1. **Define the function f(a,b):**
f(a,b)=a⌈bν⌉−b⌊aν⌋
We need to show that the number of excellent pairs (a,b) is equal to the sum of the positive divisors of m.
2. **Relate f(a,b) to f(a+b,b) and f(a,b+a):**
- If {aν}+{bν}≥1, then:
f(a+b,b)=(a+b)⌈bν⌉−b⌊(a+b)ν⌋=a⌈bν⌉−b⌊aν⌋+b(⌈bν⌉−⌊bν⌋)=f(a,b)+b
because ⌊(a+b)ν⌋=⌊aν⌋+⌊bν⌋.
- Similarly, for f(a,b+a):
f(a,b+a)=a⌈(b+a)ν⌉−(b+a)⌊aν⌋=a(⌈bν⌉+⌈aν⌉−1)−(b+a)⌊aν⌋=a⌈bν⌉−b⌊aν⌋=f(a,b)
- When {aν}+{bν}<1, we have:
f(a,b)=f(a+b,b)=f(a,b+a)−a
3. **If (a,b) is excellent, then min(a,b)≤m:**
- For c,d≥0, f(c,d)=c⌈dν⌉−d⌊cν⌋≥0.
- If (a,b) is excellent, then f(a−b,b) and f(a,b−a) are not equal to f(a,b).
- If min(a,b)=a, then:
0≤f(a,b−a)=f(a,b)−a=m−a⟹a≤m
- Similarly, if min(a,b)=b, then:
0≤f(a−b,b)=f(a,b)−b=m−b⟹b≤m
4. Define relevant pairs and branches:
- A pair (a,b) is relevant if a,b are positive integers, min(a,b)≤m, and f(a,b)≤m.
- Draw arrows from (a,b)→(a+b,b) and (a,b)→(a,b+a).
- This forms a tree due to the Euclidean Algorithm's uniqueness.
5. Prove that relevant branches are finite:
- Suppose B(a,b) is infinite. Choose an infinite path of pairs (ci,di) for i=1,2,….
- Since min(ci,di)≤m, there exists M such that ci=c for all i≥M.
- f(ci,di)≤m and is non-decreasing, so it is eventually constant.
- This leads to a contradiction because {cν}+{(dM+(i−M)ci)ν}≥1 for all i≥M, which is impossible for an irrational ν.
6. Count solutions and excellent pairs:
- Let N(a,b) be the number of solutions for x,y∈Z≥0 in ax+by+f(a,b)=m.
- Let g(a,b) be the number of excellent pairs in B(a,b).
- We claim N(a,b)=g(a,b).
7. **Lemma: N(a,b)=N(a+b,b)+N(a,b+a):**
- If {aν}+{bν}≥1:
(a+b)x+by+f(a+b,b)=m⟺ax+b(x+y+1)+f(a,b)=m⟺ax′+by′+f(a,b)=m
where x′=x and y′=x+y+1.
- Similarly:
ax+(b+a)y+f(a,b+a)=m⟺a(x+y)+by+f(a,b)=m⟺ax′+by′+f(a,b)=m
where x′=x+y and y′=y.
8. **Prove g(a,b)=N(a,b):**
- If (a,b) is excellent, then g(a,b)=1 and N(a,b)=1 because x=y=0.
- Otherwise, g(a,b)=g(a+b,b)+g(a,b+a) because the set of arrows and pairs form a tree.
- Using induction, g(a,b)=N(a,b) for all relevant (a,b).
9. Finishing the proof:
- Let k(m) be the number of excellent pairs for m.
- Partition the set of positive integer pairs (a,b) into sets T1,T2,… where (a,b)∈Tgcd(a,b).
- Then:
k(m)=i=1∑∞g(i,i)=i=1∑∞N(i,i)
- We know:
N(i,i)={0imif i∤mif i∣m
- Therefore:
k(m)=i=1∑∞N(i,i)=d∣m∑N(d,d)=d∣m∑dm=d∣m∑d=σ(m)
■
The final answer is σ(m)