Fibonacci Mystery - Numberphile
Video Statistics and Information
Channel: Numberphile
Views: 2,505,480
Rating: 4.9530668 out of 5
Keywords: Fibonacci Number (Literature Subject), numberphile, pisano, pattern
Id: Nu-lW-Ifyec
Channel Id: undefined
Length: 9min 48sec (588 seconds)
Published: Wed Sep 18 2013
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.
is there a given name for the problem of finding the length of period mentioned at the end?
One of my calculus professors asked me to prove that every Fibonacci sequence mod
[;n;]
is periodic. He didn't know he rediscovered Pisano periods but he found the pattern when he was bored as an undergraduate and said I reminded him of it.If you already know what the Fibonacci sequence is, skip to 1:30 or so.
My jaw actually dropped when I realised he was about to show that the modulus Fibbonacci sequences had the same additive property as the normal Fibbonacci.
I love numberphile.
I've written something about finding the Pisano periods algorithmically here. There's a SPOJ problem on finding them, though my Haskell for that is not nearly as fast as some others, so there's likely way better solutions.
Essentially, the pisano period (mod m) is the multiplicative order of the matrix M = [[1, 1], [1, 0]] in the group (Z/mZ)2 x 2. You then try to do magic to find this order.
Question- At 4:50 he says there could only be 1,2, or 4 zeros. Now I'm assuming that taking the fibonacci sequence mod n creates a group (which I don't really feel like checking), and he's going on the theorem stating that the order of any element in a group must divide the order of the group. But wouldn't that mean there could also be five zeros, since the group has an order of 20?
Some fooling around in excel suggest that using integer with modulo functions gives similar results dividing by non-integers. It looks like the pisano period of pi is 34; 1, 2, 3, 1, 1, 0, 2, 2, 1, 1, 2, 0, 0, 0, 0, 1, 1, 2, 1, 0, 1, 2, 1, 0, 1, 2, 1, 0, 1, 2, 0, 0, 0 -- it also looks like the zeroes rule does not apply for non-integers. =INT(MOD(Fibnum, non-int))
Fun fact, related to these periods:
The nth Fibonacci number is divisible by exactly the same number of powers of 5 as n is.
Another cool property of the Fibonacci sequence:
Take the limit of Fn+1/Fn as n--> infinity and you get the golden ratio.