I am having trouble finding any techniques that allow people to solve double sums where the indices rely on each other. For example, suppose I have the following sum:
$$\sum_{i=1}^{n-1}\sum_{j=i+1}^n i + j$$
What techniques can I use to eliminate the sums, and arrive at a simple algebraic expression in terms of only n?
Comment:
In (1) we write the index region somewhat more conveniently.
Note that we assume the index $j$ is a bound variable and so we write $(i+j)$ consequently in parenthesis.
In (2) we exchange the sums so that the inner sum starts now with $i=1$.
In (3) we use the summation formula $\sum_{k=1}^n c= n\cdot c$.
In (4) we use the summation formula $\sum_{k=1}^n k=\frac{1}{2}n(n+1)$.
In (5) we do some simplifications and start with $j=1$ without changing anything since we are adding $0$ only.
In (6) we use the formula $\sum_{k=1}^nk^2=\frac{1}{6}n(n+1)(2n+1)$ and the same as in (4).
In (7) we do some final simplifications.