Expected time analysis of interpolation merge - a simple new merging algorithm

Sumanta Guha, Arunabha Sen

Research output: Contribution to journalArticle

Abstract

Binary merge is the best general-purpose merging algorithm known to date. Binary merge consists of two components: a first component in which an array index is incremented by a number nearly equal to the ratio of the sizes of the two arrays being merged followed by a second component, binary search. In this paper we formulate a simple algorithm called interpolation merge, where the binary search component is replaced with linear search, and analyze its expected behavior over data drawn from a uniform distribution. Our results, both theoretical and experimental, indicate a constant factor (≈0.75) speed-up over straight two-way merge. Further, our analysis of interpolation merge, which uses a mechanism of incremental indexing similar to that in binary merge, will hopefully lead to a better understanding of the latter algorithm. Currently, no significant results are known about the expected behavior of binary merge over data drawn from any standard probability distribution.

Original languageEnglish (US)
Pages (from-to)277-281
Number of pages5
JournalInformation Processing Letters
Volume40
Issue number5
DOIs
StatePublished - Dec 13 1991

Fingerprint

Merging
Interpolation
Interpolate
Binary
Binary search
Linear search
Probability distributions
Uniform distribution
Indexing
Straight
Speedup
Probability Distribution

Keywords

  • Analysis of algorithms
  • merging

ASJC Scopus subject areas

  • Computational Theory and Mathematics

Cite this

Expected time analysis of interpolation merge - a simple new merging algorithm. / Guha, Sumanta; Sen, Arunabha.

In: Information Processing Letters, Vol. 40, No. 5, 13.12.1991, p. 277-281.

Research output: Contribution to journalArticle

@article{d680822754b04bcebf5e3f2245ebfdfe,
title = "Expected time analysis of interpolation merge - a simple new merging algorithm",
abstract = "Binary merge is the best general-purpose merging algorithm known to date. Binary merge consists of two components: a first component in which an array index is incremented by a number nearly equal to the ratio of the sizes of the two arrays being merged followed by a second component, binary search. In this paper we formulate a simple algorithm called interpolation merge, where the binary search component is replaced with linear search, and analyze its expected behavior over data drawn from a uniform distribution. Our results, both theoretical and experimental, indicate a constant factor (≈0.75) speed-up over straight two-way merge. Further, our analysis of interpolation merge, which uses a mechanism of incremental indexing similar to that in binary merge, will hopefully lead to a better understanding of the latter algorithm. Currently, no significant results are known about the expected behavior of binary merge over data drawn from any standard probability distribution.",
keywords = "Analysis of algorithms, merging",
author = "Sumanta Guha and Arunabha Sen",
year = "1991",
month = "12",
day = "13",
doi = "10.1016/0020-0190(91)90123-Y",
language = "English (US)",
volume = "40",
pages = "277--281",
journal = "Information Processing Letters",
issn = "0020-0190",
publisher = "Elsevier",
number = "5",

}

TY - JOUR

T1 - Expected time analysis of interpolation merge - a simple new merging algorithm

AU - Guha, Sumanta

AU - Sen, Arunabha

PY - 1991/12/13

Y1 - 1991/12/13

N2 - Binary merge is the best general-purpose merging algorithm known to date. Binary merge consists of two components: a first component in which an array index is incremented by a number nearly equal to the ratio of the sizes of the two arrays being merged followed by a second component, binary search. In this paper we formulate a simple algorithm called interpolation merge, where the binary search component is replaced with linear search, and analyze its expected behavior over data drawn from a uniform distribution. Our results, both theoretical and experimental, indicate a constant factor (≈0.75) speed-up over straight two-way merge. Further, our analysis of interpolation merge, which uses a mechanism of incremental indexing similar to that in binary merge, will hopefully lead to a better understanding of the latter algorithm. Currently, no significant results are known about the expected behavior of binary merge over data drawn from any standard probability distribution.

AB - Binary merge is the best general-purpose merging algorithm known to date. Binary merge consists of two components: a first component in which an array index is incremented by a number nearly equal to the ratio of the sizes of the two arrays being merged followed by a second component, binary search. In this paper we formulate a simple algorithm called interpolation merge, where the binary search component is replaced with linear search, and analyze its expected behavior over data drawn from a uniform distribution. Our results, both theoretical and experimental, indicate a constant factor (≈0.75) speed-up over straight two-way merge. Further, our analysis of interpolation merge, which uses a mechanism of incremental indexing similar to that in binary merge, will hopefully lead to a better understanding of the latter algorithm. Currently, no significant results are known about the expected behavior of binary merge over data drawn from any standard probability distribution.

KW - Analysis of algorithms

KW - merging

UR - http://www.scopus.com/inward/record.url?scp=0026274955&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026274955&partnerID=8YFLogxK

U2 - 10.1016/0020-0190(91)90123-Y

DO - 10.1016/0020-0190(91)90123-Y

M3 - Article

VL - 40

SP - 277

EP - 281

JO - Information Processing Letters

JF - Information Processing Letters

SN - 0020-0190

IS - 5

ER -