Contents

bit_ops

Usage:

a %| b, a %^ b, a %& b and %! a, where a and b are REAL or structures
  with REAL components with integer elements >= 0 and <= 4294967295 =
  2^32-1 nbits(x)

Keywords: operations, glm, missing values

There are 4 operators for working with integers considered as the sets
of 32 bits specified by their binary representations.

Bit Operation  Precedence  Meaning
    a %| b         1       Bitwise Or (OR)
    a %^ b         2       Bitwise Exclusive Or (XOR)
    a %& b         3       Bitwise And (AND)
      %!a          4       Bitwise Complement (COMPL)

When an operand x is not an integer or x < 0 or x > 4294967295 = 2^32-1,
the result of any of these operators is MISSING.

For '%&', a bit of the result is 1 if and only if the corresponding bits
in the operands are both 1.
  Example: 25 %& 19 is 17 because 11001b AND 10011b is 10001b

For '%|', a bit of the result is 1 if and only if at least 1 of the
corresponding bits in the operands is 1.
  Example: 25 %| 19 is 27  because 11001b OR 10011b is 11011b

For '%^', a bit of the result is 1 if and only if exactly 1 of the
corresponding bits in the operands are 1, that is, if the corresponding
bits differ.
  Example: 25 %^ 19 is 10  because 11001b XOR 10011b is 01010b

Operator '%!' operates on the immediately following variable considered
as a collection of 32 bits, changing 1's to 0's and 0's to 1's.
  Examples:
       %! 25 is 4294967270 since COMPL(0000000000000000000000000011001b)
          is 11111111111111111111111111100110b
       %! 0  is 4294967295 since COMPL(0000000000000000000000000000000b)
          is 11111111111111111111111111111111b

When an operand is LOGICAL, it is treated as having value 0 (F) or 1
(T).  The result is always REAL.

When any operand is MISSING, so is the result.

Bit operators were introduced to be useful with the output of
modelinfo(bitmodel:T).  For example, 2^(i-1) %& modelinfo(bitmodel:T)[j]
is non-zero if and only if the j-th term of the model contains the i-th
factor or variate (assuming i <= 32).

The operators were listed above in increasing order of precedence.
Moreover, they have lower precedence that all arithmetic, comparison, or
logical operators which means they are evaluated after all such
operators.

  Examples:
    Expression                 Interpretation              Value
    17 %| 29 %^ 91 %& 11       17 %| (29 %^ (91 %& 11))    23
    %!21 %| 97 %& %! 33        (%!21) %| (97 %& (%!33))    4294967274
    1 %& 3 + 4                 1 %& (3+4)                  1
    3 %^ 5 != 6                3 %^ (5 != 6)               2
    1 %| 2 == 3                1 %| (2 == 3)               1
    %!0 == 4294967295          %!(0 == 4294967295)         4294967295

To understand the last three examples, note that 5 != 6 is True and is
interpreted as 1, and that 2 == 3 and 0 == 4294967295 are both False and
are interpreted as 0. See topics 'arithmetic' and 'logic'.

See topic 'arithmetic' for a description of the "shape" of the result
when operands are not scalars.

See also modelinfo(), nbits().