On the definition of ulp(x)
暂无分享,去创建一个
Function ulp (acronym for unit in the last place) is frequently used for expressing errors in floating-point computations. We present several previously suggested definitions of that function, and analyse some of their properties.
[1] Michael L. Overton,et al. Numerical Computing with IEEE Floating Point Arithmetic , 2001 .
[2] Peter W. Markstein,et al. IA-64 and elementary functions - speed and precision , 2000 .
[3] Ansi Ieee,et al. IEEE Standard for Binary Floating Point Arithmetic , 1985 .
[4] Jean-Michel Muller,et al. Elementary Functions: Algorithms and Implementation , 1997 .
[5] John Harrison,et al. A Machine-Checked Theory of Floating Point Arithmetic , 1999, TPHOLs.