, which can be amortized by charging O(1) time to each of the ... |

49 |
Searching and storing similar lists
- Cole
- 1986
(Show Context)
Citation Context |

44 | Fast randomized point location without preprocessing in two- and three-dimensional Delaunay triangulations
- Mücke, Saias, et al.
- 1996
(Show Context)
Citation Context ...gh" technique often used in GIS applications, which can be used, for example, to achieve a sub-linear amount of additional space while still allowing for O(log n)-time queries. Recently Mucke et.=-= al. [29] proposed -=-a randomized point location method which is also based on "walk-through" idea, but their analysis was restricted to Delaunay triangulations. Because of the motivation of our re-examination o... |

43 | Parallel algorithms for higher-dimensional convex hulls
- Amato, Goodrich, et al.
- 1994
(Show Context)
Citation Context ... n + rffi log n + nffi) with probability 1 \Gamma 1=2 ffi , for any ffis1, and deterministically in polynomial time. Our proof of this lemma is based upon the general theory of geometric range spaces =-=[2, 9, 20, 26]-=-. In the context of point location data structures a range space is defined by a set S of line segments and R, the set of all combinatorially distinct ways of intersecting segments of S with trapezoid... |

43 | Dynamic Trees and Dynamic Point Location
- Goodrich, Tamassia
- 1991
(Show Context)
Citation Context ...All of our methods assume that the planar subdivision does not change over time, but environments allowing for dynamic changes to the subdivision over time are well-motivated and wellstudied, as well =-=[3, 4, 10, 11, 19, 32, 33, 34, 38]-=-. Thus, another interesting open problem is whether one can, say, achieve the adaptive query bounds of Theorem 3.5 in such dynamic environments, where one allows insertions and deletions of vertices a... |

43 |
Data Structures and Network Algorithms, volume 44
- Tarjan
- 1983
(Show Context)
Citation Context ...and they form a simple and very interesting class of "balanced" binary search trees. We highlight the term balanced, because, unlike other binary search trees such as AVL trees [1] or red-bl=-=ack trees [21, 39]-=-, splay trees do not enforce explicit global structural constraints. Sleator and Tarjan define a splay operation at a node v, which involves a series of rotations to take v to the root (but care is ta... |

33 |
Dynamic point location in general subdivision
- Baumgarten, Jung, et al.
- 1992
(Show Context)
Citation Context ...All of our methods assume that the planar subdivision does not change over time, but environments allowing for dynamic changes to the subdivision over time are well-motivated and wellstudied, as well =-=[3, 4, 10, 11, 19, 32, 33, 34, 38]-=-. Thus, another interesting open problem is whether one can, say, achieve the adaptive query bounds of Theorem 3.5 in such dynamic environments, where one allows insertions and deletions of vertices a... |

33 |
A New Approach to Planar Point Location
- Preparata
- 1981
(Show Context)
Citation Context |

28 |
Geometric partitioning made easier, even in parallel
- Goodrich
- 1993
(Show Context)
Citation Context ... n + rffi log n + nffi) with probability 1 \Gamma 1=2 ffi , for any ffis1, and deterministically in polynomial time. Our proof of this lemma is based upon the general theory of geometric range spaces =-=[2, 9, 20, 26]-=-. In the context of point location data structures a range space is defined by a set S of line segments and R, the set of all combinatorially distinct ways of intersecting segments of S with trapezoid... |

28 |
Cutting hyperplane arrangements
- Matousek
- 1991
(Show Context)
Citation Context |

26 |
New Results on Dynamic Planar Point Location
- Cheng, Janardan
- 1990
(Show Context)
Citation Context ...All of our methods assume that the planar subdivision does not change over time, but environments allowing for dynamic changes to the subdivision over time are well-motivated and wellstudied, as well =-=[3, 4, 10, 11, 19, 32, 33, 34, 38]-=-. Thus, another interesting open problem is whether one can, say, achieve the adaptive query bounds of Theorem 3.5 in such dynamic environments, where one allows insertions and deletions of vertices a... |

23 | Fully dynamic point location in a monotone subdivision
- Preparata, Tamassia
- 1989
(Show Context)
Citation Context |

21 | A unified approach to dynamic point location, ray shooting, and shortest paths in planar maps
- Chiang, Preparata, et al.
- 1993
(Show Context)
Citation Context |

17 |
A new point-location algorithm and its practical efficiency — Comparison with existing algorithms
- Edahiro, Kokubo, et al.
- 1984
(Show Context)
Citation Context ...on a beautiful hierarchical subdivision method, but the constant factor in the running time is fairly large. Because of the questions of practicality raised by this method, Edahiro, Kokubo, and Asano =-=[16] performed-=- a series of empirical benchmarking tests between the trapezoid and hierarchical subdivision methods, as well as a simple "bucketing" approach, which is not entirely comparison-based and act... |

17 | Epsilon-nets and computational geometry
- MatouSek
- 1993
(Show Context)
Citation Context ...0 is non-decreasing. Proof: The proof follows from applications of general proof techniques of Chazelle and Friedman [9] and Clarkson and Shor [12] for range spaces with finite VC-dimension (see also =-=[27]-=-). In our case, f 0 (r) is equal to the number of trapezoids defined by a trapezoidal decomposition of an r-segment subset S ` S; hence f 0 (r) is O(r). We omit the details of the proof of Lemma 2.2 i... |

9 |
Geographic Information Systems
- Bernhardsen
- 1992
(Show Context)
Citation Context ...ft of L, or to the right of L. We are interested in re-examining the point location problem, focusing on issues that arise in important application areas, such as geographic information systems (GIS) =-=[7, 6, 18, 35, 40]-=-. Our re-examination of the point-location problem focuses on a number of practical issues that have not been explicitly explored in previous work, including an examination of the exact number of poin... |

8 | Biased finger trees and three-dimensional layers of maxima
- Atallah, Goodrich, et al.
- 1994
(Show Context)
Citation Context |

6 |
Dynamic Planar Point Location with Optimal Query Time
- Preparata, Tamassia
- 1990
(Show Context)
Citation Context |

4 |
An Incremental Reconstruction Method for Dynamic Planar Point Location
- Tamassia
- 1991
(Show Context)
Citation Context |

3 |
Computer systems and low level data structures for GIS
- Franklin
- 1991
(Show Context)
Citation Context ...ft of L, or to the right of L. We are interested in re-examining the point location problem, focusing on issues that arise in important application areas, such as geographic information systems (GIS) =-=[7, 6, 18, 35, 40]-=-. Our re-examination of the point-location problem focuses on a number of practical issues that have not been explicitly explored in previous work, including an examination of the exact number of poin... |

1 |
The LEDA User Manual Version R 3.2. Max-Planck-Institut fur Informatik 66123
- Naher, Uhrig
(Show Context)
Citation Context ...well as the Edahiro et al. [16] bucket method, were implemented and experiments were conducted to compare the query times of the various methods. All algorithms were implemented in C++ using the LEDA =-=[28, 30]-=- library of data structures and algorithms (v3.2.1), and they were compiled on a SUN SPARC station ELC running SUN OS Release 4.1.1 with the g++ compiler (v4.2). Three different classes of input subdi... |

1 |
A fully dynamic planar point location technique
- Preparata, Tamassia
- 1987
(Show Context)
Citation Context |