Hello, this may be a very naive question : to compute the c.f. of a real number to high precision by using the most elementary operations, there is a way isn't ?? by using only additions on large numbers you add up the fractional part of x a certain number of times, let's say n times, just before you reach 1 then do : 1 - nx = new value n may be big sometimes but you do not have to do any inverses or multiplication. nx is computed only by adding x to itself n times. Is the cost of that algorithm bigger than known methods ? cordialement, Simon Plouffe 2010/8/6 Hans Havermann <pxp@rogers.com>
Bill Gosper:
Has anyone listed the largest terms in Hans's data?
I have. :)
http://chesswanks.com/seq/cfpi/
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun