Is this what you are looking for? http://www.jsoftware.com/jwiki/Essays/Fibonacci_Sums
----- Original Message ----- From: John Randall <[EMAIL PROTECTED]> Date: Thursday, December 4, 2008 20:33 Subject: Re: [Jprogramming] Fibonacci shift To: Programming forum <[email protected]> > Oleg Kobchenko wrote: > > > > But here's a greedy Change: > > > http://www.personal.kent.edu/~rmuhamma/Algorithms/MyAlgorithms/Greedy/greedyIntro.htm> > > change=: 2 -/\ (] - I. { [)^:([EMAIL PROTECTED])^:a: > > > > 100 25 10 1 change 289 > > 100 100 25 25 25 10 1 1 1 1 > > > Thanks, Oleg. Your formulation is similar to my z, which > is slightly > simpler on account of being monadic. At least I am > somewhat along the > right lines. > > The representation as Fibonacci numbers is like the change > problem. The > coins are Fibonacci numbers, but you only have one of > each. Requring the > representation to be calculated greedily shows it is unique: > more is > needed for existence. A consequence is that the component > Fibonaccinumbers are non-consecutive. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
