3

Here's a problem that is leading me in circles.

Consider the Fibonacci number $F_n$ defined by $F_0 = 0$, $F_1 = 1$ and $F_n = F_{n-1} + F_{n-2}$ for all $n \geq 2$.

Prove that $F_{2n-1} = F_{n}^2 + F_{n-1}^2$.

Tried an induction proof that lead me nowhere fast. So far we have only been given limited identities, and they are all basically summation formulas for the odd terms, or even terms, or all terms up to n. I also though that maybe if it was a difference, the difference of squares could factor to something, but I'm just not getting it.

4 Answers4

6

The identity may be derived from the interesting fact that

$$\left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right ) ^k = \left ( \begin{array} \\ F_{k+1} & F_k\\F_k & F_{k-1} \\ \end{array} \right ).$$

From this, we may observe that

$$\begin{align} \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right ) ^m \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right ) ^n &= \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right ) ^{m+n} \\ &= \left ( \begin{array} \\ F_{m+1} & F_m\\F_m & F_{m-1} \\ \end{array} \right ) \left ( \begin{array} \\ F_{n+1} & F_n\\F_n & F_{n-1} \\ \end{array} \right )\\ &= \left ( \begin{array} \\ F_{m+n+1} & F_{m+n}\\F_{m+n} & F_{m+n-1} \\ \end{array} \right ) \end{align} $$

Therefore we can say that, for example,

$$F_{m+n-1} = F_m F_n + F_{m-1} F_{n-1}$$

Plug in $m=n$ and the desired identity follows.

EDIT

The first identity quoted above follows from putting the Fibonacci recurrence into matrix form:

$$\left ( \begin{array} \\F_{k+2} \\ F_{k+1} \\ \end{array} \right ) = \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right ) \left ( \begin{array} \\F_{k+1} \\ F_k \\ \end{array} \right )$$

which may be immediately verified. We may repeat this matrix multiplication $k$ times to get

$$\left ( \begin{array} \\F_{k+2} \\ F_{k+1} \\ \end{array} \right ) = \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right )^k \left ( \begin{array} \\F_{2} \\ F_1 \\ \end{array} \right ) = \left ( \begin{array} \\ 1 & 1\\1 & 0 \\ \end{array} \right )^k \left ( \begin{array} \\1 \\ 1 \\ \end{array} \right )$$

Noting that $F_{k+2}= F_{k+1}+ F_k$ and $ F_{k+1} = F_k + F_{k-1}$, the stated identity follows.

Ron Gordon
  • 142,110
  • How did you come up with the 2x2 matrix identity above? What basis do you have to use that in your derivation? – Eleven-Eleven Feb 09 '13 at 05:21
  • @ChristopherErnst: this is a well-known identity, the proof of which I outlined above. I do not understand your second question. – Ron Gordon Feb 09 '13 at 06:35
  • Nice proof! But let me ask you one question: we get the following: $\begin{pmatrix} F_{k+1} & F_{k} \ F_{k} & F_{k-1} \end{pmatrix} \begin{pmatrix} 1 \ 1 \end{pmatrix}=\begin{pmatrix} 1 & 1 \ 1 & 0 \end{pmatrix}^k\begin{pmatrix} 1 \ 1 \end{pmatrix}$ But how to conclude from here that $\begin{pmatrix} F_{k+1} & F_{k} \ F_{k} & F_{k-1} \end{pmatrix}=\begin{pmatrix} 1 & 1 \ 1 & 0 \end{pmatrix}^k$ – RFZ Nov 28 '17 at 15:41
3

Using Binet's Fibonacci Number Formula,

$F_r=\frac{a^r-b^r}{a-b}$ where $a,b$ are the roots of $x^2-x-1=0\implies ab=-1,a+b=1$

So, $$F_n^2+F_{n-1}^2=\frac{(a^n-b^n)^2+(a^{n-1}-b^{n-1})^2}{(a-b)^2}$$

$$=\frac{a^{2n}+b^{2n}+a^{2n-2}+b^{2n-2}-2(ab)^{n-1}(ab+1)}{(a-b)^2}$$

$$=\frac{a^{2n-1}\left(a+\frac1a\right)+b^{2n-1}\left(b+\frac1a\right)}{(a-b)^2}\text{ as } ab=-1$$

$$=\frac{a^{2n-1}(a-b)+b^{2n-1}(b-a)}{(a-b)^2} \text{ as } \frac1a=-b,\frac1b=-a$$

$$=\frac{a^{2n-1}-b^{2n-1}}{(a-b)}=F_{2n-1}$$

  • This was exactly what I needed. I noticed, buried in my notes from class that the professor had briefly gone over the formula, but failed to show its use in applications. Now I feel dumb for having been told of its existence and not figuring it out.... – Eleven-Eleven Feb 09 '13 at 05:22
0

lab bhattacharjee has given excellent direct prove of the identity OP has seeking. I wonder my following indirect prove would count as a prove of the given identity.

If you use Euler's formula for Pythagorean triple, where you can use $n=F_{k-1}$ and $m=F_k$, then the hypotenuse of the triangle will be $c=m^2+n^2=F_k^2 + F_{k-1}^2$, which is equal to $c=F_k^2 + F_{k-1}^2=F_{2k-1}$ as shown in following table:

$$ \begin{array}{c|ccccc} k & n = F_{k-1} & m = F_k & a = F_k^2-F_{k-1}^2 & b = 2F_kF_{k-1} & c = F_k^2+F_{k-1}^2 \\ \hline 3 & 1 & 2 & 3 & 4 & 5 \\ 4 & 2 & 3 & 5 & 12 & 13 \\ 5 & 3 & 5 & 16 & 30 & 34 \\ 6 & 5 & 8 & 39 & 80 & 89 \\ 7 & 8 & 13 & 105 & 208 & 233 \\ k & F_{k-1} & F_{k} & F_{k-2}F_{k+1} & 2F_kF_{k-1} & F_{2k-1} \\ \hline \end{array} $$

Close inspection of above table reveals that $c = F_k^2+F_{k-1}^2 = F_{2k-1}$. For example, for the first line, $k=3$ so that $F_{2k-1}= F_{2\cdot3-1} = F_5=5$. For the second line, $k=4$ so that $F_{2k-1}= F_{2\cdot4-1} = F_7=13$.

Interestingly, close inspection of above table also reveals that $a = F_k^2-F_{k-1}^2 = F_{k-2}F_{k+1}$, which is also a Fibonacci identity.

0

Let me show an interesting, direct algebraic proof of your identity involving the Lucas numbers. The Lucas numbers are defined as

$$L_n = F_{n+1} + F_{n-1}$$

For the proof you need the following four results (all proven inductively): \begin{equation} F_1F_2 + F_2F_3 + ... + F_{2n-1}F_{2n} = F_{2n}^2 \quad \quad(1) \end{equation} \begin{equation} F_1F_2 + F_2F_3 + ... + F_{2n-1}F_{2n} + F_{2n}F_{2n+1} = F_{2n+1}^2 - 1 \quad \quad(2) \end{equation} \begin{equation} F_2 + F_4 + F_6+... + F_{2n} = F_{2n+1} -1 \quad \quad (3) \end{equation} \begin{equation} F_{2n} = F_{n}L_{n} \quad \quad(4) \end{equation}

From (1) and (2) we can write \begin{align*} F_n^2+F_{n-1}^2 &= 1 + (F_0F_1 + F_1F_2) + (F_2F_3 + F_3F_4) + ...(F_{n-2}F_{n-1} + F_{n-1}F_n) \\ &= 1 + F_1(F_0+F_2) + ... + F_{n-1}(F_{n-2} + F_{n}) \\ &= 1 + F_1L_1 + F_2L_2 +... + F_{n-1}L_{n-1} \end{align*} Now using (3) and (4) we can write \begin{align*} F_n^2+F_{n-1}^2 &= 1 + F_2 + F_4 ... + F_{2n-2} \\ &= 1 + [F_{2n-1} -1] \\ &= F_{2n-1} \end{align*}

HG11
  • 123
  • Also, I thought maybe I should write down a related formula here for the even Fibonacci numbers $$F_{2n} = F_{n+1}^2 - F_{n-1}^2 $$ – HG11 Dec 06 '24 at 06:44