Hi,
Here is a sequence of prime numbers:
3,5,11,17,47,257,510767,....
The algorithm used to build such sequence is :
We definie Primorial p noted #p=2*3*5*7*....*p
Start from U(0)=3 the first odd prime
Compute a(1)=int(sqrt(U(0)))+1=2
U(1)=U(0)+a(1) where a(1) is equal to primorial #2 = 2
U(1)=5
Compute a(2)=int(sqrt(U(1))+1=3
U(2)=U(1)+a(2)=5+#3=5+(2*3)=11
Compute a(3)=int(sqrt(11))+1=4 (4 is not prime so we use #3=2*3
U(3)=11+(2*3)=17
and so on
Are all the numbers of the sequence prime numbers?
The sequence is growing faster it is not easy to know.
Thank you for any help.
Are all the sequence numbers prime?
Moderators: gmalivuk, Moderators General, Prelates
Re: Are all the sequence numbers prime?
No. The next term in your sequence (after 510767) is
which has a factor of 965854931
Code: Select all
13802651106711802536344050306133362992649963656229914863058580142142610482430817949922104531639351381921564573865712490763569228376295661814770390189505137031692781919527713285374540164408571278055683171593020170233128086464775974520546386806644011091046992108509661969860784773011026549130272637
which has a factor of 965854931
 Xanthir
 My HERO!!!
 Posts: 5423
 Joined: Tue Feb 20, 2007 12:49 am UTC
 Location: The Googleplex
 Contact:
Re: Are all the sequence numbers prime?
I'm having a hard time figuring out what this sequence even *is*  is the a() function just "the primorial of the largest prime less than or equal to the sqrt of the previous sequence value"?
(defun fibs (n &optional (a 1) (b 1)) (take n (unfold '+ a b)))
Re: Are all the sequence numbers prime?
Xanthir wrote:I'm having a hard time figuring out what this sequence even *is*  is the a() function just "the primorial of the largest prime less than or equal to the sqrt of the previous sequence value"?
Yes.
If a() is prime (let us say q) then the primoiral is #q=2*3*5*3.....*q
If a() is not prime then the primorial is #q where q is the last prime < a()
Examples :
a()=10 then #q=2*3*5*7
a()=19 then #q=2*3*5*.....*17*19

 Posts: 126
 Joined: Wed May 09, 2012 6:02 pm UTC
Re: Are all the sequence numbers prime?
PM 2Ring wrote:No. The next term in your sequence (after 510767) isCode: Select all
13802651106711802536344050306133362992649963656229914863058580142142610482430817949922104531639351381921564573865712490763569228376295661814770390189505137031692781919527713285374540164408571278055683171593020170233128086464775974520546386806644011091046992108509661969860784773011026549130272637
which has a factor of 965854931
Which is to be expected.
The first few terms must be prime, simply because the method of construction rules out any factor below their square root:
5=2+3 cannot be divisible by 2 or 3
11=5+2x3 cannot be divisible by 2 or 3
17=11+2x3 cannot be divisible by 2 or 3
47=17+2x3x5 cannot be divisible by 2,3 or 5
And the following two terms are also fairly likely to be prime:
257=47+2x3x5x7 cannot be divisible by 2,3,5 or 7.
510767=257+2x3x5x7x11x13x17 cannot be divisible by 2,3,5,7,11,13 or 17.
A number in the 200's which isn't divisible by 2,3,5 or 7 has a 16/21 chance of being prime.
And a number in the 500000s which isn't divisible by anything up to 17 has a roughly 40% chance of being prime.
On the other hand, the next term is a 296digit number and the only factors which are ruled out are those under 719. So there's no reason at all to think it would be prime. And indeed, it isn't.
Who is online
Users browsing this forum: No registered users and 11 guests