Issue with hggroup and zlim

[Ben Abbott]

On Sep 2, 2008, at 7:11 PM, David Bateman wrote:

> Ben Abbott wrote:
>> On Sep 1, 2008, at 10:22 AM, David Bateman wrote:
>>> David Bateman wrote:
>>>> Ben Abbott wrote:
>>>>> I think that depends upon compatibility in other areas.
>>>>>
>>>>> Trusting my memory, plot3 (<real_matrix>, <real_matrix>,   
>>>>> <real_matrix>) produces a series of 3D lines in Matlab, but   
>>>>> produces a single line in Octave ... the Octave result will zig-  
>>>>> zag (see the code below)
>>>>>
>>>>>    [x, y] = meshgrid (0:10, 0:10);
>>>>>    z = x.*y;
>>>>>    figure(1)
>>>>>    clf
>>>>>    plot3 (x, y, z)
>>>>>
>>>>> Same for plot3 (<real_matrix>, <complex_matrix)
>>>>>
>>>>>    y = y + 1i*z;
>>>>>    figure(2)
>>>>>    clf
>>>>>    plot3 (x, y)
>>>>>
>>>>> This issue is what I think needs attention, and I have a faint   
>>>>> recollection that maintaining the complex_vector/matrix feature   
>>>>> was problematic.
>>>>>
>>>>> Ben
>>>>>
>>>> Ok, this is clearly a bug. I think the attached addresses this.  
>>>> and  now
>>>>
>>>> [x, y] = meshgrid (0:10, 0:10); z = x.*y; figure(1); clf; h =   
>>>> plot3(x,y + 1i *z); zlim(0,100)
>>>>
>>>> and
>>>>
>>>> [x, y] = meshgrid (0:10, 0:10); z = x.*y; figure(1); clf; h =   
>>>> plot3(x,y,z); zlim(0,100)
>>>>
>>>> both work correctly on my machine. Note the zlim is needed to  
>>>> get  around the issue addressed in
>>>>
>>>> http://hg.tw-math.de/octave-graphics-mq/rev/7747c2f2f3ed
>>>>
>>>> D.
>>>>
>>>>
>>> Please use the attached instead, that moves the color definition   
>>> inside the loop over the columns of the matrices so that   
>>> __next_line_color__ is called for each column.
>>>
>>> D.
>> Ok, I've pulled a fresh copy of jwe's sources (from velveeta) and   
>> built with no problems.
>> Unforunately, while the demos for plot3 run for me, but this  
>> example  fails
>> octave:1>> x=ones(10,1)*(1:10);
>> octave:2>> y = x';
>> octave:3>> z = x.*y;
>> octave:4>> plot3(x,y,z)
>>      line 330: *All* edges undefined or out of range, thus no plot.
>> I have no idea what file the line number refers to. I added this  
>> as  another demo and the line reference changed to 110. I tried  
>> fgrep to  locate this error/warning/message but found nothing :-(
>> Ben
>
>
> Try
>
> axis ([0 10 0 10 0 100])
>
> This is due to the unapplied changeset from Michael
>
> http://hg.tw-math.de/octave-graphics-mq/rev/7747c2f2f3ed
>
> for the bug that started this thread way back when ....
>
> D.

Very good. So that error was from gnuplot?

Ben