cp's OEIS Frontend

This is a front-end for the Online Encyclopedia of Integer Sequences, made by Christian Perfect. The idea is to provide OEIS entries in non-ancient HTML, and then to think about how they're presented visually. The source code is on GitHub.

Showing 1-2 of 2 results.

A215672 Fermat pseudoprimes to base 2 with three prime factors.

Original entry on oeis.org

561, 645, 1105, 1729, 1905, 2465, 2821, 4371, 6601, 8481, 8911, 10585, 12801, 13741, 13981, 15841, 16705, 25761, 29341, 30121, 30889, 33153, 34945, 41665, 46657, 52633, 57421, 68101, 74665, 83665, 87249, 88561, 91001, 93961, 113201, 115921, 121465, 137149
Offset: 1

Views

Author

Marius Coman, Aug 20 2012

Keywords

Comments

Fermat pseudoprimes to base 2 are also called Poulet numbers.
Most of the terms shown can be written in one of the following two ways:
(1) p*(p*(n + 1) - n)*(p*(m + 1) - m);
(2) p*(p*n - (n + 1))*(p*m - (m + 1)),
where p is the smallest of the three prime factors and n, m natural numbers.
Exempli gratia for Poulet numbers from the first category:
10585 = 5*29*73 = 5*(5*7 - 6)*(5*18 - 17);
13741 = 7*13*151 = 7*(7*2 - 1)*(7*25 - 24);
13981 = 11*31*41 = 11*(11*3 - 2)*(11*4 - 3);
29341 = 13*37*61 = 13*(13*3 - 2)*(13*5 - 4);
137149 = 23*67*89 = 23*(23*3 - 2)*(23*4 - 3).
Exempli gratia for Poulet numbers from the second category:
6601 = 7*23*41 = 7*(7*4 - 5)*(7*7 - 8).
Note: from the numbers from the sequence above, just the numbers 30889, 88561 and 91001 can't be written in one of the two ways.
What these three numbers have in common: they all have a prime divisor q of the form 30*k + 23 (i.e. 23, 53, 83) and can be written as q*((r + 1)*q - r), where r is a natural number.
Conjecture: Any Poulet number P with three or more prime divisors has at least one prime divisor q for that can be written as P = q*((r + 1)*q - r), where r is a natural number.
Note: it can be proved that a Carmichael number can be written this way for any of its prime divisors - see the sequence A213812.
Note: there are also many Poulet numbers with two prime divisors that can be written this way, but here are few exceptions: 7957, 23377, 42799, 49981, 60787.
The conjecture fails for a(80) = 617093 = 43 * 113 * 127. - Charles R Greathouse IV, Dec 07 2014
First differs from A074380 at n=56. - Amiram Eldar, Jun 28 2019

Crossrefs

The even terms form A296117.

Programs

  • Mathematica
    Select[Range[10^5], PrimeNu[#] == 3 && PowerMod[2, (# - 1), #] == 1 &] (* Amiram Eldar, Jun 28 2019 *)
  • PARI
    is(n)=Mod(2,n)^n==2 && bigomega(n)==3 \\ Charles R Greathouse IV, Dec 07 2014

A306487 Poulet numbers which are not super-Poulet numbers.

Original entry on oeis.org

561, 645, 1105, 1729, 1905, 2465, 2821, 4371, 6601, 8481, 8911, 10585, 11305, 12801, 13741, 13981, 15841, 16705, 18705, 23001, 25761, 29341, 30121, 30889, 33153, 34945, 39865, 41041, 41665, 46657, 52633, 55245, 57421, 62745, 63973, 68101, 72885, 74665, 75361
Offset: 1

Views

Author

Amiram Eldar, Feb 18 2019

Keywords

Comments

Subsequence of A080747 from which this differs for the first time at n=78, with A080747(78) = 294409, a term not present here.
Is this sequence infinite?
According to Sierpinski there are infinitely many Poulet numbers which are not super-Poulet numbers. But his definition of Poulet numbers includes the even pseudoprimes to base 2 (A006935), and the proof is based on the infinitude of this sequence and that super-Poulet numbers are never even.

Examples

			561 is in the sequence because 2^561 % 561 == 2 but 33|561 and 2^33 % 33 = 8 <> 2. - _David A. Corneth_, Feb 28 2019
		

References

  • W. Sierpinski, Elementary Theory of Numbers, ed. A. Schinzel, North-Holland Mathematical Library (2nd ed.), Amsterdam: North Holland, 1988, Chapter V, p. 234, Exercise 1.

Crossrefs

Cf. A215672 (differs from a(13) = 11305 on, which is not in A215672).

Programs

  • Mathematica
    Select[Select[Range[3, 100000, 2], !PrimeQ[ # ] && PowerMod[2, (# - 1), # ] == 1 &], Union[PowerMod[2, Rest[Divisors[#]], #]] != {2}& ]
  • PARI
    is_A001567(n) = {Mod(2, n)^(n-1)==1 && !isprime(n) && n>1}; \\ From A001567 by M. F. Hasler
    is_A050217(n) = if(isprime(n), 0, fordiv(n, d, if(Mod(2, d)^d!=2, return(0))); (n>1)); \\ After Charles R Greathouse IV's Aug 27 2016 PARI-program in A050217.
    is_A306487(n) = (is_A001567(n) && !is_A050217(n)); \\ (Probably could be reduced to a simpler program). - Antti Karttunen, Feb 28 2019
    
  • PARI
    is(n) = {if(isprime(n) || n < 2 || n%2 == 0, return(0)); if(Mod(2, n)^n!=2, return(0) , d = divisors(n); for(i = 1, #d-1, if(Mod(2, d[i])^d[i]!=2, return(1) ) ) ); 0 } \\ David A. Corneth, Feb 28 2019
Showing 1-2 of 2 results.