A moment’s thought: physical derivations of Fibonacci summations
David Treeby
A moments thought: physical derivations of Fibonacci summations - - PowerPoint PPT Presentation
A moments thought: physical derivations of Fibonacci summations David Treeby Two pretty formulas n n 2 j 3 = j j =1 j =1 n n 2 F 3 j F 3 F 2 j +1 = j F j +1 j =1 j =1 Two models of the Fibonacci
David Treeby
n
j3 =
j 2
n
F 3
j F 3 j+1 =
F 2
j Fj+1
2
Theorem
The number of ways to tile a board of length j with squares and dominoes is fj where f0 = f1 = 1 and fj = fj−1 + fj−2. j square domino
f4 = 5
Proof.
Consider a j-board. Suppose that this can be tiled in fj ways. j
Proof.
Consider a j-board. Suppose that this can be tiled in fj ways. j Case 1. If the first tile is a square then there are fj−1 ways to tile the remaining (j − 1)-board. j − 1
Proof.
Consider a j-board. Suppose that this can be tiled in fj ways. j Case 1. If the first tile is a square then there are fj−1 ways to tile the remaining (j − 1)-board. j − 1 Case 2. If the first tile is a domino then there are fj−2 ways to tile the remaining (j − 2)-board. j − 2
Proof.
Consider a j-board. Suppose that this can be tiled in fj ways. j Case 1. If the first tile is a square then there are fj−1 ways to tile the remaining (j − 1)-board. j − 1 Case 2. If the first tile is a domino then there are fj−2 ways to tile the remaining (j − 2)-board. j − 2 Therefore fj = fj−1 + fj−2.
Theorem
f2
0 + f2 1 + f2 2 + · · · + f2 n = fnfn+1
Proof.
(n + 1)-board? n + 1 n
Proof.
(n + 1)-board? n + 1 n Answer 1. There are fn and fn+1 tilings of the first and second board, respectively. Therefore there are fnfn+1 tilings of both boards.
Answer 2. Condition on the position j corresponding to the last common edge of each tiling. j
Answer 2. Condition on the position j corresponding to the last common edge of each tiling. j To avoid future common edges, there is exactly one way to finish the tiling.
Answer 2. Condition on the position j corresponding to the last common edge of each tiling. j To avoid future common edges, there is exactly one way to finish the tiling. Prior to this, the first and second board can each be tiled fj ways, so both can be tiled f2
j ways.
Answer 2. Condition on the position j corresponding to the last common edge of each tiling. j To avoid future common edges, there is exactly one way to finish the tiling. Prior to this, the first and second board can each be tiled fj ways, so both can be tiled f2
j ways.
Summing over all possible values of j gives n
j=0 fj2 tilings.
Construct a rectangle comprising two adjacent squares of side F1 = 1 and F2 = 1.
F1 F2
For every j ≥ 2 we construct a square of side Fj on the larger side
F1F3 F2
For every j ≥ 2 we construct a square of side Fj on the larger side
F1F3 F2 F4
For every j ≥ 2 we construct a square of side Fj on the larger side
F1F3 F5 F2 F4
For every j ≥ 2 we construct a square of side Fj on the larger side
F1F3 F5 F2 F4 F6
For every j ≥ 2 we construct a square of side Fj on the larger side
F1F3 F5 F7 F2 F4 F6
The jth square has side Fj where F1 = F2 = 1 and Fj = Fj−1 + Fj−2 for n ≥ 2.
F1F3 F5 F7 F2 F4 F6
Theorem
F 2
1 + F 2 2 + · · · + F 2 n = FnFn+1
F1F3 F5 F7 F2 F4 F6
Proof.
The total area is equal to the sum of its parts.
Fibonacci numbers?
Fibonacci numbers?
number there has to be. However, the answer is not expressible as the product of Fibonacci numbers.
determining the sum of cubes of Fibonacci numbers?
determining the sum of cubes of Fibonacci numbers?
Theorem
n
F 3
j = Fn+1F 2 n+2 + (−1)nFn − 2F 3 n
2
Proof.
International Conference on Fibonacci Numbers and their Applications, (2009), 45-51
For the geometric proof we require one preliminary result,
n
F 2
j Fj+1 = 1
2FnFn+1Fn+2.
x = n
j=1 Ajxj
A and y = n
j=1 Ajyj
A
Example
Fj Fj+1 Fj−1
Example
Fj Fj+1 Fj−1 1. F 2
j+2 = 4FjFj+1 + F 2 j−1
Example
1 4 2 3 Fj Fj+1 Fj−1 5
Fj+2 2
O Fj+2
Example
1 4 2 3 Fj Fj+1 Fj−1 5
Fj+2 2
O Fj+2 2F 2
j−1Fj + 4FjFj+1Fj+2 = F 3 j+2 − F 3 j−1
Theorem
n
FjFj+1Fj+2 = F 3
j + F 3 j+1 + F 3 j+2 − Fj−1FjFj+1 − 2
4 .
Example
Fj Fj+1 Fj+1 Fj
Example
Fj Fj+1 Fj+1 Fj F 2
j+2 = 2FjFj+1 + F 2 j + F 2 j+1
Example
1 2 3 4 Fj Fj+1 Fj Fj+1 O
Fj+2 2
Fj+2 F 3
j + 3FjFj+1Fj+2 = F 3 j+2 − F 3 j+1
Theorem
n
F 3
j =
3F 2
j+1Fj − F 3 j+1 − F 3 j + 1
2 .
Starting point: Tn = 1 + 2 + · · · + n = 1 2n(n + 1)
1
1 2Tn
Tn 2 · · · j · · · n
j xj Tj−1 Tj
j
1 2j2
n
j3 =
j 2
The same trick works more generally. Suppose our starting point is
n
F 2
j Fj+1 = 1
2FnFn+1Fn+2.
F 2
1 F2F 2 2 F3
F 2
3 F4
· · · F 2
nFn+1 1 4FnFn+1Fn+2 1 2FnFn+1Fn+2
F 2
j Fj+1
xj Sj−1 Sj
F 2
j Fj+1 1 2FjF 2 j+1
n
F 3
j F 3 j+1 = 1
4F 2
nF 2 n+1F 2 n+2
n
F 3
j F 3 j+1 =
F 2
j Fj+1
2
n
F 5
j F 5 j+1F2j+1 = 1
8F 4
nF 4 n+1F 4 n+2