.NET port of the QD library implementing the double-double and quad-double technique for achieving almost 128-bit and 256-bit floating point precision types.
See original paper by David H. Bailey Yozo Hida and Xiaoye S. Li for mathematical details. An unofficial copy of the original C++/Fortran implementation is available e.g. here.
Note that the port is in it's early stages so there may be some bugs.
dotnet add package NetQD
- Comprehensive set of unit tests that check that nothing has been broken during porting
- Add more math functions (Exp, Log, Pow...)
- Fit dotnet IFormatProvider into parsing and printing code
Pull Requests are welcome.