The following is from Clark Kimberling’s problems page.

14. Never 3?

Let G = (1 + 51/2)/2 and f(n) = floor(n2G) – n floor(nG), for n = 1,2,3, . . . .

Examples:

f(n) = 0 for n = 0, 1, 2, 5, 13, 34, . . .
f(n) = 1 for n = 4, 10, 16, 68, 178, . . .
f(n) = 2 for n = 3, 7, 18, 47, 123, . . .

1. Reward of $20 for a proof that f(n) is never 3.

2. Reward of $10 for the least n such that f(n) = 3, if there is such an n.

3. Reward of $30 for a closed-form formula (with proof) for all positive integers k for f(n) is never k.

Accepted solution

f(n) is never 3

Remark
Note that f(n) = floor(g(n)) where g(n) = n × frac(nG)   [frac() = fractional part].

Hence G can be replaced by frac(G), and this is done below.

If we look at the values of g(n) we find empirically that they have a sequence of discrete accumulation points, thus:

g(n) tends from above to 0.477 =? 1/51/2 for n = 1, 2, 5, 13, …
g(n) tends from above to 1.789 =? 4/51/2 for n = 4, 10, 26, 68, …
g(n) tends from above to 2.236 =? 5/51/2 for n = 3, 7, 18, 47, …
g(n) tends from above to 4.025 =? 9/51/2 for n = 6, 15, 39, 102, …
g(n) tends from above to 4.919 =? 11/51/2 for n = 9, 12, 23, 31, …

Since g(1), g(4) and g(3) are all less than 3 the result follows. So we need to prove this empirical finding.

 

Proof
Let a = (51/2 + 1)/2,  b = (51/2 – 1)/2 = frac(G).    Note that ab = 1.

The nth Fibonacci number Fn = 5–1/2[an – (–b)n]    (in the notation F0 = 0, F1 = F2 = 1, …)

and a bit of calculation shows that frac(F2m+1b) = F2m+1b – F2m = b2m+1.

Lemma Every positive integer n can be expressed as a combination of distinct odd-indexed Fibonacci numbers:

n = F2p+1 ± F2q+1 ± … + F2z+1

where each coefficient is ±1 and the largest and smallest terms both have coefficient +1.

Proof of lemma, by induction

1 = F1
2 = F3
3 = F3 + F1
4 = F5 – F3 + F1
5 = F5
6 = F5 + F1
7 = F5 + F3
8 = F5 + F3 + F1

Having got F2m = F2m–1 + F2m–3 + … + F1, continue to F2m+1 by expressing F2m = F2m+1 – F2m–1 and using the previous results. The term F2m–1 may cancel:

9 = F7 – F5 + F1
10 = F7 – F5 + F3
11 = F7 – F5 + F3 + F1
12 = F7 [– F5 + F5] – F3 + F1
13 = F7 [– F5 + F5]

Now continue to F2m+2 with leading term F2m+1 plus the previous results:

14 = F7 + F1

21 = F7 + F5 + F3 + F1

(The representation is not always unique; see “By the way” near the end of this solution.)

End of lemma.

To get frac(nb), express n as in the lemma and use frac(F2m+1b) = b2m+1. We just add the terms ±b2m+1. This works because

b + b3 + b5 + b7 + … = 1

so that (i) sum < 1 (ii) sum >= 0, because the coefficient of the largest term (smallest power of b) is +1.

Next look at g(n) = n × frac(nb). For simplicity, this will be done by means of a typical example, in which

n = F2m+7 – F2m+3 + F2m+1 (m = 0, 1, 2, …)

corresponding to n = 12, 31, … . We have (using ab = 1)

g(n) = 5–1/2(a2m+7 + b2m+7 – a2m+3 – b2m+3 + a2m+1 + b2m+1)(b2m+7 – b2m+3 + b2m+1)

= 5–1/2[3 + (a6 + b6) – (a4 + b4) – (a2 + b2) + (b2m+7 – b2m+3 + b2m+1)2]

For k = 1, 2, 3, … let Ek = a2k + b2k = F2k–1 + F2k+1. Thus

E1 = 3, E2 = 7, E3 = 18, E4 = 47, … . Then

g(n) = 5–1/2(3 + E3 – E2 – E1 + R) = 5–1/2(11 + R),

where 0 < R < 1 and R tends to 0 as m tends to infinity. This confirms part of the empirical finding above.

Note that the constant 11 is obtained as (1 – a2 + a6)(1 – b2 + b6). Since ab = 1 we also have 11 = (1 – a4 + a6)(1 – b4 + b6) corresponding to n = F2m+7 – F2m+5 + F2m+1 = 9, 23, …. The accumulation point 11/51/2 (unlike the smaller ones) is the limit of two interleaved series.

Let the accumulation points be L/51/2 where empirically we’ve noticed L = 1, 4, 5, 9, 11. In general the L values are:

(1)(1) = 1
(1 + a2)(1 + b2) = 5
(1 – a2 + a4)(1 – b2 + b4) = 4
(1 + a4)(1 + b4) = 9
(1 – a2 + a6)(1 – b2 + b6) = 11
(1 – a4 + a6)(1 – b4 + b6) = 11
and so on for all similar combinations of even powers, in which the coefficient of the highest power used is +1.

If the highest power used is 2k (k = 1, 2, 3…) then the lowest possible L value is

(1 – a2 – a4 … – a2k–2 + a2k) (1 – b2 – b4 … – b2k–2 + b2k)

which after a bit of calculation is found to be

F2k–3 + F2k–5 + 2 (with the natural definitions F–1 = 1, F–3 = 2).

For k = 1, 2, 3, 4, 5, … this gives L = 5, 4, 5, 9, 40, … with the values increasing after L = 4. (By the way, L = 5 is repeated because F2m+7 – F2m+5 – F2m+3 + F2m+1 = F2m+5 + F2m+3.)

So the empirical list at the start includes all L values <= 9. This proves that either g(n) <= g(3) = 2.562 or g(n) > 9/51/2 = 4.025, and hence that f(n) = floor(g(n)) is never 3, QED.