How To Increase Precision in Scilab

[Vincent Newell]

Is there any way to get Scilab to use more significant digits? I need
something that can hold all of the data up to 10^12, but SciLab seems
to only hold 8 digits. How do I get it to display with more accuracy?

[philippe]

Hi Vincent,

Vincent Newell a écrit :

> Is there any way to get Scilab to use more significant digits? I need
> something that can hold all of the data up to 10^12, but SciLab seems
> to only hold 8 digits. How do I get it to display with more accuracy?

use the "format" function :

-->%pi
%pi =

3.1415927

-->format('v',20)

-->%pi
%pi =

3.14159265358979312

philippe

[SG]

On 29 juin, 21:06, philippe <rou...@sfr.fr> wrote:
> Hi Vincent,

>
> -->format('v',20)
>
> -->%pi
>  %pi  =
>
>     3.14159265358979312

Arrgg:-)

%eps =
2.220D-16

// > 1e-20

help number_properties

Samuel

[Francis]

Scilab uses double precision numbers:
http://en.wikipedia.org/wiki/Double_precision_floating-point_format.
When printing the number scilab uses a standard format usually 8
digits, but in the calculations it definitively uses double precision
unless you changed it (not sure you can do single precision in scilab).

[janaagel...@gmail.com]

Scilab 6 comes with int64 and uint64 - 18 digits
--> a=10^12;
--> disp(int64(a));
1000000000000


J