Re: [OctDev] Re: 'text' problems

[David B. <Dav...@mo...>]

Stefan van der Walt wrote:

>On Fri, Dec 03, 2004 at 03:12:40AM -0600, Pascal A. Dupuis wrote:
>  
>
>>atstr = sprintf("%f,%f", position(1),position(2));
>>
>>str = sprintf("%f", 1e-9) 
>>results in ... 0!
>>
>>The following patch seems to work for me:
>>
>>---------------------------------------------------
>>--- text.m.orig 2004-12-03 10:08:10.000000000 +0100
>>+++ text.m      2004-12-03 10:08:42.000000000 +0100
>>@@ -147,9 +147,9 @@
>>     font = "";
>>   endif
>>   if position(3)!=0,
>>-    atstr = sprintf("%f,%f,%f", position(1),position(2),position(3));
>>+    atstr = sprintf("%e,%e,%e", position(1),position(2),position(3));
>>   else
>>-    atstr = sprintf("%f,%f", position(1),position(2));
>>+    atstr = sprintf("%e,%e", position(1),position(2));
>>   endif
>>   command = sprintf('gset label "%s" at %s %s %s %s%s',
>>                    str, units, atstr, align, rotate, font);
>>----------------------------------------------------------
>>    
>>
>
>If you do that, bigger numbers will also be printed in engineering
>notation.  I.e.
>
>octave:22> sprintf("%e\n", 1024)
>ans = 1.024000e+03
>
>What is the most common behaviour required of "text"?
>
>Regards
>Stefan
>  
>
Stefan the "%e" was replaced with "%g" in octave-forge and committed 
when Pascale reported the problem. Using "%g" should give the best of 
both worlds...

Regards
David

-- 
David Bateman                                Dav...@mo...
Motorola CRM                                 +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 1 69 35 77 01 (Fax) 
91193 Gif-Sur-Yvette FRANCE

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary