Re: calculation accuracy
- From: "Beliavsky" <beliavsky@xxxxxxx>
- Date: 28 Mar 2007 11:41:54 -0700
On Mar 28, 2:33 pm, Thomas Koenig <Thomas.Koe...@xxxxxxxxx> wrote:
Richard Maine <nos...@xxxxxxxxxxxxx> wrote:
Yes. It seems sort of pointless to declare pi_d as double precision and
also properly specify the literal as double precision, but then give
only 8 digits of it. It isn't as though the compiler is going to be able
to guess "oh, that's close to pi, and the variable name has pi in it, so
maybe I should use a more accurate value of pi instead of the one
specified).
Can we have a standardized way to specify pi? Please?
Usually I don't use plain pi repeatedly in my code but 2*pi or sqrt(pi/
2) or ... . Should the committee standardize all of these? I use Paul
van Delst's "Fundamental Constants" for pi and related quantities. It
is part of his "Utility Source Code" at http://www.ssec.wisc.edu/~paulv/Fortran90/Utility/
..
.
- References:
- calculation accuracy
- From: Ben
- Re: calculation accuracy
- From: glen herrmannsfeldt
- Re: calculation accuracy
- From: Richard Maine
- Re: calculation accuracy
- From: Thomas Koenig
- calculation accuracy
- Prev by Date: Re: unable to view symbols in gdb
- Next by Date: Re: BOZ literal constants, a puzzle for us Standard-loving types.
- Previous by thread: Re: calculation accuracy
- Next by thread: RE: calculation accuracy
- Index(es):
Relevant Pages
|
|