home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math
- Path: sparky!uunet!europa.asd.contel.com!emory!swrinde!elroy.jpl.nasa.gov!decwrl!deccrl!news.crl.dec.com!dbased.nuo.dec.com!nntpd.lkg.dec.com!math.zk3.dec.com!edp
- From: edp@math.zk3.dec.com (Eric Postpischil)
- Subject: Re: Pi in beatiful form --lost
- Message-ID: <1992Dec16.144508.19371@nntpd.lkg.dec.com>
- Sender: usenet@nntpd.lkg.dec.com (USENET News System)
- Reply-To: edp@math.zk3.dec.com (Eric Postpischil)
- Organization: Digital Equipment Corporation
- References: <BzBt9q.2HM@unccsun.uncc.edu> <1gmvieINN7al@aludra.usc.edu>
- Date: Wed, 16 Dec 1992 14:45:08 GMT
- Lines: 24
-
- In article <1gmvieINN7al@aludra.usc.edu>, rmurphy@aludra.usc.edu (Bob
- Murphy) writes:
-
- >(pi - 3)/4 = 1/(2*3*4) + 1/(4*5*6) + 1/(6*7*8) + ...
-
- The first term is 1/24, which is already greater than (pi-3)/4. Should
- that have been an alternating series?
-
- A form I like is:
-
- pi = 2 sum from i=0 to infinity product j=1 to i j/(2j+1) .
-
- This series has the advantage that each term is positive and less than
- half the preceding term (making a maximum error easy to calculate), each
- term can be computed from the preceding one by one multiplication by j
- and one division by 2j+1. Thus, tens of thousands of terms can be
- computed with routines that multiply or divide an extended-precision
- number by an ordinary-precision number, which are not difficult to
- write.
-
-
- -- edp (Eric Postpischil)
- "Always mount a scratch monkey."
- edp@rusure.enet.dec.com
-