English

Tight Bounds for Sorting Under Partial Information

Data Structures and Algorithms 2024-08-01 v3

Abstract

Sorting has a natural generalization where the input consists of: (1) a ground set XX of size nn, (2) a partial oracle OPO_P specifying some fixed partial order PP on XX and (3) a linear oracle OLO_L specifying a linear order LL that extends PP. The goal is to recover the linear order LL on XX using the fewest number of linear oracle queries. In this problem, we measure algorithmic complexity through three metrics: oracle queries to OLO_L, oracle queries to OPO_P, and the time spent. Any algorithm requires worst-case log2e(P)\log_2 e(P) linear oracle queries to recover the linear order on XX. Kahn and Saks presented the first algorithm that uses Θ(loge(P))\Theta(\log e(P)) linear oracle queries (using O(n2)O(n^2) partial oracle queries and exponential time). The state-of-the-art for the general problem is by Cardinal, Fiorini, Joret, Jungers and Munro who at STOC'10 manage to separate the linear and partial oracle queries into a preprocessing and query phase. They can preprocess PP using O(n2)O(n^2) partial oracle queries and O(n2.5)O(n^{2.5}) time. Then, given OLO_L, they uncover the linear order on XX in Θ(loge(P))\Theta(\log e(P)) linear oracle queries and O(n+loge(P))O(n + \log e(P)) time -- which is worst-case optimal in the number of linear oracle queries but not in the time spent. For c1c \geq 1, our algorithm can preprocess OPO_P using O(n1+1c)O(n^{1 + \frac{1}{c}}) queries and time. Given OLO_L, we uncover LL using Θ(cloge(P))\Theta(c \log e(P)) queries and time. We show a matching lower bound, as there exist positive constants (α,β)(\alpha, \beta) where for any constant c1c \geq 1, any algorithm that uses at most αn1+1c\alpha \cdot n^{1 + \frac{1}{c}} preprocessing must use worst-case at least βcloge(P)\beta \cdot c \log e(P) linear oracle queries. Thus, we solve the problem of sorting under partial information through an algorithm that is asymptotically tight across all three metrics.

Keywords

Cite

@article{arxiv.2404.08468,
  title  = {Tight Bounds for Sorting Under Partial Information},
  author = {Ivor van der Hoog and Daniel Rutschmann},
  journal= {arXiv preprint arXiv:2404.08468},
  year   = {2024}
}

Comments

To appear at FOCS 2024

R2 v1 2026-06-28T15:52:30.266Z