Jump to content

Key clustering

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by I dream of horses (talk | contribs) at 00:35, 25 June 2019 (top: Cleaning up a randomely generated list, added underlinked tag). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Key or hash function should avoid clustering, the mapping of two or more keys to consecutive slots. Such clustering may cause the lookup cost to skyrocket, even if the load factor is low and collisions are infrequent. The popular multiplicative hash[3] is claimed to have particularly poor clustering behaviour.[9]