[GiNaC-list] Change precision of numeric

Richard B. Kreckel kreckel at ginac.de
Sat Nov 3 00:56:07 CET 2007

Timo Schlüßler wrote:
> When setting "Digits = 2" for the 1.0 numeric, the result (of multiply)
> stil has 4 decimal places ...

Actually, it has even more decimal digits but the printing routine 
truncates fractional trailing zeros.

The CLN preprocessor macro intDsize determines the granularity of binary 
digits. E.g. on x86, the precision is always a multiple of 32 bit. On 
amd64, it is a multiple of 64 bit.

Richard B. Kreckel

More information about the GiNaC-list mailing list