By Warren D. Smith 1 Oct 2006
Puzzle: Let the set {0,1,2,...,N-1} be covered by sums of exactly 2 elements (not necessarily distinct) of a subset Q containing q elements. For each N: what is the minimal q?
Example: N=9, q=4, set={0,1,3,4} means {0,1,2,...,8} is covered thus by pair-sums: 0=0+0, 1=0+1, 2=1+1, 3=0+3, 4=0+4=1+3, 5=1+4, 6=3+3, 7=3+4, 8=4+4 and the cardinality q=4 is minimal since (by the pigeonhole principle) no 3-element subset of {0,1,2,...,8} can work.
N | q | set |
---|---|---|
3 | 2 | 0,1 |
4 | 3 | 0,1,2 |
5 | 3 | 0,1,2 |
6 | 4 | 0,1,2,3 |
7 | 4 | 0,1,2,3 |
8 | 4 | 0,1,3,4 |
9 | 4 | 0,1,3,4 |
10 | 5 | 0,1,2,4,5 |
11 | 5 | 0,1,2,4,5 |
12 | 5 | 0,1,3,5,6 |
13 | 5 | 0,1,3,5,6 |
14 | 6 | 0,1,2,4,6,7 |
15 | 6 | 0,1,2,4,6,7 |
16 | 6 | 0,1,3,5,7,8 |
17 | 6 | 0,1,3,5,7,8 |
18 | 7 | 0,1,2,4,6,8,9 |
19 | 7 | 0,1,2,4,6,8,9 |
20 | 7 | 0,1,3,5,7,9,10 |
21 | 7 | 0,1,3,5,7,9,10 |
22 | 8 | 0,1,2,4,6,8,10,11 |
23 | 8 | 0,1,2,4,6,8,10,11 |
24 | 8 | 0,1,3,5,7,9,11,12 |
25 | 8 | 0,1,3,5,7,9,11,12 |
26 | 8 | 0,1,3,4,9,10,12,13 |
27 | 8 | 0,1,3,4,9,10,12,13 |
28 | 9 | 0,1,3,5,7,9,11,13,14 |
29 | 9 | 0,1,3,5,7,9,11,13,14 |
30 | 9 | 0,1,3,4,9,10,12,14,15 |
31 | 9 | 0,1,3,4,9,10,12,14,15 |
32 | 9 | 0,1,2,5,8,11,14,15,16 |
33 | 9 | 0,1,2,5,8,11,14,15,16 |
34 | 10 | 0,1,3,4,9,10,12,14,16,17 |
35 | 10 | 0,1,3,4,9,10,12,14,16,17 |
36 | 10 | 0,1,3,5,6,12,13,15,17,18 |
37 | 10 | 0,1,3,5,6,12,13,15,17,18 |
38 | 10 | 0,1,3,5,6,13,14,16,18,19 |
39 | 10 | 0,1,3,5,6,13,14,16,18,19 |
40 | 10 | 0,1,3,4,9,11,16,17,19,20 |
41 | 10 | 0,1,3,4,9,11,16,17,19,20 |
42 | 11 | 0,1,3,5,6,13,14,16,18,20,21 |
43 | 11 | 0,1,3,5,6,13,14,16,18,20,21 |
44 | 11 | 0,1,3,4,6,11,13,18,19,21,22 |
45 | 11 | 0,1,3,4,6,11,13,18,19,21,22 |
46 | 11 | 0,1,2,3,7,11,15,19,21,22,24 |
47 | 11 | 0,1,2,3,7,11,15,19,21,22,24 |
48 | 12 | 0,1,3,5,7,8,16,17,19,21,23,24 |
49 | 12 | 0,1,3,5,7,8,16,17,19,21,23,24 |
50 | 12 | 0,1,3,5,7,8,17,18,20,22,24,25 |
51 | 12 | 0,1,3,5,7,8,17,18,20,22,24,25 |
52 | 12 | 0,1,3,5,6,12,13,20,21,23,25,26 |
53 | 12 | 0,1,3,5,6,12,13,20,21,23,25,26 |
54 | 12 | 0,1,3,5,6,13,14,21,22,24,26,27 |
55 | 12 | 0,1,3,5,6,13,14,21,22,24,26,27 |
56 | 13 | 0,1,3,4,8,9,11,20,21,23,25,27,28 |
57 | 13 | 0,1,3,4,8,9,11,20,21,23,25,27,28 |
58 | 13 | 0,1,3,5,6,13,14,21,22,24,26,28,29 |
59 | 13 | 0,1,3,5,6,13,14,21,22,24,26,28,29 |
60 | 13 | 0,1,3,5,6,13,15,17,24,25,27,29,30 |
61 | 13 | 0,1,3,5,6,13,15,17,24,25,27,29,30 |
62 | 13 | 0,2,3,5,9,13,17,21,25,26,28,30,31 |
63 | 13 | 0,1,2,3,7,11,15,19,23,27,29,30,32 |
64 | 13 | 0,1,3,4,9,11,16,21,23,28,29,31,32 |
65 | 13 | 0,1,3,4,9,11,16,21,23,28,29,31,32 |
The sequence 3, 5, 9, 13, 17, 21, 27, 33, 41, 47, 55, ... does not match anything previously in the Online Encyclopedia of Integer Sequences. Note the N=46, N=47 and N=63 lines in the table above, which are the first known examples where it is necessary for Q-cardinality-minimality that Q include a number greater than N/2 (shown blue).
Asymptotics: We can easily bound N(q) between two quadratics:
A least-squares fit of a quadratic in q=2,3,4,...,12 to the data N=3,5,9,...,55 yields N = 0.291*q2 + 1.139*q - 0.527 approximately, suggesting that perhaps the lower bound on N is closer to the truth than the upper bound, i.e. indicating the trivial 2-digit construction is actually not so bad.