File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Random number generators produce collisions: Why, how many and more

TitleRandom number generators produce collisions: Why, how many and more
Authors
Keywordscollisions
expected number of collisions
floating point numbers
probability of collision
Random numbers
ties
Issue Date2021
Citation
American Statistician, 2021, v. 75, n. 4, p. 394-402 How to Cite?
AbstractIt seems surprising that when applying widely used random number generators to generate one million random numbers on modern architectures, one obtains, on average, about 116 collisions. This article explains why, how to mathematically compute such a number, why they often cannot be obtained in a straightforward way, how to numerically compute them in a robust way and, among other things, what would need to be changed to bring this number below 1. The probability of at least one collision is also briefly addressed, which, as it turns out, again needs a careful numerical treatment. Overall, the article provides an introduction to the representation of floating-point numbers on a computer and corresponding implications in statistics and simulation. All computations are carried out in R and are reproducible with the texttt included in this article.
Persistent Identifierhttp://hdl.handle.net/10722/325485
ISSN
2023 Impact Factor: 1.8
2023 SCImago Journal Rankings: 0.675
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorHofert, Marius-
dc.date.accessioned2023-02-27T07:33:41Z-
dc.date.available2023-02-27T07:33:41Z-
dc.date.issued2021-
dc.identifier.citationAmerican Statistician, 2021, v. 75, n. 4, p. 394-402-
dc.identifier.issn0003-1305-
dc.identifier.urihttp://hdl.handle.net/10722/325485-
dc.description.abstractIt seems surprising that when applying widely used random number generators to generate one million random numbers on modern architectures, one obtains, on average, about 116 collisions. This article explains why, how to mathematically compute such a number, why they often cannot be obtained in a straightforward way, how to numerically compute them in a robust way and, among other things, what would need to be changed to bring this number below 1. The probability of at least one collision is also briefly addressed, which, as it turns out, again needs a careful numerical treatment. Overall, the article provides an introduction to the representation of floating-point numbers on a computer and corresponding implications in statistics and simulation. All computations are carried out in R and are reproducible with the texttt included in this article.-
dc.languageeng-
dc.relation.ispartofAmerican Statistician-
dc.subjectcollisions-
dc.subjectexpected number of collisions-
dc.subjectfloating point numbers-
dc.subjectprobability of collision-
dc.subjectRandom numbers-
dc.subjectties-
dc.titleRandom number generators produce collisions: Why, how many and more-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1080/00031305.2020.1782261-
dc.identifier.scopuseid_2-s2.0-85088833858-
dc.identifier.volume75-
dc.identifier.issue4-
dc.identifier.spage394-
dc.identifier.epage402-
dc.identifier.eissn1537-2731-
dc.identifier.isiWOS:000553407300001-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats