Xref: utzoo sci.math:16411 comp.compression:237
Newsgroups: sci.math,comp.compression
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!think.com!snorkelwacker.mit.edu!stanford.edu!agate!garnet.berkeley.edu!greg
From: greg@garnet.berkeley.edu (Greg Kuperberg)
Subject: A program to compute E
Message-ID: <1991Apr4.192717.21850@agate.berkeley.edu>
Followup-To: sci.math
Sender: usenet@agate.berkeley.edu (USENET Administrator)
Reply-To: greg@math.berkeley.edu
Organization: U.C. Berkeley
Date: Thu, 4 Apr 91 19:27:17 GMT
Lines: 19

Since my program to compute Pi has been topped by another program which
is simpler, uses less space, and runs about as fast, I am compelled to
present another program to compute E:

main(){long int d[3301],i=0,j=3300,c=0;printf("2.");
while(j>1)d[j--]=1;while(i++<2000){for(j=3300;j;j--)
{c+=d[j]*1e5;d[j]=c%j;c/=j;}printf("%05d",c);}printf("\n");}

This time I feel safe that the code is hard to beat for brevity,
give or take a few characters.

The algorithm is based on the same idea as the two to compute Pi.
It converts .11111... from base 2,3,4,5,6,... to base 10.  I must
give credit to Kevin Nomura for telling me this algorithm many years
ago.  It inspired my program for computing Pi.  (Alas, my understanding
of Pi was not up to par with my understanding of this trick.)
----
Greg Kuperberg
greg@math.berkeley.edu