Home

[This page is auto-generated; please prefer to read the article’s PDF Form.]



[Prev]   [Up]   [Next]   

Continued Fraction of ϕ

We first note that ϕ is the positive solution to the quadratic equation:

             x   1 + x
             --= -----
      2      1     x
⇔    x − x − 1 = 0
So, ϕ = (1 + √ --
  5)2, and
        1-
ϕ = 1 + ϕ

Please refer to the conventions used in section “Background” of [1] about continued fractions. The continued fraction of ϕ is:

        ---1----
ϕ = 1 + 1+  -1--= [1;1,1,1,...]
            1+...

which is special since all ai terms are 1.

We define p1 = 1 and q1 = 0. For this continued fraction, we have p0 = 1,q0 = 1. And due to the recurrence-relation (2) in [1], for all k 0:

pk+1 = ak+1pk + pk−1 = pk + pk−1
qk+1 = ak+1qk + qk− 1 = qk + qk−1

Thus, for this special case of ϕ, the sequences of pk and qk values each follow the same recurrence-relation as the Fibonacci Sequence, which is defined as:

F0 = 0
F1 = 1

Fn = Fn− 2 + Fn− 1 for all n ≥ 2
So, in the continued fraction of ϕ, for all k 0:
pk = Fk+2

qk = Fk+1
Thus the convergents which approximate ϕ are: pk∕qk = Fk+2∕Fk+1.

Now, consider any integer N 1 and its corresponding unique integers t, r and s, as defined by relations (3) and (4) in [1]. The relation (3) in [1], which defines t for N, now becomes:

     q   + q  ≤ N <  q + q
      t−1   t         t   t+1
⇔   Ft + Ft+1 ≤ N <  Ft+1 + Ft+2
⇔        Ft+2 ≤ N <  Ft+3                                         (2)

So, with α = ϕ, the t 0 corresponding to N is such that Ft+2 is the largest Fibonacci number not exceeding N.

The range of r as defined by relation (4) in [1] is: 1 r at+1. With ϕ, at+1 = 1 for all t 0. So in this case we have r = 1 for all N.

[Prev]   [Up]   [Next]