Could you clarify a bit more on what you mean by hash quality and expected comparisons? I took a look at the devel::peek module, specifically the outputs of its dump function and I couldn't find what you were looking for. It sounds like you are attempting to implement a hashing function for an object but I can't be sure. The CPAN documentation for devel::peek (http://search.cpan.org/~ilyaz/Devel-Peek-0.96/Peek.pm) doesn't mention quality either.
Were you referring to another module? Sean Tobin [EMAIL PROTECTED] ________________________________ To: [email protected] Subject: "Quality Percentage" of hashes as reported by Devel::Peek From: [EMAIL PROTECTED] Date: Mon, 3 Dec 2007 12:28:03 -0600 Gurus, I tried directing this query to the module's author, but the e-mail address as listed in the module's documentation is no longer valid--not surprising, after nearly a decade, I guess. Anyhoo, I thought I'd ask if any of you folks might know the answer... In the documentation for this module's Dump subroutine, a hash's "quality" is described as a ratio of "total" comparisons versus "expected" comparisons. There is a formula given for total comparisons, n + n(n-1)/2k, but no formula is given to determine expected comparisons. The question is then, how can a user determine mathematically what an expected quality percentage might be? I know that Dump outputs the percentage, I'd just like to be able to calculate it predictively, instead of fiddling keys(%hash) = X assignments and TIAS-ing for differing values of 'X.' Is there a formula for expected comparisons available? Thanks, Deane Rothenmaier Programmer/Analyst Walgreens Corp. 847-914-5150 A word to the wise ain't necessary, it's the stupid ones who need the advice. -- Bill Cosby _________________________________________________________________ Put your friends on the big screen with Windows Vista® + Windows Live™. http://www.microsoft.com/windows/shop/specialoffers.mspx?ocid=TXT_TAGLM_CPC_MediaCtr_bigscreen_102007 _______________________________________________ ActivePerl mailing list [email protected] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
