Sorry, Fibonacci.

Find the least positive integer d d for which there exists an infinite arithmetic progression satisfying the following properties:

  1. Each term of the progression is a positive integer.
  2. The common difference of the progression is d d .
  3. No term of the progression appears in the Fibonacci sequence.

Details and assumptions

The Fibonacci sequence is defined by F 1 = 1 , F 2 = 1 F_1 = 1, F_2 = 1 and F n + 2 = F n + 1 + F n F_{n+2} = F_{n+1} + F_{n} for n 1 n \geq 1 .

The arithmetic progression has infinitely many terms.


The answer is 8.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

7 solutions

Michael Tang
Aug 25, 2013

Notice that the consecutive integers 1 , 2 , 3 1, 2, 3 appear in the Fibonacci sequence. Therefore, d d cannot be 1 , 2 1, 2 or 3 3 ; otherwise, the sequence would contain one of the integers.

Now note that all terms in the arithmetic progression are congruent modulo d . d. So, if the Fibonacci sequence contains every possible residue modulo d , d, then it is impossible to find any such progression. We check the first few values of d d by taking the Fibonacci sequence mod d d :

  • In mod 4 , 4, the Fibonacci sequence is 1 , 1 , 2 , 3 , 1 , 0 , 1,1, 2, 3, 1, 0,\ldots which contains all residues.

  • In mod 5 , 5, the Fibonacci sequence is 1 , 1 , 2 , 3 , 0 , 3 , 3 , 1 , 4 , 1, 1, 2, 3, 0, 3, 3, 1, 4, \ldots which contains all residues.

  • In mod 6 , 6, the Fibonacci sequence is 1 , 1 , 2 , 3 , 5 , 2 , 1 , 3 , 4 , 1 , 5 , 0 , 1, 1, 2, 3, 5, 2, 1, 3, 4, 1, 5, 0, \ldots which contains all residues.

  • In mod 7 , 7, the Fibonacci sequence is 1 , 1 , 2 , 3 , 5 , 1 , 6 , 0 , 6 , 6 , 5 , 4 , 1, 1, 2, 3, 5, 1, 6, 0, 6, 6, 5, 4, \ldots which contains all residues.

  • In mod 8 , 8, the Fibonacci sequence is 1 , 1 , 2 , 3 , 5 , 0 , 5 , 5 , 2 , 7 , 1 , 0 , 1 , 1 , 1, 1, 2, 3, 5, 0, 5, 5, 2, 7, 1, 0, 1, 1, \ldots Notice that this sequence has now repeated, without containing the residues 4 4 or 6. 6. Therefore, d = 8 \boxed{d=8} is the minimum possible value of d . d. \blacksquare


Note: A possible progression would be 4 , 12 , 20 , 28 , 4, 12, 20, 28, \ldots since all terms are 4 ( m o d 8 ) . 4 \pmod8.

Moderator note:

Nice job!

Several people expressed a desire for a more conceptual solution, without calculating the Fibonacci sequence modulo small integers. Alas, such solution probably does not exist.

You have to show that for mod 4 5 6 7 they're periodic as well. The problem with your solution is this argument: Say in mod 7, residue 6 only happens up until a certain point and then doesn't happen anymore, then I can take my arithmetic sequence AFTER that point.

Hendrata Dharmawan - 7 years, 9 months ago

Log in to reply

Oh, yeah. Thanks for your help. Showing the periodicity is simply just continuing the sequence, so that's not hard.

Michael Tang - 7 years, 9 months ago

Log in to reply

Right, or you could make the argument that since the number of pairs of residue is finite, so eventually it's going to repeat, so the sequence is periodic (and this applies for all d d ). And now since for d=4,5,6,7 all the residues show up early on, there's no need to continue until the period point, since we know they'll eventually repeat.

Hendrata Dharmawan - 7 years, 9 months ago

Please can you explain the line "In mod 8, the Fibonacci sequence is 1,1,2,3,5,0,5,5,2,7,1,0,1,1,… Notice that this sequence has now repeated, without containing the residues 4 or 6.", I didn't understand that if 4 and 6 are not residues then it is a possible solution.

Adya Jaiswal - 7 years, 9 months ago

Log in to reply

@Adya Jaiswal As I said, if the Fibonacci sequence contains every residue an infinite number of times, then no such sequence can exist. I found that 4 and 6 never appear in mod 8, therefore all arithmetic sequences whose terms are all 4 or 6 mod 8 can never intersect the Fibonacci sequence.

Michael Tang - 7 years, 9 months ago

The cycle of residues of fibonacci sequence m o d ( n ) mod(n) is known as the Pisano period of n n . I did the problem by observing the table of the first few Pisano periods , noting that 8 8 is the smallest integer whose Pisano period does not contain all its residues. I am sure that there must be a more elegant method though.


Sreejato Bhattacharya - 7 years, 9 months ago

Not a very elegant way, but I guess it works. Even though it probably is obvious, at least having some proof for the values repeating would be nice (namely that the residues of the previous two terms are added together to make the next term, thus when you have two of the same residues in a row then it is guaranteed to repeat again)

Jiao Wang - 7 years, 9 months ago

Food for thought: Is it a coincidence that 8 8 is a Fibonacci number?

Michael Tong - 7 years, 9 months ago

Log in to reply

This is most definitely a coincidence.

Alexander Borisov - 7 years, 9 months ago

I think it would be... I can't think of a reason otherwise.

Michael Tang - 7 years, 9 months ago

I followed this way too. I thought that the solution(s) posted would be a bit more "mathematical".

Arnab Animesh Das - 7 years, 9 months ago

Log in to reply

oh, yeah

Ewerton Cassiano - 7 years, 9 months ago

Sorry but could someone clarify what exactly the terminology "mod" and "residues" means in this case as these are terms I've not come across in these circumstances.

Also, I may have read the question wrong, but I thought 8 was impossible as I assumed the progression we are creating began at 0 ( as it can't start at 1 because one is a fib. number and so this contradicts the 3rd property immediately) and if it did begin at 0, 0+8=8 which is also a fib. number and so we have the same problem.....

I'm sure my understanding of arithmetic progression is flawed or I've misunderstood the question but some clarification on these points would be appreciated.

Luke Milsom - 7 years, 9 months ago

Log in to reply

"mod," short for "modulo," is a special number system. In modulo n , n, the only integers you use are 0 , 1 , 2 , , n 1 , 0, 1, 2, \ldots, n-1, which are also the remainders when you divide something by n . n. Those integers are also called the residues mod n . n. Every number has a unique residue mod n , n, which you determine by dividing the number by n n and taking the remainder. When I took the Fibonacci sequence in mods 4 , 5 , 6 , 7 , 8 , 4, 5, 6, 7, 8, I reduced each number in the sequence to its residue. (Of course, I didn't actually list out all the terms in normal base 10 10 first; I used the fact that you can add residues as well, as long as you reduce again.)

Arithmetic progressions can start anywhere you want, and then they go on forever (I believe - not too clear on this either). Then it is okay for the difference to be 8. 8.

Michael Tang - 7 years, 9 months ago

I also did it in this way,but can we find out a better way?

Yunhao King - 7 years, 9 months ago
Adam Buck
Aug 27, 2013

Let b n b_{n} be a sequence with minimal d d subject to the constraints given. For each N = 1 , 2 , 3 , . . . N = 1, 2, 3, ... consider the sequence ( a N , n ) (a_{N,n}) defined a N , n : = F n M o d N a_{N,n} := F_{n} Mod N .

Claim: If there exists integers i i and j j , with i < j i < j such that a i = a j a_i = a_j and a i + 1 = a j + 1 a_{i+1} = a_{j+1} , then sequence (\ (a_N,n ) repeats the sequence a i , a i + 1 , . . . , a j 1 a_i, a_{i+1}, ... , a_{j-1} repeats indefinitely.

Proof: This follows from how ( a N , n ) (a_{N,n}) is defined as well as the properties of Mod and the fact that the Fibonacci sequence is a recursion relation with two seed values.

Claim: There is a repeating sequence for each N.

Proof: Since each sequence ( a N , n ) (a_{N,n}) can take on at most N N values, there are only N 2 N^2 distinct pairs of successive terms a i a_i and a i + 1 a_{i+1} , therefore there must eventually be an occurrence of the situation described in the previous claim.

Consider ( b 1 ) (b_1) of the desired sequence. If d is the common difference then, a d , n a_{d,n} must contain only finitely many of some m m in 0 , 1 , 2 , . . . d 1 {0, 1, 2, ... d-1} . Otherwise b n b_n will hit every Fibonacci number greater than b 1 b_1 such that b 1 M o d d = 0 b_1 Mod d = 0 .

Thus the smallest N N such that the repeating part of ( a N , n ) (a_{N,n}) is missing one of the values 0 , 1 , 2 , . . . , N 1 0, 1, 2, ..., N-1 is the d d we are looking for. We begin listing the sequence a N , n a_{N,n} until repetition is discovered.

We start with N = 2: 1, 1, 0 , 1, 1,...

N = 3: 1, 1, 2, 0, 2, 2, 1, 0, 1, 1, ...

N = 4: 1, 1, 2, 3, 1, 0, 1, 1, ....

N = 5: 1, 1, 2, 3, 0, 3, 3, 1, 4, 0, 4, 4, 3, 2, 0, 2, 2, 4, 1, 0, 1, 1, ...

N = 6: 1, 1, 2, 3, 5, 1, 0, 1, 1, ...

N = 7: 1, 1, 2, 3, 5, 1, 6, 0, 6, 6, 5, 4, 2, 6, 1, 0, 1, 1, ...

N = 8: 1, 1, 2, 3, 5, 0, 5, 5, 2, 7, 1, 0, 1, 1,...

We see that N = 8 N = 8 has ( a N , n (a_{N,n} with no instances of 4 or 6. So d = 8 d = 8 .

Let ( a n ) (a_n) be an arithmetic progression that satisfies the given conditions. Let d d be its common difference. Assume that the set of congruence classes of ( F n ) (F_n) modulo d d consists of all integers 0 , 1 , , d 1. 0,1,\dots, d-1. Since each residue is obtained from the residues of the previous terms (by using the recurrence relation of the Fibonacci sequence), each residue occurs infinitely many times with a fixed pattern. For example, the Fibonacci numbers mod 3 3 are 1 , 1 , 2 , 0 , 2 , 2 , 1 , 0 , 1 , 1 , 2 , 0 , 2 , 2 , 1 , 0 , 1 , 1 , 2 , 0 , 2 , 2 , 1 , 0 , . 1, 1, 2, 0, 2, 2, 1, 0, 1, 1, 2, 0, 2, 2, 1, 0, \overline{1,1,2,0,2,2,1,0,} \dots. This allows us to choose a positive integer m m such that F m a 1 and F m a 1 ( m o d d ) . F_m \ge a_1 \quad \text{and} \quad F_m \equiv a_1 \pmod{d}. This implies F m a 1 = k d F_m-a_1 = kd or F m = a 1 + k d F_m = a_1 + kd for some nonnegative integer k k , which is the ( k + 1 ) (k+1) st term of the arithmetic progression ( a n ) (a_n) . This is a contradiction to condition 3.

Hence we must look for the smallest integer d d where the set of congruence classes of ( F n ) (F_n) modulo d d is not complete. After some numerical experiments, we find that such smallest number is d = 8 , \boxed{d=8}, where the residues of ( F n ) (F_n) mod d d are 0 , 1 , 2 , 3 , 5 , 0,1,2,3,5, and 7. 7. Therefore, if we, for example, let ( a n ) (a_n) be defined by a n = 4 + ( n 1 ) 8 = 8 n 4 , a_n = 4+(n-1)8 = 8n-4, then ( a n ) (a_n) is an arithmetic progression with no term of the progression appearing in the Fibonacci sequence, since a n 4 ( m o d 8 ) a_n \equiv 4 \pmod{8} for all n n but F n ≢ 4 ( m o d 8 ) F_n \not\equiv 4 \pmod{8} for all n n .

Tran Dinh Duy Vu
Aug 30, 2013

I tried to write down the remainder when Fibonacci numbers is divided by d (d>2) with d=8 the sequence of remainder is periodic and in one period the sequence lacks of " 4" and "6" It is 1 1 2 3 5 0 5 5 2 7 1 0 1 1 2 3..... From d=2 to 7 the period contains all the number less than d So if we choose the starting point as 4 or 6 the sequece of remainder contains no zero. That is no fibonacci number -6 (or 4) is divisible by 8 Or every Fibonacci numbers \neq 6(or 4) + 8\times n So the answer is 8

Russell Few
Aug 26, 2013

So I am assuming that the progression is infinite, otherwise there would be no definite answer.

Note that if we have an arithmetic progression, there exists a a and b b such that all the terms in that progression is congruent to a m o d b a mod b . Note that if we consider the terms of the fibonacci sequence m o d b mod b , a each term that appears at least once would appear infinitely many times. This is because there is a finite number of possibilities for 2 2 consecutive terms and each term depends solely on the 2 2 previous terms. So we just have to look for the lowest d d such that there are no terms in the fibonacci sequences that are congruent to a m o d d a mod d for some integer a a .

We need not to consider m o d 1 , 2 , 3 mod 1, 2, 3 because the first 3 3 terms are 1 , 2 , 3 1, 2, 3 .

mod 4: 1, 1, 2, 3, 1, 0. all exists. mod 5: 1, 1, 2, 3, 0, 3, 3, 1, 4, all exists. mod 6: 1, 1, 2, 3, 5, 2, 7, 1, all exists. mod 7: 1, 1, 2, 3, 5, 1, 6, 0, 6, 6, 5, 4, 2, all exists.

But for mod 8, 1, 1, 2, 3, 5, 0, 5, 5, 2, 7, 1, 0, 1, 1. Note that the 1, 1 already repeated but there still has no 6. Thus the 6 will never come in the sequence because the 1, 1, 2, 3, 5, 0, 5, 5, 2, 7, 1, 0, pattern will keep on repeating as each term is dependent solely on the previous 2 terms.

Indeed, if some residue appears in a Fibonacci sequence once, it will appear infinitely many times. However, the reason for this is a bit more subtle. It has to do with the fact that one can reconstruct two consecutive residues of the Fibonacci sequence from the next two consecutive ones uniquely.

Alexander Borisov - 7 years, 9 months ago

LaTeX " a \pmod b " for stating mods.

Michael Tong - 7 years, 9 months ago

Log in to reply

ok, thanks!

Russell FEW - 7 years, 9 months ago

I think also numbers that are 4(mod 8) will satisfy this equation

Rindell Mabunga - 7 years, 9 months ago
Ton de Moree
Aug 29, 2013

Our sequence can be written as a + d n a+dn ( n N (n \in \mathbb{N} ).

We are looking for the smallest d d so that there is an a a that does not appear in the Fibonacci sequence modulo d d .

This can be done with trial and error. The Fibonacci sequence modulo 8 8 is

1 , 1 , 2 , 3 , 5 , 0 , 5 , 5 , 2 , 7 , 1 , 0 , 1 , 1 , . . . 1, 1, 2, 3, 5, 0, 5, 5, 2, 7, 1, 0, 1, 1, ...

Both 4 4 and 6 6 don't appear in this sequence, so the sequences defined by 4 + 8 n 4+8n ( n N ) (n \in \mathbb{N}) and 6 + 8 n 6+8n ( n N ) (n \in \mathbb{N}) both don't contain a number from the Fibonacci sequence.

I heavily relied on prior knowledge that the Fibonacci sequence is periodic modulo any number n n .

For a short proof of that: Given n n , there are only a finite number of possible duo's. Therefore there must be a point in the sequence where a duo of numbers appears for the second time. From here on out, the Fibonacci sequence repeats itself.

Ton de Moree - 7 years, 9 months ago
Freddy Román
Aug 28, 2013

All of the terms of an arithmetic progression with common difference d d leave the same remainder when divided by d d . Using that fact, if we find a d d such that the Fibonacci sequence modulo d d enters a cycle that skips at least one of the possible remainders, there is an arithmetic sequence with common difference d d that satisfies the problem's constraints.

By trial and error, we can see that the first d d with these properties is 8.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...