|
|
A063990
|
|
Amicable numbers.
|
|
117
|
|
|
220, 284, 1184, 1210, 2620, 2924, 5020, 5564, 6232, 6368, 10744, 10856, 12285, 14595, 17296, 18416, 63020, 66928, 66992, 67095, 69615, 71145, 76084, 79750, 87633, 88730, 100485, 122265, 122368, 123152, 124155, 139815, 141664, 142310
(list;
graph;
refs;
listen;
history;
text;
internal format)
|
|
|
OFFSET
|
1,1
|
|
COMMENTS
|
A pair of numbers x and y is called amicable if the sum of the proper divisors of either one is equal to the other. The smallest pair is x = 220, y = 284.
The sequence lists the amicable numbers in increasing order. Note that the pairs x, y are not adjacent to each other in the list. See also A002025 for the x's, A002046 for the y's.
Theorem: If the three numbers p = 3*(2^(n-1)) - 1, q = 3*(2^n) - 1 and r = 9*(2^(2n-1)) - 1 are all prime where n >= 2, then p*q*(2^n) and r*(2^n) are amicable numbers. This 9th century theorem is due to Thabit ibn Kurrah (see for example, the History of Mathematics by David M. Burton, 6th ed., p. 510). - Mohammad K. Azarian, May 19 2008
The first time a pair ordered by its first element is not adjacent is x = 63020, y = 76084 which correspond to a(17) and a(23), respectively. - Omar E. Pol, Jun 22 2015
Sierpiński (1964), page 176, mentions Erdős's work on the number of pairs of amicable numbers <= x. - N. J. A. Sloane, Dec 27 2017
Kanold (1954) proved that the asymptotic upper density of amicable numbers is < 0.204 and Erdős (1955) proved that it is 0. - Amiram Eldar, Feb 13 2021
|
|
REFERENCES
|
Scott T. Cohen, Mathematical Buds, Ed. Harry D. Ruderman, Vol. 1, Chap. VIII, pp. 103-126, Mu Alpha Theta, 1984.
Clifford A. Pickover, The Math Book, Sterling, NY, 2009; see p. 90.
Wacław Sierpiński, Elementary Theory of Numbers, Panst. Wyd. Nauk, Warsaw, 1964.
David Wells, The Penguin Dictionary of Curious and Interesting Numbers, Penguin Books, 1987, pp. 145-147.
|
|
LINKS
|
Mariano García, Jan Munch Pedersen, Herman te Riele, Amicable pairs, a survey, Report MAS-R0307, Centrum Wiskunde & Informatica.
|
|
FORMULA
|
Pomerance shows that there are at most x/exp(sqrt(log x log log log x)/(2 + o(1))) terms up to x for sufficiently large x. - Charles R Greathouse IV, Jul 21 2015
Sum_{n>=1} 1/a(n) is in the interval (0.0119841556, 215) (Nguyen and Pomerance, 2019; an upper bound 6.56*10^8 was given by Bayless and Klyve, 2011). - Amiram Eldar, Oct 15 2020
|
|
MAPLE
|
F:= proc(t) option remember; numtheory:-sigma(t)-t end proc:
select(t -> F(t) <> t and F(F(t))=t, [$1.. 200000]); # Robert Israel, Jun 22 2015
|
|
MATHEMATICA
|
s[n_] := DivisorSigma[1, n] - n; AmicableNumberQ[n_] := If[Nest[s, n, 2] == n && ! s[n] == n, True, False]; Select[Range[10^6], AmicableNumberQ[ # ] &] (* Ant King, Jan 02 2007 *)
|
|
PROG
|
(PARI) aliquot(n)=sigma(n)-n
isA063990(n)={local(a); a=aliquot(n); a<>n && aliquot(a)==n} \\ Michael B. Porter, Apr 13 2010
(Python)
from sympy import divisors
A063990 = [n for n in range(1, 10**5) if sum(divisors(n))-2*n and not sum(divisors(sum(divisors(n))-n))-sum(divisors(n))] # Chai Wah Wu, Aug 14 2014
|
|
CROSSREFS
|
A180164 (gives for each pair (x, y) the value x+y = sigma(x)+sigma(y)).
|
|
KEYWORD
|
nonn
|
|
AUTHOR
|
|
|
STATUS
|
approved
|
|
|
|