Re: [Caml-list] Printing floating point numbers & precision

[Remi Vanicat <remi.vanicat@gmail.com>]

On Thu, 30 Sep 2004 12:52:00 -0400, mff <mff@research.att.com> wrote:
> 
> Greetings,
> 
> I have a question regarding the printing of floating point numbers.
> Here's an example:
> 
> lt-mff: /usr/bin/ocaml
>        Objective Caml version 3.07+2
> # max_float;;
> - : float = 1.79769313486231571e+30
> 
> # Printf.sprintf "%e" max_float;;
> - : string = "1.797693e+308"
> 
> So using the default %e format to print a floating point number loses 11 digits
> of precision.    I could be more precise about the number of fractional digits:
> 
> # Printf.sprintf "%.17e" max_float;;
> - : string = "1.79769313486231571e+308"
> 
> But this does not seem to work in general:
> 
> # Printf.sprintf "%.17e" 12345.1;;
> - : string = "1.23451000000000004e+04"
> 
> I've combed the documentation to try to understand the relationship between
> printing and precision (and accuracy!) but have come up empty handed.

This is not a ocaml problem, this a problem with base 2 represention
of decimal number : 1/10 is not representable exactly using standard
ieee number, the more exact precicsion you could have is 0.100000004
(with may be some more 0 between the 1 and the 4).

By the way it is a FAQ, and you could find dicsution about it int hte
archive of this very mailing-list.

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners