There are red cards and blue cards. Each card has a positive integer between and (inclusive) written on it. Prove that we can choose a (non-empty) subset of the red cards and a (non-empty) subset of the blue cards, so that the sum of the numbers on the chosen red cards equals the sum of the numbers on the chosen blue cards.
Problem 1518
Official solution
1. Define the labels and cumulative sums:
Let be the labels of the red cards and be the labels of the blue cards. Define the cumulative sums:
2. **Construct the set of points :**
Consider the set of points . For each point , label it with .
3. Apply the Pigeonhole Principle:
We claim that some two points in have the same label. To show this, we will use the Pigeonhole Principle.
4. **Define the sets :**
For each , let be the set of points such that . This means contains points on the boundary of the square grid at distance from the point .
5. Label the points and apply discrete continuity:
Let be the label of the point , which is . For every , note that since and are both in , by "discrete continuity" there must exist some such that the label of is in the interval .
6. Count the distinct labels:
Since the 's are mutually disjoint and there are of them, we now know the existence of distinct points in which all have labels in . By the Pigeonhole Principle, there must be at least two points in with the same label.
7. Identify the points with the same label:
Hence, we can pick such that and and and have the same label.
8. Derive the subset sums:
By the definition of our table, we have:
Rearranging, we get:
This implies that the sum of the numbers on the chosen red cards (from to ) equals the sum of the numbers on the chosen blue cards (from to ).
Therefore, we have shown that it is possible to choose a (non-empty) subset of the red cards and a (non-empty) subset of the blue cards such that the sum of the numbers on the chosen red cards equals the sum of the numbers on the chosen blue cards.