MCPcopy Create free account
hub / github.com/Open-GD/OpenGD / compute_error_scaled

Function compute_error_scaled

Source/external/fast_float.h:1695–1703  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

1693// for significant digits already multiplied by 10 ** q.
1694template <typename binary>
1695fastfloat_really_inline FASTFLOAT_CONSTEXPR14
1696adjusted_mantissa compute_error_scaled(int64_t q, uint64_t w, int lz) noexcept {
1697 int hilz = int(w >> 63) ^ 1;
1698 adjusted_mantissa answer;
1699 answer.mantissa = w << hilz;
1700 int bias = binary::mantissa_explicit_bits() - binary::minimum_exponent();
1701 answer.power2 = int32_t(detail::power(int32_t(q)) + bias - hilz - lz - 62 + invalid_am_bias);
1702 return answer;
1703}
1704
1705// w * 10 ** q, without rounding the representation up.
1706// the power2 in the exponent will be adjusted by invalid_am_bias.

Callers

nothing calls this directly

Calls 1

powerFunction · 0.85

Tested by

no test coverage detected