Method half lowest

include/half.hpp:2327–2407

	/// Numeric limits for half-precision floats.
	/// **See also:** Documentation for [std::numeric_limits](https://en.cppreference.com/w/cpp/types/numeric_limits)
	template<> class numeric_limits<half_float::half>
	{
	public:
		/// Is template specialization.
		static HALF_CONSTEXPR_CONST bool is_specialized = true;

		/// Supports signed values.
		static HALF_CONSTEXPR_CONST bool is_signed = true;

		/// Is not an integer type.
		static HALF_CONSTEXPR_CONST bool is_integer = false;

		/// Is not exact.
		static HALF_CONSTEXPR_CONST bool is_exact = false;

		/// Doesn't provide modulo arithmetic.
		static HALF_CONSTEXPR_CONST bool is_modulo = false;

		/// Has a finite set of values.
		static HALF_CONSTEXPR_CONST bool is_bounded = true;

		/// IEEE conformant.
		static HALF_CONSTEXPR_CONST bool is_iec559 = true;

		/// Supports infinity.
		static HALF_CONSTEXPR_CONST bool has_infinity = true;

		/// Supports quiet NaNs.
		static HALF_CONSTEXPR_CONST bool has_quiet_NaN = true;

		/// Supports signaling NaNs.
		static HALF_CONSTEXPR_CONST bool has_signaling_NaN = true;

		/// Supports subnormal values.
		static HALF_CONSTEXPR_CONST float_denorm_style has_denorm = denorm_present;

		/// Supports no denormalization detection.
		static HALF_CONSTEXPR_CONST bool has_denorm_loss = false;

	#if HALF_ERRHANDLING_THROWS
		static HALF_CONSTEXPR_CONST bool traps = true;
	#else
		/// Traps only if [HALF_ERRHANDLING_THROW_...](\ref HALF_ERRHANDLING_THROW_INVALID) is acitvated.
		static HALF_CONSTEXPR_CONST bool traps = false;
	#endif

		/// Does not support no pre-rounding underflow detection.
		static HALF_CONSTEXPR_CONST bool tinyness_before = false;

		/// Rounding mode.
		static HALF_CONSTEXPR_CONST float_round_style round_style = half_float::half::round_style;

		/// Significant digits.
		static HALF_CONSTEXPR_CONST int digits = 11;

		/// Significant decimal digits.
		static HALF_CONSTEXPR_CONST int digits10 = 3;