OFFSET
1,1
COMMENTS
It is interesting that the number 2 occurs deep into the sequence indicating a twin prime pair. It is reasonable to ask if this will ever occur again. Similarly, the analogous sequence A074383, "Difference between (1+10^n)-th and (10^n)-th primes" has 2 occurring shallow into the sequence. It is reasonable to ask if the number 2 will ever occur again in that sequence. The link provides an excellent algorithm, primex(n), that I developed to find the n-th prime using Gram's approximation of Riemann's approximation R(x) for Pi(x). Primex(n) will give about n/2 exact digits for prime(n). For A006988 (18), primex(18) is 44211790234127235469.62904554...This is only as good as R (x) but nevertheless is superior to the exact formulas out there from a practical stand point. If we apply the code gpx(n) = for(x=1,n,y=nextprime(primex(10^x))-nextprime (primex(10^x-1));print1(floor(y)",")), we will get the erratic concoction 2,0,8,14,22,28,26,0,72,18,22,0,0,0,0,0,32,0,80,78,60,0 as an analytical counterpart of the sequence given.
LINKS
C. Hilliard, Nth prime approx [broken link].
EXAMPLE
CROSSREFS
KEYWORD
nonn,more
AUTHOR
Cino Hilliard, Jun 04 2007
EXTENSIONS
a(19) from Max Alekseyev, May 13 2009
a(20) from Max Alekseyev, May 30 2013
a(21),a(22) from Max Alekseyev, Dec 04 2014
a(23)-a(24) from Chai Wah Wu using terms in A006988, Sep 18 2018
STATUS
approved