home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ukma!gatech!concert!duke!news.duke.edu!quirky.phy.duke.edu
- From: phusek@quirky.phy.duke.edu (Paul Husek)
- Newsgroups: comp.programming
- Subject: Re:how about Prime Numbers?
- Message-ID: <9201@news.duke.edu>
- Date: 28 Jan 93 16:55:13 GMT
- Sender: news@news.duke.edu
- Lines: 14
- Nntp-Posting-Host: quirky.phy.duke.edu
-
-
- I thought the way they came up with large prime numbers was to perform a
- test on them which had a 1/x probability of revealing a prime number if it
- were prime, where x is of the order of 2. If you perform this test (with
- varying parameters) n times then you have a 1/x^n chance of mistaking a
- nonprime number for a prime. Do it 100 times and your pretty damn certain
- if it passes all these tests its prime.
-
- I think Corman, Rivest, etc in Intro to Algorithms has a
- discussion on this.
- --
- Paul Husek 1-919-383-3535
- Duke University Dept. of Physics phusek@phy.duke.edu
- Durham, N.C. 27706
-