View Single Post
Old 2015-09-16, 14:09   #1
fivemack
(loop (#_fork))
 
fivemack's Avatar
 
Feb 2006
Cambridge, England

2·29·109 Posts
Default On polynomials without roots modulo small p

A friend of mine observed that the centred hexagonal numbers (3x^2+3x+1) never looked 'nice'. This is true, and the reason is that they're never divisible by 2, 3, 5 or 11.

So I've been looking for polynomials which take only values not divisible by small primes. You can construct these by the Chinese Remainder Theorem, but I'm more interested in the asymptotics of where they turn up in nature for polynomials with small coefficients.

For quadratics, going up by minimum-value-of-maximum-coefficient:

Code:
a2  a1  a0  smallest prime with root mod p
48   42   73      53
81  141  -139   59
178   74  179    67
358  222  331  73
786  912  -811 83
2106  546 3037  89
3840 1290 2089  97
For cubics
Code:
a3  a2  a1  a0   p_min
12  3  33    -1  37
16  9  36    -1  41
4  48 -10   47  47
1  49 -52    1  53
60 52 -10   71  71
6  44 164 -131  79
For monic quadratics
Code:
1  1415   -673   67
1  6663  -1361   71
1  8433  -8171   73
1 12431  -6079   83
1 12787  -3187  101
1 16159 -10093  107
1 41569 -19993  127
1 97195 -93463  131
1 139831 -124513  137
1 221601 -204983  149
For quartics
Code:
2 -4 -2 4 -1  23
4 0 -6 0 1  31
3 -6 -4 7 1  41
5 6 18 15 -1  43
9 13 19 13 -1  53
29 0 -29 0 -1  61
4 22 -45 19 -1  83
3 0 85 0 1  89
7 -14 -90 97 -1 103
29 0 37 0 107  107
61 -122 -41 102 109  109
For ax^4+bx^2+c
Code:
4 0 -6 0 1  31
19 0 19 0 -1  37
21 0 -21 0 -1  41
25 0 -27 0 1  47
29 0 -29 0 -1  61
61 0 53 0 -1  83
3 0 85 0 1  89
29 0 37 0 107  107
19 0 -139 0 -137  113
156 0 -174 0 -131  131
9 0 259 0 1  139
486 0 -6 0 -223  149

Last fiddled with by fivemack on 2015-09-17 at 08:19 Reason: add some quadratic-in-x^2 cases
fivemack is offline   Reply With Quote