Re: fprintf, %e

[Jakub Selesovsky]

Thank you Tom,

I supposed it would be a problem of this kind. I have found another  
solution for both types (+/-) exponents. I will use the sed stream editor  
to replace the 'e+' or the 'e-' string by the 'e+0', resp. 'e-0' string in  
the octave output file. I wanted only to ask if it is possible to solve  
this problem directly in octave.

Jakub

On Wed, 30 Aug 2006 22:31:40 +0200, Tom Holroyd (NIH/NIMH) [E]  
<address@hidden> wrote:

> The problem is with the underlying C library's conversion function, not  
> Octave.
> But you can do this (for positive exponents only):
>
>         x = 1.2345678910e+002;
>         ex = floor(log(x)/log(10));
>         y = x / 10**ex;
>         fprintf('%10.10f+%03d\n', y, ex)
>         => 1.2345678910+002
>
> Jakub Selesovsky wrote:
> > Hello,
> >  I have a little problem with the fprintf function. For another  
> > processing  I need to have the results from octave in this form:
> >  1.2345678910e+002
> >  I am able to control the precision and the width of the column via   
> > '%17.10e' but I don't know how to control the number of digits behind   
> > letter 'e'. I am always getting the output with two digits (e+02) in   
> > exponent, but I have to get three digits (e+002).
> >  Than you for any help.
> > Jakub



--
Jakub Selesovsky
--
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/