Author |
Topic: prime triplet (Read 1379 times) |
|
towr
wu::riddles Moderator
Uberpuzzler
Some people are average, some are just mean.
Gender: 
Posts: 13730
|
 |
Re: prime triplet
« Reply #1 on: May 15th, 2013, 1:08pm » |
 Quote  Modify
|
For p=191 (in both sequences) the sum of the triple is not prime ( 7 divides 3*191+8 and 11 divides 3*191+10).
And many, many other aren't either.
|
|
 IP Logged |
Wikipedia, Google, Mathworld, Integer sequence DB
|
|
|
Christine
Full Member
Posts: 159
|
|
Re: prime triplet
« Reply #2 on: May 15th, 2013, 1:55pm » |
Quote Modify
|
Is there an analytical solution that explains for which condition(s) the sum would prime?
|
|
IP Logged |
|
|
|
towr
wu::riddles Moderator
Uberpuzzler
Some people are average, some are just mean.
Gender:
Posts: 13730
|
|
Re: prime triplet
« Reply #3 on: May 15th, 2013, 11:15pm » |
Quote Modify
|
Maybe if the Riemann hypothesis is true. 
But as far as I'm concerned it's arbitrary.
|
|
IP Logged |
Wikipedia, Google, Mathworld, Integer sequence DB
|
|
|
JohanC
Senior Riddler
Posts: 460
|
|
Re: prime triplet
« Reply #4 on: May 23rd, 2013, 2:01pm » |
Quote Modify
|
The sums of such triplets are neither divisible by 2 nor by 3. Therefore, small triplets have a slightly higher probability of having a prime sum.
The larger the triplet, the lower this influence.
|
|
IP Logged |
|
|
|
whizen
Newbie
hidden:
Gender:
Posts: 13
|
|
Re: prime triplet
« Reply #5 on: May 29th, 2013, 3:47pm » |
Quote Modify
|
First non-prime prime triplet sum for
p + (p+2) + (p+6)
is at p = 107, sum = 329
( 107, 191, 227, 461, 821, 881 are other primes under 1000 which break this rule)
p + (p+4) + (p+6) is not prime
for p = 13, sum = 49
(13, 37, 97, 103, 193, 223, 277, 307, 613, 823, 853, 877 are bad primes for this series)
There are 189 and 196 such primes below 100k respectively.
It does seem to be the case that the latter primes are more than the former.
Upto 1 million: 1087 vs 1151, diff 64
Upto 10 million: 6989 vs 7119, diff 130
It will be interesting to note if the non-primes in one of the categories is strictly more than the other, in an asymptotic sense.
Suboptimal python code to do this...
Code:
isPrime = lambda x : x%2 == 1 and all((x % z for z in xrange(3, int(pow(x, 0.5)) + 1, 2)))
def primeTriples(d1, d2, lim=10000):
nonpts = []
d3 = d1 + d2
for x in range(3, lim, 2):
if isPrime(x) and isPrime(x + d1) and isPrime(x + d2):
if not isPrime(3*x + d3):
nonpts.append(x)
return nonpts
def printOutput(lim):
pt26 = primeTriples(2, 6, lim)
pt46 = primeTriples(4, 6, lim)
l26, l46 = len(pt26), len(pt46)
print lim, l26, l46, l46 - l26
[printOutput(int(lim)) for lim in (1E3, 1E4, 1E5)]
|
|
|
| « Last Edit: May 29th, 2013, 3:47pm by whizen » |
IP Logged |
|
|
|
|
RIDDLES SITE
WRITE MATH!
Home 
Help 
Search 
Members 
Login 
Register
Reply
Notify of replies
Send Topic
Print