subtract_with_carry_engine Class
The new home for Visual Studio documentation is Visual Studio 2017 Documentation on docs.microsoft.com.
The latest version of this topic can be found at subtract_with_carry_engine Class.
Generates a random sequence by the subtract-with-carry (lagged Fibonacci) algorithm.
template <class UIntType, size_t W, size_t S, size_t R> class subtract_with_carry_engine;
Parameters
UIntType
The unsigned integer result type. For possible types, see <random>.
W
Word size. Size of each word, in bits, of the state sequence. Precondition: 0 < W ≤ numeric_limits<UIntType>::digits
S
Short lag. Number of integer values. Precondition: 0 < S < R
R
Long lag. Determines recurrence in the series generated.
subtract_with_carry_engine::subtract_with_carry_engine | subtract_with_carry_engine::min | subtract_with_carry_engine::discard |
subtract_with_carry_engine::operator() | subtract_with_carry_engine::max | subtract_with_carry_engine::seed |
default_seed is a member constant, defined as 19780503u, used as the default parameter value for subtract_with_carry_engine::seed and the single value constructor. |
For more information about engine members, see <random>.
The substract_with_carry_engine template class is an improvement over the linear_congruential_engine. Neither for these engines is as fast or with as high quality results as the mersenne_twister_engine.
This engine produces values of a user-specified unsigned integral type using the recurrence relation ( period) x(i) = (x(i - R) - x(i - S) - cy(i - 1)) mod M, where cy(i) has the value 1 if x(i - S) - x(i - R) - cy(i - 1) < 0, otherwise 0, and M has the value 2W. The engine's state is a carry indicator plus R values. These values consist of the last R values returned if operator() has been called at least R times, otherwise the N values that have been returned and the last R - N values of the seed.
The template argument UIntType must be large enough to hold values up to M - 1.
Although you can construct a generator from this engine directly, you can also use one of these predefined typedefs:
ranlux24_base: Used as a base for ranlux24.typedef subtract_with_carry_engine<unsigned int, 24, 10, 24> ranlux24_base;
ranlux48_base: Used as a base for ranlux48.typedef subtract_with_carry_engine<unsigned long long, 48, 5, 12> ranlux48_base;
For detailed information about the subract with carry engine algorithm, see the Wikipedia article Lagged Fibonacci generator.
Header: <random>
Namespace: std