Fully Retroactive Approximate Range and Nearest Neighbor Searching

Abstract

We describe fully retroactive dynamic data structures for approximate range reporting and approximate nearest neighbor reporting. We show how to maintain, for any positive constant d, a set of n points in R d indexed by time such that we can perform insertions or deletions at any point in the timeline in O(log n) amortized time. We support, for any small constant > 0, (1 + )-approximate range reporting queries at any point in the timeline in O(log n + k) time, where k is the output size. We also show how to answer (1 + )-approximate nearest neighbor queries for any point in the past or present in O(log n) time. arXiv:1109.0312v1 [cs.CG] 1 Sep 2011 2. Represent each d-dimensional point as a line segment in d + 1 dimensions.

Fully Retroactive Priority Queues using Persistent Binary Search Trees

“…The notion of retroactive data structure helps to solve problems, such as dynamic shortest path problem (Sunita and Garg, 2018) and geometric problems, such as cloning Voronoi diagrams (Dickerson et al, 2010) and nearest neighbor search (Goodrich and Simons, 2011 time per operation. The data structure also supports the operation of determining the time at which an element was deleted from the data structure in O(lg 2 m).…”

Windows into Geometric Events: Data Structures for Time-Windowed Querying of Temporal Point Sets

“…We will denote this linear ordering by the symbol ≤ Z . Considering points in their Zorder is a dimension reduction technique that is often used for proximity based data structures [7,19].…”

“…Then for each of the inner cells I ∈ I u ∪ I v we perform a 2-dimensional range reporting query with the rectangle [p i , p j ] × [z 0 , z 1 ], where z 0 and z 1 are the first and last point in I in Z-order, and record the union of the points reported in O(log w + k) time where k is output size. By a simple packing argument, the total number of inner cells is bounded by a function of the constants ε and d [19]. Thus, the total time for the query is O(log w + k).…”