OFFSET
0,4
COMMENTS
These integers are generated as a subset of interesting "areas" in typical microprocessor designs often based on 8, 16, 24, 32, 40, 56, 64 or 80 bits of precision and includes those areas where the sign representation could cause some design error. The numbers -1, 0, 1 represent a special case since they should and will be represented by a higher order number in the limited number space of N bits.
LINKS
Index entries for linear recurrences with constant coefficients, signature (-1,-1,0,0,0,16,16,16).
FORMULA
Union of integers: -1, 0, 1, (2^((4n)-1)) +/- { 0, 1 }, (2^(4n)) +/- { 0, 1 }
From Chai Wah Wu, Dec 23 2019: (Start)
a(n) = - a(n-1) - a(n-2) + 16*a(n-6) + 16*a(n-7) + 16*a(n-8) for n > 10.
G.f.: (-16*x^10 - 32*x^9 - 48*x^8 - 56*x^7 - 48*x^6 - 24*x^5 - 16*x^4 - 8*x^3 + x + 1)/((4*x^3 - 1)*(4*x^3 + 1)*(x^2 + x + 1)). (End)
MATHEMATICA
Join[{-1, 0, 1}, Sort[ Flatten[ NestList[{# - 1, # + 1} &, Flatten[ Table[{2^(4n - 1), 2^(4n)}, {n, 6}]], 1]]]] (* Robert G. Wilson v, Jun 14 2005 *)
CROSSREFS
KEYWORD
sign,easy
AUTHOR
Henrik Lundquist (sploinker(AT)sploink.dk), Jun 13 2005
EXTENSIONS
Edited by Robert G. Wilson v, Jun 14 2005
STATUS
approved