Let n≥4 be an integer. We need to find all functions W:{1,…,n}2→R such that for every partition [n]=A∪B∪C into disjoint sets, the following condition holds:
a∈A∑b∈B∑c∈C∑W(a,b)W(b,c)=∣A∣∣B∣∣C∣.
To solve this, we denote the function W by f for simplicity. We start by considering specific partitions of [n].
First, consider the partition P({1},{2},{3,4,5,…}):
f(1,2)f(2,4)+f(1,2)f(2,5)+⋯+f(1,2)f(2,n)=(n−2)−f(1,2)f(2,3).
Next, consider the partition P({1},{3},{2,4,5,…}):
f(1,3)f(3,4)+f(1,3)f(3,5)+⋯+f(1,3)f(3,n)=(n−2)−f(1,3)f(3,2).
Now, consider the partition P({1},{2,3},{4,5,…}):
f(1,2)f(2,4)+f(1,2)f(2,5)+⋯+f(1,2)f(2,n)+f(1,3)f(3,4)+f(1,3)f(3,5)+⋯+f(1,3)f(3,n)=2(n−3).
This simplifies to:
(n−2)−f(1,2)f(2,3)+(n−2)−f(1,3)f(3,2)=2(n−3)⟹f(1,2)f(2,3)+f(1,3)f(3,2)=2.
Similarly, for any distinct a,b,c:
f(a,b)f(b,c)+f(a,c)f(c,b)=2.
Considering P({3,4,5,…},{2},{1}), P({2,4,5,…},{3},{1}), and P({4,5,…},{2,3},{1}), we get:
f(3,2)f(2,1)+f(2,3)f(3,1)=2,
which generalizes to:
f(a,b)f(b,c)+f(b,a)f(a,c)=2.
Thus, we see that:
f(a,c)f(c,b)=f(b,a)f(a,c)⟹f(a,c)=0orf(c,b)=f(b,a).
Suppose f(a,c)=0 for some a=c. Considering P({a},{c},[n]−{a,c}), we get 0=n−2, a contradiction. Hence, f(c,b)=f(b,a) for all distinct a,b,c.
This implies that f(x,b) and f(b,x) are constants for each b and all x=b. Consequently, f(x,y) is a constant k if x=y. From the condition 2k2=2, we find k=1 or k=−1.
Therefore, all solutions are of the form where W(a,a) can be any value, and for all distinct a,b∈[n], W(a,b) equals a constant k, where k=1 or k=−1.
The answer is: W(a,b) = k for all distinct } a, b and } k = 1 or } k = -1.}