0

Assume that $n \ge 1$ is an integer and that $a_1, a_2, ..., a_n$ is a sequence of $n$ integers. Prove that there are $i$ and $j$, with $1 \le i \le j \le n$, such that $a_i+a_{i+1}+ \cdots + a_j$ is divisible by $n$.

Here is what I have so far, I feel that I am very close to completing the proof, I just do not know how to finish it.

I am going to use the idea of the pigeonhole principle to prove this by assuming the opposite is true and finding a contradiction.
Let's assume there does not exist $i$ and $j$ such that $a_i+a_{i+1}+ \cdots + a_j$ is divisible by $n$

When we divide $a_i+a_{i+1}+ \cdots + a_j$ by $n$ there are $n-1$ remainders ($r$) we can get, $r={1, 2, \dots , n-1}$.
Now we look at the first $n$ possible sub sequences we can have when $i=1$, these are $(a_i), (a_i+a_{i+1}), \dots , (a_i+a_{i+1}+\dots+a_{i+n-1})$
By the pigeonhole principle when we divide each the sums of these sub sequences by $n$, since we have $n$ sub sequences and $n-1$ possible remainders, either one of them is divisible by $n$ or at least two of them share the same remainder. Let's call these two sums $a_h$ and $a_k$ where $a_k \gt a_h$.
Since $a_k$ and $a_h$ share the same remainder we can subtract them to get a number that is divisible by $n$, $a_k - a_h$ is divisible by $n$.

Now there is suppose to be a contradiction somewhere but I don't know what to do from here, why does this mean that there is an $i$ and a $j$ where $a_i+a_{i+1}+ \cdots + a_j$ is divisible by $n$?

idknuttin
  • 2,475

2 Answers2

3

Start with the sums: $$S_1=a_1\quad S_2=a_1+a_2\quad\cdots \quad S_n=a_1+\cdots +a_n$$

There are $n$ of these.

Case I. They are all different $\pmod n$

Then, as we have $n$ distinct classes $\pmod n$ we must hit them all, including $0$.

Case II. Two of them are congruent $\pmod n$.

Then if $S_k\equiv S_l\pmod n$ with $k>l$ we have $S_k-S_l\equiv 0 \pmod n$ but $S_k-S_l=a_{l+1}+\cdots +a_k$.

lulu
  • 70,402
  • So in case two, $S_k-S_l$ will always equal the sum of another consecutive sub sequence, correct? That's what I was not sure about. – idknuttin Sep 20 '16 at 00:31
  • 1
    Yes. it's the consecutive sum between $a_{l+1}$ and $a_k$. – lulu Sep 20 '16 at 00:32
1

consider the $n+1$ numbers $$b_0=0,$$ $$b_1=a_1,$$ $$b_2=a_1+a_2,$$ $$\dots$$ $$b_n=a_1+a_2+\dots +a_n.$$ There must be two of them, say $b_i,b_j$ such that $b_i\equiv b_j $ mod $n$.

Q-Zhang
  • 1,608