|
|
A236183
|
|
Primes p such that for all primes q dividing p-1 every residue mod p is the sum of two q-th powers mod p.
|
|
1
|
|
|
2, 3, 5, 13, 17, 19, 37, 73, 97, 101, 109, 151, 163, 181, 193, 197, 211, 241, 251, 257, 271, 281, 337, 379, 397, 401, 421, 433, 449, 487, 491, 541, 577, 601, 631, 641, 661, 673, 701, 727, 751, 757, 769, 811, 881, 883, 991, 1009
(list;
graph;
refs;
listen;
history;
text;
internal format)
|
|
|
OFFSET
|
1,1
|
|
COMMENTS
|
Alternative definition: increasing list of primes p such that for ALL primes q, every residue mod p is the sum of two q-th powers mod p (if q does not divides p-1 then every residue mod p is a q-th power, so only the case q divides p-1 is not trivial).
Related to the conjecture:
For every prime q there are only finitely many primes p such that not every residue mod p is the sum of two q-th powers mod p.
|
|
LINKS
|
|
|
EXAMPLE
|
p=7 is not in the list because q=3 divides p-1 and the only residues mod 7 which are the sum of two cubic residues mod 7 are: 0,1,-1,2,-2.
To check that p=13 is in the sequence, since 2 and 3 are the only primes dividing p-1=12, we only need to see that every residue mod 13 is both the sum of two quadratic residues mod 13 and the sum of two cubic residues mod 13.
|
|
PROG
|
(Sage)
for p in prime_range(a, b):
c=1
C=GF(p)
u=combinations_with_replacement(C, 2)
v=[x for x in u]
for q in prime_divisors(p-1):
w=(k[0]^q+k[1]^q for k in v)
s=set(w)
l=len(s)
if l!=p:
c=0
break
if c==1:
print(p)
(PARI) is(p)=if(!isprime(p), return(0)); my(f=factor(p-1)[, 1], v, u); for(i=1, #f, u=vector(p); v=vector(p, j, j^f[i]%p); for(j=1, p, for(k=j, p, u[(v[j]+v[k])%p+1]=1)); if(!vecmin(u), return(0))); 1 \\ Charles R Greathouse IV, Jan 23 2014
|
|
CROSSREFS
|
|
|
KEYWORD
|
nonn
|
|
AUTHOR
|
|
|
STATUS
|
approved
|
|
|
|