IEEE Std 1788-2015 pdf download – IEEE Standard for Interval Arithmetic.
domain: (af) For a function with arguments taken from some set, the domain comprises those points in the set at which the function has a value. The domain of a point operation is part of its definition. E.g., the (point) operation of division x/y, in this standard, has arguments (x, y) in R 2 , and its domain is { (x, y) ∈ R 2 | y 6 = 0 }. See also natural domain. elementary function: (af) Synonymous with arithmetic operation. enclose: (af) Synonymous with contain. exception, exceptional condition: (af) An exception is an event that occurs, and may be signaled, when an operation on some particular operands has no outcome suitable for every reasonable application. A signaled exception is handled in a language- or implementation-defined way. NOTE—Details in 7.5 . An exceptional condition is one of the events handled by the decoration system; it is not an exception. NOTE—Details in Clause 8 . (s) In the set-based standard, exceptions may occur in interval constructors and in the intervalPart oper- ation. NOTE—Details in 12.12.11 and in 12.1.3, 13.4 respectively . expression: (af) A symbolic form used to define a function. As used in this standard, it is a mathematical construct extracted from a section of run-time dataflow of a program, from which statements about the behavior ofthis function may be deduced by using the FTIA or FTDIA ofa given flavor. It may be represented by a computational graph, a code list or a normal algebraic expression. An arithmetic expression is one whose operations are all arithmetic operations. NOTE—Details in Clause 6 . explicit type: (s) An interval type that has a uniquely defined interval hull operation. floating-point format: (af) A number format like those of IEEE 754, whose numbers have the form x = s × d 0 .d 1 . . . d p × b e where integer b ≥ 2 is the fixed radix, integer p ≥ 0 is the fixed precision, s = ±1 is the sign, d 0 .d 1 . . . d p (a radix-b fraction) is the significand or mantissa, and e is an integer in a fixed exponent range emin ≤ e ≤ emax.