[R] precision in operations

Douglas Bates bates at stat.wisc.edu
Fri Nov 7 14:39:29 CET 2003

Jason Turner <jasont at indigoindustrial.co.nz> writes:

> javier garcia - CEBAS wrote:
> > Hi all;
> > could you remind me what is the function to change the precision of
> > the operations done in R? I can't remember nor find it.
> >
> There isn't one.  R does all its calculations with system defined
> double-precision or integers (on most platforms where R is used, these
> are both 32 bits).  See help(is.single)

Usually the double-precision representation is 64 bits in memory and,
for the IA-32 processors (Intel and AMD x86 family), 80 bits in the
floating point registers.

Javier may be looking for

options(digits = nn)

where the default value of nn is 7.
If you want more precision in printed results try setting

options(digits = 12)

More information about the R-help mailing list