arXiv subjects

Pawel Gawrychowski

Publications and source records attributed to Pawel Gawrychowski.

At least 19 recordsLinked to original sources

arXiv2026

Near-Linear and Parameterized Approximations for Maximum Cliques in Disk Graphs

A \emph{disk graph} is the intersection graph of (closed) disks in the plane. We consider the classic problem of finding a maximum clique in a disk graph. For general disk graphs, the complexity of this problem is still open, but for unit disk graphs, it is well known to be in P. The currently fastest algorithm runs in time $O(n^{7/3+ o(1)})$, where $n$ denotes the number of disks~\cite{EspenantKM23, keil_et_al:LIPIcs.SoCG.2025.63}. Moreover, for the case of disk graphs with $t$ distinct radii, the problem has also recently been shown to be in XP. More specifically, it is solvable in time $O^*(n^{2t})$~\cite{keil_et_al:LIPIcs.SoCG.2025.63}. In this paper, we present algorithms with improved running times by allowing for approximate solutions and by using randomization: - for unit disk graphs, we give an algorithm that, with constant success probability, computes a $(1-\varepsilon)$-approximate maximum clique in expected time $\tilde{O}(n/\varepsilon^2)$; and - for disk graphs with $t$ distinct radii, we give a parameterized approximation scheme that, with a constant success probability, computes a $(1-\varepsilon)$-approximate maximum clique in expected time $\tilde{O}(f(t)\cdot (1/\varepsilon)^{O(t)} \cdot n)$, for some (exponential) function $f(t)$.

cs.CG↗

arXiv2024

Revisiting Weighted Information Extraction: A Simpler and Faster Algorithm for Ranked Enumeration

Information extraction from textual data, where the query is represented by a finite transducer and the task is to enumerate all results without repetition, and its extension to the weighted case, where each output element has a weight and the output elements are to be enumerated sorted by their weights, are important and well studied problems in database theory. On the one hand, the first framework already covers the well-known case of regular document spanners, while the latter setting covers several practically relevant tasks that cannot be described in the unweighted setting. It is known that in the unweighted case this problem can be solved with linear time preprocessing $O(|D|)$ and output-linear delay $O(|s|)$ in data complexity, where $D$ is the input data and $s$ is the current output element. For the weighted case, Bourhis, Grez, Jachiet, and Riveros [ICDT 2021] recently designed an algorithm with linear time preprocessing, but the delay of $O(|s| \cdot \log|\mathsf{D}|)$ depends on the size of the data. We first show how to leverage the existing results on enumerating shortest paths to obtain a simple alternative algorithm with linear preprocessing and a delay of $O(|s_i| + \min\{ \log i, \log|\mathsf{D}|\})$ for the $i^{\text{th}}$ output element $s_i$ (in data complexity); thus, substantially improving the previous algorithm. Next, we develop a technically involved rounding technique that allows us to devise an algorithm with linear time preprocessing and output-linear delay $O(|s|)$ with high probability. To this end, we combine tools from algebra, high-dimensional geometry, and linear programming.

cs.DS↗

arXiv2024

Enumerating m-Length Walks in Directed Graphs with Constant Delay

In this paper, we provide a novel enumeration algorithm for the set of all walks of a given length within a directed graph. Our algorithm has worst-case constant delay between outputting succinct representations of such walks, after a preprocessing step requiring linear time relative to the size of the graph. We apply these results to the problem of enumerating succinct representations of the strings of a given length from a prefix-closed regular language (languages accepted by a finite automaton which has final states only).

cs.DS↗

arXiv2021

Faster Exponential Algorithm for Permutation Pattern Matching

The Permutation Pattern Matching problem asks, given two permutations $σ$ on $n$ elements and $π$, whether $σ$ admits a subsequence with the same relative order as $π$ (or, in the counting version, how many such subsequences are there). This natural problem was shown by Bose, Buss and Lubiw [IPL 1998] to be NP-complete, and hence we should seek exact exponential time solutions. The asymptotically fastest such solution up to date, by Berendsohn, Kozma and Marx [IPEC 2019], works in $\mathcal{O}(1.6181^n)$ time. We design a simple and faster $\mathcal{O}(1.415^{n})$ time algorithm for both the detection and the counting version. We also prove that this is optimal among a certain natural class of algorithms.

cs.DS↗

arXiv2021

Efficiently Testing Simon's Congruence

Simon's congruence $\sim_k$ is defined as follows: two words are $\sim_k$-equivalent if they have the same set of subsequences of length at most $k$. We propose an algorithm which computes, given two words $s$ and $t$, the largest $k$ for which $s\sim_k t$. Our algorithm runs in linear time $O(|s|+|t|)$ when the input words are over the integer alphabet $\{1,\ldots,|s|+|t|\}$ (or other alphabets which can be sorted in linear time). This approach leads to an optimal algorithm in the case of general alphabets as well. Our results are based on a novel combinatorial approach and a series of efficient data structures.

cs.FL↗

arXiv2017

Submatrix Maximum Queries in Monge Matrices are Equivalent to Predecessor Search

We present an optimal data structure for submatrix maximum queries in n x n Monge matrices. Our result is a two-way reduction showing that the problem is equivalent to the classical predecessor problem in a universe of polynomial size. This gives a data structure of O(n) space that answers submatrix maximum queries in O(loglogn) time. It also gives a matching lower bound, showing that O(loglogn) query-time is optimal for any data structure of size O(n polylog(n)). Our result concludes a line of improvements that started in SODA'12 with O(log^2 n) query-time and continued in ICALP'14 with O(log n) query-time. Finally, we show that partial Monge matrices can be handled in the same bounds as full Monge matrices. In both previous results, partial Monge matrices incurred additional inverse-Ackerman factors.

cs.DS↗

arXiv2017

Improved Submatrix Maximum Queries in Monge Matrices

We present efficient data structures for submatrix maximum queries in Monge matrices and Monge partial matrices. For $n\times n$ Monge matrices, we give a data structure that requires O(n) space and answers submatrix maximum queries in $O(\log n)$ time. The best previous data structure [Kaplan et al., SODA`12] required $O(n \log n)$ space and $O(\log^2 n)$ query time. We also give an alternative data structure with constant query-time and $ O(n^{1+\varepsilon})$ construction time and space for any fixed $\varepsilon<1$. For $n\times n$ {\em partial} Monge matrices we obtain a data structure with O(n) space and $O(\log n \cdot α(n))$ query time. The data structure of Kaplan et al. required $O(n \log n \cdot α(n))$ space and $O(\log^2 n)$ query time. Our improvements are enabled by a technique for exploiting the structure of the upper envelope of Monge matrices to efficiently report column maxima in skewed rectangular Monge matrices. We hope this technique can be useful in obtaining faster search algorithms in Monge partial matrices. In addition, we give a linear upper bound on the number of breakpoints in the upper envelope of a Monge partial matrix. This shows that the inverse Ackermann $α(n)$ term in the analysis of the data structure of Kaplan et. al is superfluous.

cs.DS↗

arXiv2017

Optimal Query Time for Encoding Range Majority

We revisit the range $τ$-majority problem, which asks us to preprocess an array $A[1..n]$ for a fixed value of $τ\in (0,1/2]$, such that for any query range $[i,j]$ we can return a position in $A$ of each distinct $τ$-majority element. A $τ$-majority element is one that has relative frequency at least $τ$ in the range $[i,j]$: i.e., frequency at least $τ(j-i+1)$. Belazzougui et al. [WADS 2013] presented a data structure that can answer such queries in $O(1/τ)$ time, which is optimal, but the space can be as much as $Θ(n \lg n)$ bits. Recently, Navarro and Thankachan [Algorithmica 2016] showed that this problem could be solved using an $O(n \lg (1/τ))$ bit encoding, which is optimal in terms of space, but has suboptimal query time. In this paper, we close this gap and present a data structure that occupies $O(n \lg (1/τ))$ bits of space, and has $O(1/τ)$ query time. We also show that this space bound is optimal, even for the much weaker query in which we must decide whether the query range contains at least one $τ$-majority element.

cs.DS↗

arXiv2017

Near-Optimal Compression for the Planar Graph Metric

The Planar Graph Metric Compression Problem is to compactly encode the distances among $k$ nodes in a planar graph of size $n$. Two naïve solutions are to store the graph using $O(n)$ bits, or to explicitly store the distance matrix with $O(k^2 \log{n})$ bits. The only lower bounds are from the seminal work of Gavoille, Peleg, Prennes, and Raz [SODA'01], who rule out compressions into a polynomially smaller number of bits, for {\em weighted} planar graphs, but leave a large gap for unweighted planar graphs. For example, when $k=\sqrt{n}$, the upper bound is $O(n)$ and their constructions imply an $Ω(n^{3/4})$ lower bound. This gap is directly related to other major open questions in labelling schemes, dynamic algorithms, and compact routing. Our main result is a new compression of the planar graph metric into $\tilde{O}(\min (k^2 , \sqrt{k\cdot n}))$ bits, which is optimal up to log factors. Our data structure breaks an $Ω(k^2)$ lower bound of Krauthgamer, Nguyen, and Zondiner [SICOMP'14] for compression using minors, and the lower bound of Gavoille et al. for compression of weighted planar graphs. This is an unexpected and decisive proof that weights can make planar graphs inherently more complex. Moreover, we design a new {\em Subset Distance Oracle} for planar graphs with $\tilde O(\sqrt{k\cdot n})$ space, and $\tilde O(n^{3/4})$ query time. Our work carries strong messages to related fields. In particular, the famous $O(n^{1/2})$ vs. $Ω(n^{1/3})$ gap for distance labelling schemes in planar graphs {\em cannot} be resolved with the current lower bound techniques.

cs.DS↗

arXiv2015

Testing k-binomial equivalence

Two words $w_1$ and $w_2$ are said to be $k$-binomial equivalent if every non-empty word $x$ of length at most $k$ over the alphabet of $w_1$ and $w_2$ appears as a scattered factor of $w_1$ exactly as many times as it appears as a scattered factor of $w_2$. We give two different polynomial-time algorithms testing the $k$-binomial equivalence of two words. The first one is deterministic (but the degree of the corresponding polynomial is too high) and the second one is randomised (it is more direct and more efficient). These are the first known algorithms for the problem which run in polynomial time.

cs.FL↗

arXiv2015

Longest Common Extensions in Trees

The longest common extension (LCE) of two indices in a string is the length of the longest identical substrings starting at these two indices. The LCE problem asks to preprocess a string into a compact data structure that supports fast LCE queries. In this paper we generalize the LCE problem to trees and suggest a few applications of LCE in trees to tries and XML databases. Given a labeled and rooted tree $T$ of size $n$, the goal is to preprocess $T$ into a compact data structure that support the following LCE queries between subpaths and subtrees in $T$. Let $v_1$, $v_2$, $w_1$, and $w_2$ be nodes of $T$ such that $w_1$ and $w_2$ are descendants of $v_1$ and $v_2$ respectively. \begin{itemize} \item $\LCEPP(v_1, w_1, v_2, w_2)$: (path-path $\LCE$) return the longest common prefix of the paths $v_1 \leadsto w_1$ and $v_2 \leadsto w_2$. \item $\LCEPT(v_1, w_1, v_2)$: (path-tree $\LCE$) return maximal path-path LCE of the path $v_1 \leadsto w_1$ and any path from $v_2$ to a descendant leaf. \item $\LCETT(v_1, v_2)$: (tree-tree $\LCE$) return a maximal path-path LCE of any pair of paths from $v_1$ and $v_2$ to descendant leaves. \end{itemize} We present the first non-trivial bounds for supporting these queries. For $\LCEPP$ queries, we present a linear-space solution with $O(\log^{*} n)$ query time. For $\LCEPT$ queries, we present a linear-space solution with $O((\log\log n)^{2})$ query time, and complement this with a lower bound showing that any path-tree LCE structure of size $O(n \polylog(n))$ must necessarily use $Ω(\log\log n)$ time to answer queries. For $\LCETT$ queries, we present a time-space trade-off, that given any parameter $τ$, $1 \leq τ\leq n$, leads to an $O(nτ)$ space and $O(n/τ)$ query-time solution. This is complemented with a reduction to the the set intersection problem implying that a fast linear space solution is not likely to exist.

cs.DS↗

arXiv2015

Optimal Encodings for Range Top-k, Selection, and Min-Max

We consider encoding problems for range queries on arrays. In these problems the goal is to store a structure capable of recovering the answer to all queries that occupies the information theoretic minimum space possible, to within lower order terms. As input, we are given an array $A[1..n]$, and a fixed parameter $k \in [1,n]$. A range top-$k$ query on an arbitrary range $[i,j] \subseteq [1,n]$ asks us to return the ordered set of indices $\{\ell_1, ..., \ell_{k}\}$ such that $A[\ell_m]$ is the $m$-th largest element in $A[i..j]$, for $1 \le m \le k$. A range selection query for an arbitrary range $[i,j] \subseteq [1,n]$ and query parameter $k' \in [1,k]$ asks us to return the index of the $k'$-th largest element in $A[i..j]$. We completely resolve the space complexity of both of these heavily studied problems---to within lower order terms---for all $k = o(n)$. Previously, the constant factor in the space complexity was known only for $k=1$. We also resolve the space complexity of another problem, that we call range min-max, in which the goal is to return the indices of both the minimum and maximum elements in a range.

cs.DS↗

arXiv2015

Encodings of Range Maximum-Sum Segment Queries and Applications

Given an array A containing arbitrary (positive and negative) numbers, we consider the problem of supporting range maximum-sum segment queries on A: i.e., given an arbitrary range [i,j], return the subrange [i' ,j' ] \subseteq [i,j] such that the sum of the numbers in A[i'..j'] is maximized. Chen and Chao [Disc. App. Math. 2007] presented a data structure for this problem that occupies Θ(n) words, can be constructed in Θ(n) time, and supports queries in Θ(1) time. Our first result is that if only the indices [i',j'] are desired (rather than the maximum sum achieved in that subrange), then it is possible to reduce the space to Θ(n) bits, regardless the numbers stored in A, while retaining the same construction and query time. We also improve the best known space lower bound for any data structure that supports range maximum-sum segment queries from n bits to 1.89113n - Θ(lg n) bits, for sufficiently large values of n. Finally, we provide a new application of this data structure which simplifies a previously known linear time algorithm for finding k-covers: i.e., given an array A of n numbers and a number k, find k disjoint subranges [i_1 ,j_1 ],...,[i_k ,j_k ], such that the total sum of all the numbers in the subranges is maximized.

cs.DS↗

arXiv2015

Strong inapproximability of the shortest reset word

The Černý conjecture states that every $n$-state synchronizing automaton has a reset word of length at most $(n-1)^2$. We study the hardness of finding short reset words. It is known that the exact version of the problem, i.e., finding the shortest reset word, is NP-hard and coNP-hard, and complete for the DP class, and that approximating the length of the shortest reset word within a factor of $O(\log n)$ is NP-hard [Gerbush and Heeringa, CIAA'10], even for the binary alphabet [Berlinkov, DLT'13]. We significantly improve on these results by showing that, for every $ε>0$, it is NP-hard to approximate the length of the shortest reset word within a factor of $n^{1-ε}$. This is essentially tight since a simple $O(n)$-approximation algorithm exists.

cs.FL↗

arXiv2015

Lock-in Problem for Parallel Rotor-router Walks

The rotor-router model, also called the Propp machine, was introduced as a deterministic alternative to the random walk. In this model, a group of identical tokens are initially placed at nodes of the graph. Each node maintains a cyclic ordering of the outgoing arcs, and during consecutive turns the tokens are propagated along arcs chosen according to this ordering in round-robin fashion. The behavior of the model is fully deterministic. Yanovski et al.(2003) proved that a single rotor-router walk on any graph with m edges and diameter $D$ stabilizes to a traversal of an Eulerian circuit on the set of all 2m directed arcs on the edge set of the graph, and that such periodic behaviour of the system is achieved after an initial transient phase of at most 2mD steps. The case of multiple parallel rotor-routers was studied experimentally, leading Yanovski et al. to the conjecture that a system of $k \textgreater{} 1$ parallel walks also stabilizes with a period of length at most $2m$ steps. In this work we disprove this conjecture, showing that the period of parallel rotor-router walks can in fact, be superpolynomial in the size of graph. On the positive side, we provide a characterization of the periodic behavior of parallel router walks, in terms of a structural property of stable states called a subcycle decomposition. This property provides us the tools to efficiently detect whether a given system configuration corresponds to the transient or to the limit behavior of the system. Moreover, we provide polynomial upper bounds of $O(m^4 D^2 + mD \log k)$ and $O(m^5 k^2)$ on the number of steps it takes for the system to stabilize. Thus, we are able to predict any future behavior of the system using an algorithm that takes polynomial time and space. In addition, we show that there exists a separation between the stabilization time of the single-walk and multiple-walk rotor-router systems, and that for some graphs the latter can be asymptotically larger even for the case of $k = 2$ walks.

cs.DM↗

arXiv2014

Weighted ancestors in suffix trees

The classical, ubiquitous, predecessor problem is to construct a data structure for a set of integers that supports fast predecessor queries. Its generalization to weighted trees, a.k.a. the weighted ancestor problem, has been extensively explored and successfully reduced to the predecessor problem. It is known that any solution for both problems with an input set from a polynomially bounded universe that preprocesses a weighted tree in O(n polylog(n)) space requires Ω(loglogn) query time. Perhaps the most important and frequent application of the weighted ancestors problem is for suffix trees. It has been a long-standing open question whether the weighted ancestors problem has better bounds for suffix trees. We answer this question positively: we show that a suffix tree built for a text w[1..n] can be preprocessed using O(n) extra space, so that queries can be answered in O(1) time. Thus we improve the running times of several applications. Our improvement is based on a number of data structure tools and a periodicity-based insight into the combinatorial structure of a suffix tree.

cs.DS↗

arXiv2014

Euclidean TSP with few inner points in linear space

Given a set of $n$ points in the Euclidean plane, such that just $k$ points are strictly inside the convex hull of the whole set, we want to find the shortest tour visiting every point. The fastest known algorithm for the version when $k$ is significantly smaller than $n$, i.e., when there are just few inner points, works in $O(k^{11\sqrt{k}} k^{1.5} n^{3})$ time [Knauer and Spillner, WG 2006], but also requires space of order $k^{c\sqrt{k}}n^{2}$. The best linear space algorithm takes $O(k! k n)$ time [Deineko, Hoffmann, Okamoto, Woeginer, Oper. Res. Lett. 34(1), 106-110]. We construct a linear space $O(nk^2+k^{O(\sqrt{k})})$ time algorithm. The new insight is extending the known divide-and-conquer method based on planar separators with a matching-based argument to shrink the instance in every recursive call. This argument also shows that the problem admits a quadratic bikernel.

cs.DS↗

arXiv2014

Order-preserving pattern matching with k mismatches

We study a generalization of the recently introduced order-preserving pattern matching, where instead of looking for an exact copy of the pattern, we only require that the relative order between the elements is the same. In our variant, we additionally allow up to k mismatches between the pattern and the text, and the goal is to construct an efficient algorithm for small values of k. For a pattern of length m and a text of length n, our algorithm detects an order-preserving occurrence with up to k mismatches in O(n(loglogm + kloglogk)) time.

cs.DS↗