Goldbach's conjecture: every even number greater than 4 can be written as the sum of two odd primes.
Let's start by creating a list of odd primes up to N.
|
[3, 5, 7, 11, 13, 17, 19] [3, 5, 7, 11, 13, 17, 19] |
We could do better. For a start, we could increment p by 2 instead of 1 each time.
|
[3, 5, 7, 11, 13, 17, 19] [3, 5, 7, 11, 13, 17, 19] |
We can do even better using the next_prime() method or function.
|
1507984 1507984 |
38.84908 38.84908 |
|
![]() |
![]() |
![]() |
Let's figure out which points on the plot are champions, i.e. are higher than any point on its left.
|
|
52 52 |
[[6, 1], [8, 2], [10, 3], [16, 4], [22, 5], [24, 6], [34, 7], [36, 8], [48, 10], [60, 12], [78, 14], [84, 16], [90, 18], [114, 20], [120, 24], [168, 26], [180, 28], [210, 38], [300, 42], [330, 48], [390, 54], [420, 60], [510, 64], [630, 82], [780, 88], [840, 102], [990, 104], [1050, 114], [1140, 116], [1260, 136], [1470, 146], [1650, 152], [1680, 166], [1890, 182], [2100, 194], [2310, 228], [2730, 256], [3150, 276], [3570, 308], [3990, 326], [4200, 330], [4410, 342], [4620, 380], [5250, 396], [5460, 436], [6090, 444], [6510, 482], [6930, 536], [7980, 548], [8190, 584], [9030, 606], [9240, 658]] [[6, 1], [8, 2], [10, 3], [16, 4], [22, 5], [24, 6], [34, 7], [36, 8], [48, 10], [60, 12], [78, 14], [84, 16], [90, 18], [114, 20], [120, 24], [168, 26], [180, 28], [210, 38], [300, 42], [330, 48], [390, 54], [420, 60], [510, 64], [630, 82], [780, 88], [840, 102], [990, 104], [1050, 114], [1140, 116], [1260, 136], [1470, 146], [1650, 152], [1680, 166], [1890, 182], [2100, 194], [2310, 228], [2730, 256], [3150, 276], [3570, 308], [3990, 326], [4200, 330], [4410, 342], [4620, 380], [5250, 396], [5460, 436], [6090, 444], [6510, 482], [6930, 536], [7980, 548], [8190, 584], [9030, 606], [9240, 658]] |
![]() |
It looks like being divisible by small primes 3,5,7 etc has an impact on the number of representations. How much of an effect does divisibility by 3 have?
|
![]() |
![]() |
Let's redo the plots, but divide the height by 2 when n is divisible by 3.
|
![]() |
![]() |
Let's try something similar for divisibility by 5
|
|
![]() |
It appears experimentally that divisibility by 3 increases the number of representations by a factor of 2, and divisibility by 5 increases by a factor of 4/3.
Let's combine them.
|
![]() |
Let's take a wild guess that the correction for divisibilty by $p$ turns out to be $(p-1)/(p-2)$, and that the corrections act roughly independently:
define $hl(n)$ by
\[ hl(n)=\prod_{p|n} \frac{p-1}{p-2} \]
over odd primes dividing $n$.
|
16/5 16/5 |
<type 'sage.rings.integer.Integer'> <type 'sage.rings.integer.Integer'> |
|
![]() |
|
|
![]() |
Let's throw in some upper and lower curves on either side, to see if they look about the right distance away from the curve.
![]() |
Let's try Li(x) instead of x/log(x).
|
|
![]() |
Traceback (click to the left of this block for traceback) ... TypeError: 'sage.rings.integer.Integer' object is not callable Traceback (most recent call last): File "<stdin>", line 1, in <module> File "_sage_input_48.py", line 10, in <module> exec compile(u'open("___code___.py","w").write("# -*- coding: utf-8 -*-\\n" + _support_.preparse_worksheet_cell(base64.b64decode("ZigxMCk="),globals())+"\\n"); execfile(os.path.abspath("___code___.py")) File "", line 1, in <module> File "/tmp/tmpoCHWmL/___code___.py", line 3, in <module> exec compile(u'f(_sage_const_10 ) File "", line 1, in <module> File "/tmp/tmpKFp1ox/___code___.py", line 4, in f return(Li(t)*_sage_const_2 (Li(_sage_const_2 *t)-Li(t))/t ) TypeError: 'sage.rings.integer.Integer' object is not callable |
|