minor bug in semilogy.m

[Ben Abbott]

On Jan 17, 2008, at 1:14 AM, Matthias Brennwald wrote:

> On 17.01.2008, at 02:06, bug-octave-request at octave.org wrote:
>
>> "semilogy" gives an erroneous warning.
>>
>> I'll get to it in a day or two. However, if anyone would like to fix
>> it please let me know.
>>
>> 	octave:1> x = 2:100;
>> 	octave:2> semilogy(x,x)
>> 	warning: axis: omitting nonpositive data in log plot
>>
>> Ben
>
> This warning message does not occur with my Octave 3.0.0 on Mac OS  
> X...
>
> Matthias

I should have mentioned, it did not on mine either. However on 3.0.0+  
I have it :-(

I just built the binary again last night ...

	octave:91> x = 1:100;
	octave:92> semilogy(x,x)
	warning: axis: omitting nonpositive data in log plot

This warning apparently originates in graphics.cc (presenlty line 1495)

*** graphics.cc ***
1494    {
1495      warning ("axis: omitting nonpositive data in log plot");
1496      min_val = min_pos;
1497    }

I'll have to look further, but the line below implies to me that  
semilogy is getting mixed up with loglog or semilogx, and yet, the x  
values are all positive ? ? ?

1480 	 double min_val = xmin;

When I have the chance, I'll go through the recent ChangeLog and look  
for a likely suspect.

Ben

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.cae.wisc.edu/pipermail/bug-octave/attachments/20080117/81e992b0/attachment.html