octaviz-help — installing and using octaviz

Re: [Octaviz-help] Specify view from the command line or script.

[Jonathan S. <jjs...@vc...>]

Peter A. Gustafson wrote:
> Hi,
> 
> On Friday 05 October 2007 09:59:05 am you wrote:
>>> Hi,
>>> vtk_contour and vtk_contourf.  I think vtk_contourf actually does what I
>>> want, however in some plots the black contour lines do not show up.
>> Can you provide an example where the lines do not show?
> 
> After more testing... I think the lines are present in general but not visible 
> in the screen when viewed from different perspectives.  If/when I get a 
> better handle on this, I will provide more detail.
> 
>>> When using either as a standalone function, it would be nice if I could
>>> turn the zaxis labels etc off.  I think this can be done but I don't
>>> know how.
>> If you look in vtk_contour.m, there is a commented line near the bottom:
>>
>> %f.axes.YAxisVisibilityOff();
>>
> 
> I tried this and it seems to work _most_ of the time, though not all the time.  
> I would turn it back on if I were you, as I would say it is better to error 
> on the side of working most of the time than having extra labels present all 
> of the time.
> 

Changed in CVS.

>>> In working on a view function, I'm running into trouble printing to pdf
>>> (tex) where labels are clipped.  I see:
>>>
>>> f.renderer.GetActiveCamera().SetClippingRange(...);
>>>
>>> Is this how I would ensure that all labels are printed?
>> Not sure off hand.  Can you provide an example where you are having
>> trouble?
> 
> Try the following:
> 
> [x,y,z]=peaks();
> f=vtk_figure(0);
> vtk_blackonwhite;
> vtk_meshc(x,y,z);
> vtk_axis("square");
> f.renderer.GetActiveCamera().SetPosition(-1, -1.25, .3);
> f.renderer.GetActiveCamera().SetFocalPoint(0,0,0);
> f.renderer.GetActiveCamera().SetViewAngle(5);
> f.renderer.GetActiveCamera().SetViewUp(0., 0., 1.);
> f.renderer.GetActiveCamera().SetClippingRange(0, 10);
> f.renderer.GetActiveCamera().SetParallelScale(0.87);
> f.window.SetSize(400,400);
> vtk_update(f);
> vtk_xlabel("x");
> vtk_ylabel("y");
> vtk_zlabel("zzzzzzzz");
> vtk_print("test.tex","-dpdf","-tex","-vect");
> 
> For me, the zlabel doesn't show. It does show up if it is a single character 
> z, or if I move the plot with the mouse so that the zzzzs are in the middle 
> of the plot.

This is an issue with the postscript/pdf converter (GL2PS).  I get the 
same behavior (the zzzz label not showing) if I just print straight to 
pdf without the -tex flag.  The suggestion I have at the moment is to 
change the aspect ratio of the window like this:

f.window.SetSize(500,400);

This solves the problem for your particular example.  It should also be 
possible to use SetFocalPoint, SetDistance, etc., to adjust the 
placement of the figure in the window, but I haven't yet figured out how 
to then update the rendering without resetting all of these...  Let me 
know if you have any success with those commands.

> 
> Another question, can fon't sizing be turned off.  I want identical font sizes 
> managed by tex.  
> 

At the moment I can only suggest that you post-process your .tex file. 
Perhaps something like:

sed -i -e /fontsize/d test.tex

will work for you.

Regards,
Jonathan

Re: [Octaviz-help] octaviz on Mac OS X

[Jonathan S. <jjs...@vc...>]

Thomas Weber wrote:
> Am Freitag, den 05.10.2007, 09:02 -0600 schrieb Jonathan Stickel:
>> Thomas Weber wrote:
>>> Am Freitag, den 05.10.2007, 08:16 -0600 schrieb Jonathan Stickel:
>>>> Thomas Weber wrote:
>>>>> Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:

>>>>>> Problem 2:
>>>>>> Linking of libraries does not work automatically.  As a workaround, I added
>>>>>>
>>>>>> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon
>>>>> Can you post the mkoctfile script? 
>>>>>
>>>> Do you mean the link.txt file?  It is attached.
>>> No, the mkoctfile script that ships with Octave. It should be a normal
>>> shell script.
>>>
>> OK, it is attached.
> 
> Hmm, Octave is missing an rpath to its libraries. On Linux, the relevant
> line is
> : ${RLD_FLAG="-Wl,-rpath -Wl,/usr/lib/octave-2.9.14"}
> 
> Did you compile Octave yourself or came it packaged like this?
> 

Right now I have octave installed via Fink.  Previously I self-compiled 
octave and had the same problem with the library linking of octaviz.  It 
may be that -rpath is treated differently in Mac OS X.

I needed those flags in link.txt to compile liboctaviz.dylib itself as 
well as the oct files, though, and so the problem isn't just with 
mkoctfile.  Isn't there a way to add those flags through appropriate 
modification of a CMakeLists.txt file?

Thanks,
Jonathan

Re: [Octaviz-help] octaviz on Mac OS X

[Thomas W. <tho...@gm...>]

Am Freitag, den 05.10.2007, 09:02 -0600 schrieb Jonathan Stickel:
> Thomas Weber wrote:
> > Am Freitag, den 05.10.2007, 08:16 -0600 schrieb Jonathan Stickel:
> >> Thomas Weber wrote:
> >>> Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:
> >>>> Dragan and Thomas
> >>>>
> >>>> OK, I finally managed to get octaviz installed on my Mac.  It required a
> >>>> few hacks to make files _post cmake_.  I detail the issues below so
> >>>> that we can try to resolve them at the cmake step.  For the record, I 
> >>>> installed octave and VTK via Fink.
> >>>>
> >>>> Are you subscribed to the octaviz-help mailing list?  If so, I will not 
> >>>> cc you in future emails about this.
> >>> I did subscribe now. I wasn't aware there's a help mailing list.
> >>>
> >>> First question: did you use current CVS for building?
> >> Yes.  Since I had the some of the issues with the modified build system 
> >> that you sent me, I decided to work with what is currently in CVS.
> > 
> > 
> >> OK.  If we get the others resolved, maybe we can do something like "if 
> >> mac os x, exclude building the problem libraries" in CMakeLists.txt. 
> >> The only one that might be missed is TIFFwriter.
> > Might be an option.
> > 
> > 
> >>>> Problem 2:
> >>>> Linking of libraries does not work automatically.  As a workaround, I added
> >>>>
> >>>> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon
> >>> Can you post the mkoctfile script? 
> >>>
> >> Do you mean the link.txt file?  It is attached.
> > 
> > No, the mkoctfile script that ships with Octave. It should be a normal
> > shell script.
> > 
> 
> OK, it is attached.

Hmm, Octave is missing an rpath to its libraries. On Linux, the relevant
line is
: ${RLD_FLAG="-Wl,-rpath -Wl,/usr/lib/octave-2.9.14"}

Did you compile Octave yourself or came it packaged like this?

	Thomas

Re: [Octaviz-help] octaviz on Mac OS X

[Jonathan S. <jjs...@vc...>]

Thomas Weber wrote:
> Am Freitag, den 05.10.2007, 08:16 -0600 schrieb Jonathan Stickel:
>> Thomas Weber wrote:
>>> Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:
>>>> Dragan and Thomas
>>>>
>>>> OK, I finally managed to get octaviz installed on my Mac.  It required a
>>>> few hacks to make files _post cmake_.  I detail the issues below so
>>>> that we can try to resolve them at the cmake step.  For the record, I 
>>>> installed octave and VTK via Fink.
>>>>
>>>> Are you subscribed to the octaviz-help mailing list?  If so, I will not 
>>>> cc you in future emails about this.
>>> I did subscribe now. I wasn't aware there's a help mailing list.
>>>
>>> First question: did you use current CVS for building?
>> Yes.  Since I had the some of the issues with the modified build system 
>> that you sent me, I decided to work with what is currently in CVS.
> 
> 
>> OK.  If we get the others resolved, maybe we can do something like "if 
>> mac os x, exclude building the problem libraries" in CMakeLists.txt. 
>> The only one that might be missed is TIFFwriter.
> Might be an option.
> 
> 
>>>> Problem 2:
>>>> Linking of libraries does not work automatically.  As a workaround, I added
>>>>
>>>> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon
>>> Can you post the mkoctfile script? 
>>>
>> Do you mean the link.txt file?  It is attached.
> 
> No, the mkoctfile script that ships with Octave. It should be a normal
> shell script.
> 

OK, it is attached.

> 
>>>> to Common/CMakeFiles/octaviz.dir/link.txt.  I notice that the equivalent 
>>>> flags are not included or needed when compiling in Linux.  Nonetheless, 
>>>> it shouldn't hurt to add them universally.  Can you help me make changes 
>>>> to the appropriate CMakeLists.txt file to add these?
>>> What does 
>>> 	$ objdump -p /usr/bin/octave-2.9.14
>>> say? This will only give meaningful results if the file is a binary.
>>> Under Linux, the binary is compiled with an rpath that points
>>> to /usr/lib/octave-2.9.14, so the octave binary 'knows' where its
>>> symbols are.
>>>
>> I don't seem to have the objdump command.  What package is it typically 
>> included with?
> 
> binutils
> Though it might be linux-specific. 
> 

I only get this for output:

   $ objdump -p /sw/bin/octave-2.9.14

/sw/bin/octave-2.9.14:     file format mach-o-le


> 
> I'm short on time right now, though I'll check my e-mails over the
> weekend.
> 

No problem.  I do have octaviz working at the moment.  We can take our
time resolving these issues so that others can build on Mac.

Thanks,
Jonathan

Re: [Octaviz-help] octaviz on Mac OS X

[Thomas W. <tho...@gm...>]

Am Freitag, den 05.10.2007, 08:16 -0600 schrieb Jonathan Stickel:
> Thomas Weber wrote:
> > Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:
> >> Dragan and Thomas
> >>
> >> OK, I finally managed to get octaviz installed on my Mac.  It required a
> >> few hacks to make files _post cmake_.  I detail the issues below so
> >> that we can try to resolve them at the cmake step.  For the record, I 
> >> installed octave and VTK via Fink.
> >>
> >> Are you subscribed to the octaviz-help mailing list?  If so, I will not 
> >> cc you in future emails about this.
> > 
> > I did subscribe now. I wasn't aware there's a help mailing list.
> > 
> > First question: did you use current CVS for building?
> 
> Yes.  Since I had the some of the issues with the modified build system 
> that you sent me, I decided to work with what is currently in CVS.


> OK.  If we get the others resolved, maybe we can do something like "if 
> mac os x, exclude building the problem libraries" in CMakeLists.txt. 
> The only one that might be missed is TIFFwriter.
Might be an option.


> >> Problem 2:
> >> Linking of libraries does not work automatically.  As a workaround, I added
> >>
> >> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon
> > 
> > Can you post the mkoctfile script? 
> > 
> Do you mean the link.txt file?  It is attached.

No, the mkoctfile script that ships with Octave. It should be a normal
shell script.


> >> to Common/CMakeFiles/octaviz.dir/link.txt.  I notice that the equivalent 
> >> flags are not included or needed when compiling in Linux.  Nonetheless, 
> >> it shouldn't hurt to add them universally.  Can you help me make changes 
> >> to the appropriate CMakeLists.txt file to add these?
> > 
> > What does 
> > 	$ objdump -p /usr/bin/octave-2.9.14
> > say? This will only give meaningful results if the file is a binary.
> > Under Linux, the binary is compiled with an rpath that points
> > to /usr/lib/octave-2.9.14, so the octave binary 'knows' where its
> > symbols are.
> > 
> I don't seem to have the objdump command.  What package is it typically 
> included with?

binutils
Though it might be linux-specific. 


> >> Problem 3:
> >> For some reason, the location of these files are listed to be (and 
> >> theire .cc files also)
> >>
> >> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt8Array.oct"
> <snip>
> > Bug in VTK, see
> > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=388736
> > 
> 
> OK, good to know.  I will manually edit vtkCommonKit.cmake as a 
> temporary fix.

I'm short on time right now, though I'll check my e-mails over the
weekend.

Bye
	Thomas

Re: [Octaviz-help] octaviz on Mac OS X

[Jonathan S. <jjs...@vc...>]

Thomas Weber wrote:
> Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:
>> Dragan and Thomas
>>
>> OK, I finally managed to get octaviz installed on my Mac.  It required a
>> few hacks to make files _post cmake_.  I detail the issues below so
>> that we can try to resolve them at the cmake step.  For the record, I 
>> installed octave and VTK via Fink.
>>
>> Are you subscribed to the octaviz-help mailing list?  If so, I will not 
>> cc you in future emails about this.
> 
> I did subscribe now. I wasn't aware there's a help mailing list.
> 
> First question: did you use current CVS for building?

Yes.  Since I had the some of the issues with the modified build system 
that you sent me, I decided to work with what is currently in CVS.

> 
>> Problem 1:
>> A few of the vtk functions will not compile, giving this error:
>>
<snip>
>> I suspect that this might be a compiler problem.  I am using Apple Xcode
>> tools 2.4.1 with gcc 4.0.  I might try a different gcc eventually, but 
>> that may involve rebuilding all kinds of dependencies :(
> 
> I can't remember having these problems under Linux. Then again, I don't
> know when I last used gcc-4.0 (Debian is currently switching from 4.1 to
> 4.2).
> 

OK.  If we get the others resolved, maybe we can do something like "if 
mac os x, exclude building the problem libraries" in CMakeLists.txt. 
The only one that might be missed is TIFFwriter.


>> Problem 2:
>> Linking of libraries does not work automatically.  As a workaround, I added
>>
>> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon
> 
> Can you post the mkoctfile script? 
> 

Do you mean the link.txt file?  It is attached.

> 
>> to Common/CMakeFiles/octaviz.dir/link.txt.  I notice that the equivalent 
>> flags are not included or needed when compiling in Linux.  Nonetheless, 
>> it shouldn't hurt to add them universally.  Can you help me make changes 
>> to the appropriate CMakeLists.txt file to add these?
> 
> What does 
> 	$ objdump -p /usr/bin/octave-2.9.14
> say? This will only give meaningful results if the file is a binary.
> Under Linux, the binary is compiled with an rpath that points
> to /usr/lib/octave-2.9.14, so the octave binary 'knows' where its
> symbols are.
> 

I don't seem to have the objdump command.  What package is it typically 
included with?

> 
>> Problem 3:
>> For some reason, the location of these files are listed to be (and 
>> theire .cc files also)
>>
>> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt8Array.oct"
<snip>
> Bug in VTK, see
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=388736
> 

OK, good to know.  I will manually edit vtkCommonKit.cmake as a 
temporary fix.

Thanks,
Jonathan

Re: [Octaviz-help] octaviz on Mac OS X

[Thomas W. <tho...@gm...>]

Am Donnerstag, den 04.10.2007, 07:55 -0600 schrieb Jonathan Stickel:
> Dragan and Thomas
> 
> OK, I finally managed to get octaviz installed on my Mac.  It required a
> few hacks to make files _post cmake_.  I detail the issues below so
> that we can try to resolve them at the cmake step.  For the record, I 
> installed octave and VTK via Fink.
> 
> Are you subscribed to the octaviz-help mailing list?  If so, I will not 
> cc you in future emails about this.

I did subscribe now. I wasn't aware there's a help mailing list.

First question: did you use current CVS for building?

> Problem 1:
> A few of the vtk functions will not compile, giving this error:
> 
> /sw/include/vtk-5.0/vtkStreamTracer.h: In member function 'virtual void
> vtkStreamTracer::SetIntegrationDirection(int)':
> /sw/include/vtk-5.0/vtkStreamTracer.h:263: error:
> 'vtkStreamTracer::<anonymous enum>' is/uses anonymous type
> /sw/include/vtk-5.0/vtkStreamTracer.h:263: error:   trying to
> instantiate 'template<class T> template<class U> octave_int::octave_int(U)'
> 
> As a workaround, I added these lines to CMakeLists.txt of the octaviz 
> directory:
> 
>    # below won't build on Mac OS X, JJS 10/2/07
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkStreamTracer 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkTensorGlyph 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkDEMReader 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkTIFFWriter 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkCameraInterpolator 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkQuaternionInterpolator 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkTransformInterpolator 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkProjectedTerrainPath 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkImagePlaneWidget 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkDistributedStreamTracer 1)
>    SET(VTK_CLASS_WRAP_EXCLUDE_vtkPStreamTracer 1)
> 
> I suspect that this might be a compiler problem.  I am using Apple Xcode
> tools 2.4.1 with gcc 4.0.  I might try a different gcc eventually, but 
> that may involve rebuilding all kinds of dependencies :(

I can't remember having these problems under Linux. Then again, I don't
know when I last used gcc-4.0 (Debian is currently switching from 4.1 to
4.2).

> Problem 2:
> Linking of libraries does not work automatically.  As a workaround, I added
> 
> -L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon

Can you post the mkoctfile script? 


> to Common/CMakeFiles/octaviz.dir/link.txt.  I notice that the equivalent 
> flags are not included or needed when compiling in Linux.  Nonetheless, 
> it shouldn't hurt to add them universally.  Can you help me make changes 
> to the appropriate CMakeLists.txt file to add these?

What does 
	$ objdump -p /usr/bin/octave-2.9.14
say? This will only give meaningful results if the file is a binary.
Under Linux, the binary is compiled with an rpath that points
to /usr/lib/octave-2.9.14, so the octave binary 'knows' where its
symbols are.


> Problem 3:
> For some reason, the location of these files are listed to be (and 
> theire .cc files also)
> 
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt8Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt16Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt32Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt64Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt8Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt16Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt32Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt64Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeFloat32Array.oct"
> "/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeFloat64Array.oct"
> 
> As a workaround, I manually edited
> 
> Common/CMakeFiles/Common.dir/build.make
> Common/CMakeFiles/Common.dir/cmake_clean.cmake
> Common/cmake_install.cmake
> 
> and removed the "/sw/.../VTK/" part of the paths.  Any idea how to 
> resolve this?  I find this bug (I think it is a bug) a bit strange since 
> all the other VTK translated header files do not have this problem.

Bug in VTK, see
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=388736


> Problem 4:
> I needed to add "-lvtkRendering" to the build line for
> vtkRaiseWindow.cc 
> in Common/CMakeFiles/Common.dir/build.make.  Not sure why this isn't 
> needed in Linux, but again it wouldn't hurt to add the flag
> universally. 
>   I'm guessing that this can be done by simply adding the flag to the 
> ADD_CUSTOM_COMMAND for vtkRaiseWindow in Common/CMakeLists.txt.  I
> can 
> try this the next time I compile octaviz.

	Thomas

Re: [Octaviz-help] Specify view from the command line or script.

[Jonathan S. <jjs...@vc...>]

Peter A. Gustafson wrote:
> 2) When plotting vtk_contour() followed by vtk_contour() (I want lines and 
> colors in the same plot), they do not end up on the same z position.  Is this 
> expected?
> 

Which two functions do you want to use together?

Jonathan

Re: [Octaviz-help] octaviz on Mac OS X

[Jonathan S. <jjs...@vc...>]

Dragan and Thomas

OK, I finally managed to get octaviz installed on my Mac.  It required a
few hacks to make files _post cmake_.  I detail the issues below so
that we can try to resolve them at the cmake step.  For the record, I 
installed octave and VTK via Fink.

Are you subscribed to the octaviz-help mailing list?  If so, I will not 
cc you in future emails about this.

Thanks,
Jonathan


Problem 1:
A few of the vtk functions will not compile, giving this error:

/sw/include/vtk-5.0/vtkStreamTracer.h: In member function 'virtual void
vtkStreamTracer::SetIntegrationDirection(int)':
/sw/include/vtk-5.0/vtkStreamTracer.h:263: error:
'vtkStreamTracer::<anonymous enum>' is/uses anonymous type
/sw/include/vtk-5.0/vtkStreamTracer.h:263: error:   trying to
instantiate 'template<class T> template<class U> octave_int::octave_int(U)'

As a workaround, I added these lines to CMakeLists.txt of the octaviz 
directory:

   # below won't build on Mac OS X, JJS 10/2/07
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkStreamTracer 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkTensorGlyph 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkDEMReader 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkTIFFWriter 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkCameraInterpolator 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkQuaternionInterpolator 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkTransformInterpolator 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkProjectedTerrainPath 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkImagePlaneWidget 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkDistributedStreamTracer 1)
   SET(VTK_CLASS_WRAP_EXCLUDE_vtkPStreamTracer 1)

I suspect that this might be a compiler problem.  I am using Apple Xcode
tools 2.4.1 with gcc 4.0.  I might try a different gcc eventually, but 
that may involve rebuilding all kinds of dependencies :(


Problem 2:
Linking of libraries does not work automatically.  As a workaround, I added

-L/sw/lib/octave-2.9.14 -loctinterp -loctave -lcruft -lvtkCommon

to Common/CMakeFiles/octaviz.dir/link.txt.  I notice that the equivalent 
flags are not included or needed when compiling in Linux.  Nonetheless, 
it shouldn't hurt to add them universally.  Can you help me make changes 
to the appropriate CMakeLists.txt file to add these?


Problem 3:
For some reason, the location of these files are listed to be (and 
theire .cc files also)

"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt8Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt16Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt32Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeInt64Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt8Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt16Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt32Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeUInt64Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeFloat32Array.oct"
"/sw/src/fink.build/vtk-py25-5.0.2-1003/VTK/Common/vtkTypeFloat64Array.oct"

As a workaround, I manually edited

Common/CMakeFiles/Common.dir/build.make
Common/CMakeFiles/Common.dir/cmake_clean.cmake
Common/cmake_install.cmake

and removed the "/sw/.../VTK/" part of the paths.  Any idea how to 
resolve this?  I find this bug (I think it is a bug) a bit strange since 
all the other VTK translated header files do not have this problem.


Problem 4:
I needed to add "-lvtkRendering" to the build line for vtkRaiseWindow.cc 
in Common/CMakeFiles/Common.dir/build.make.  Not sure why this isn't 
needed in Linux, but again it wouldn't hurt to add the flag universally. 
  I'm guessing that this can be done by simply adding the flag to the 
ADD_CUSTOM_COMMAND for vtkRaiseWindow in Common/CMakeLists.txt.  I can 
try this the next time I compile octaviz.

Re: [Octaviz-help] Specify view from the command line or script.

[Peter A. G. <pe...@sp...>]

On Monday 01 October 2007 10:07:29 am Jonathan Stickel wrote:
> Hi Pete
>
> Glad you like octaviz.  There is not yet a "view" function, but you are
> welcome to make one and submit it to one of the "trackers" on the
> sourceforge project page.  The code to use is something like
>
> f=vtk_figure(0);
> f.renderer.GetActiveCamera().SetPosition(x,y,z);
>
> Actually, you can just use this code to set your camera without making a
> view function, but a contribution would be welcome!  I am stuck with my
> development efforts at the moment since I can't get octaviz to compile
> on my Mac...
>
> Regards,
> Jonathan


Hi again,

OK, I'm looking into a view function but have to admit some of the vtk syntax 
is (as of yet) non-intuitive to me.  I'll keep plugging away as I find time.

A few other questions: 
1) When using vtk_plot3() followed by vtk_axis("square"), the points are 
scaled and look like ovoids... is there a way to avoid this?
2) When plotting vtk_contour() followed by vtk_contour() (I want lines and 
colors in the same plot), they do not end up on the same z position.  Is this 
expected?

Thanks again, keep up the good work.
Pete

Re: [Octaviz-help] Specify view from the command line or script.

[Jonathan S. <jjs...@vc...>]

Hi Pete

Glad you like octaviz.  There is not yet a "view" function, but you are 
welcome to make one and submit it to one of the "trackers" on the 
sourceforge project page.  The code to use is something like

f=vtk_figure(0);
f.renderer.GetActiveCamera().SetPosition(x,y,z);

Actually, you can just use this code to set your camera without making a 
view function, but a contribution would be welcome!  I am stuck with my 
development efforts at the moment since I can't get octaviz to compile 
on my Mac...

Regards,
Jonathan


Peter Gustafson wrote:
> Hi,
> 
> Kudos to the developers, I'm finding vtk to be a promising alternative
> for plotting in octave.
> 
> A quick question... most of my plots are intended to be printed and are
> generated from scripts.  Is there a view() function which allows
> specification of the viewing angles?  I try to avoid interactive
> generation of plots since I'm generating many, particularly I want to
> avoid the mouse.
> 
> Thanks,
> 
> Pete
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Octaviz-help mailing list
> Oct...@li...
> https://lists.sourceforge.net/lists/listinfo/octaviz-help
> 
> 

[Octaviz-help] Specify view from the command line or script.

[Peter G. <pe...@sp...>]

Hi,

Kudos to the developers, I'm finding vtk to be a promising alternative
for plotting in octave.

A quick question... most of my plots are intended to be printed and are
generated from scripts.  Is there a view() function which allows
specification of the viewing angles?  I try to avoid interactive
generation of plots since I'm generating many, particularly I want to
avoid the mouse.

Thanks,

Pete

[Octaviz-help] Hi

[Herna S. <to...@sa...>]

Hi,

Vaiagra $3. 35
Vaulium $1. 25
Cuialis $3. 75
Abmbien $2. 90
Suoma   $1. 15

http://www.33!rx.com

Important: Remove "!" in the above link

Re: [Octaviz-help] PHxysARMA

[Jairus L. <the...@pa...>]

Hi,

Vniagra 3, 35
Vnalium 1, 25
Cnialis 3, 75
Anmbien 2, 90

http://inur.22rx,com 

Important: Replace "," with "." in the above link

--

The insane smile lit his features once more, and his head drooped onto
his shoulder as Winky wailed and sobbed at his side.
CHAPTER THIRTY-SIX

[Octaviz-help] Help: octaviz makes octave segfault

[Reynald B. <rey...@ep...>]

Hi everybody,

I am facing a segfault with octaviz running on top of octave version
2.9.8. As soon as I type some vtk_* command octave quits with the
following error:

octave:5> vtk_demo
panic: Segmentation fault -- stopping myself...
attempting to save variables to `octave-core'...
error: octave_base_value::save_binary(): wrong type argument `vtk_object'
save to `octave-core' complete
Segmentation fault

Does anyone have an idea about this problem ? I'm using VTK version
4.4.2 on Gentoo Linux with gcc 4.1.1.

Thanks in advance,
Reynald

P.S: I also have a more octave specific problem as "all_va_args" is no
more available with octave 2.9.8 (but it's available with 2.1.72). Any
hints on this ? I'm gonna ask this on the octave mailing-list also.

[Octaviz-help] Re: pcolor function

[Fabian B. <f.b...@gm...>]

Hi Jonathan,

* On 03 Dec 2005 * Jonathan Stickel wrote:

> A number of new m-file functions, as well as many improvements to 
> existing functions, have been added to CVS since the last release.  We 
> would really like to make a new release, but we have kind of worked 
> ourselves into a corner where we are waiting for VTK to officially 
> release a new version.  Anyway, if you are not familiar with using CVS 
> directly, you can download individual files via http:
> 
> http://cvs.sourceforge.net/viewcvs.py/octaviz/octaviz/Scripts/

Thanks! Did see that.

Greetings!
Fabian

Re: [Octaviz-help] Re: pcolor function

[Jonathan S. <jjs...@sb...>]

A number of new m-file functions, as well as many improvements to 
existing functions, have been added to CVS since the last release.  We 
would really like to make a new release, but we have kind of worked 
ourselves into a corner where we are waiting for VTK to officially 
release a new version.  Anyway, if you are not familiar with using CVS 
directly, you can download individual files via http:

http://cvs.sourceforge.net/viewcvs.py/octaviz/octaviz/Scripts/

Jonathan


Fabian Braennstroem wrote:
> Hi Jonathan,
> 
> * On 01 Dec 2005 * Jonathan Stickel wrote:
> 
> 
>>OK, I am not at all familiar with what you are trying to do.  I do see a 
>>vtk_pcolor function in Octaviz, though.  Have you tried it?
> 
> 
> No, I can't find the function. Where should it be? There is no
> vtk_pcolor.m in my /usr/share/octave/2.1.71/site/m/octaviz
> directory.
> 
> Greetings!
> Fabian
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
> _______________________________________________
> Octaviz-help mailing list
> Oct...@li...
> https://lists.sourceforge.net/lists/listinfo/octaviz-help
> 

[Octaviz-help] Re: pcolor function

[Fabian B. <f.b...@gm...>]

Hi Jonathan,

* On 01 Dec 2005 * Jonathan Stickel wrote:

> OK, I am not at all familiar with what you are trying to do.  I do see a 
> vtk_pcolor function in Octaviz, though.  Have you tried it?

No, I can't find the function. Where should it be? There is no
vtk_pcolor.m in my /usr/share/octave/2.1.71/site/m/octaviz
directory.

Greetings!
Fabian

Re: [Octaviz-help] Re: pcolor function

[Jonathan S. <jjs...@sb...>]

OK, I am not at all familiar with what you are trying to do.  I do see a 
vtk_pcolor function in Octaviz, though.  Have you tried it?

Jonathan


Fabian Braennstroem wrote:
> Hi Jonathan,
> 
> * On 30 Nov 2005 * Jonathan Stickel wrote:
> 
> 
>>Fabian
>>
>>I am not really sure what you are asking, perhaps because I do not know 
>>what 'tecplot' is or what its data is for.  The octaviz functions 
>>(vtk_something) are attempts to mimic high-level plotting commands that 
>>are available in Matlab.  Is there something in Matlab that you like 
>>that isn't yet in Octaviz?  Please point that out, and we can try to 
>>implement it.
> 
> 
> Sorry, I was not clear enough; I try it better:
> 
> I have some 3D data (coordinates and velocities) from a CFD simulation (actually in
> tecplot format, but since vtk does not have any reader for
> tecplot, the format is not important). I load these datas
> using 'load' in octave. Now, I would like to display the
> velocity field at a certain plane. Using original vtk
> commands and a vtk reader I would use the vtkCutter, e.g.:
> 
>   cutter = vtkCutter();
>   cutter.SetInput(reader.GetOutput());
> 
> In matlab I used the function 'pcolor', which does about the
> same as the 'cutter'.
> 
> As mentioned there exists no reader for my formated data, so
> I was asking, if it somehow possible to use the 'loaded'
> data aus input for the vtkCutter, e.g.:
> 
>   load data;
>   
>   x=data(:,1);
>   y=data(:,2);
>   z=data(:,3);
>   u=data(:,4);
>   
>   cutter= vtkCutter();
>   cutter.SetInput(x,y,z,u);
> 
> Using the pcolor function in matlab I had to 'reshape' the
> velocity: 
> 
>   U=reshape(u,size(x),size(y));
>   pcolor(x,y,U)
> 
> Maybe, you have an idea how to do that!?
> Probably the best way would be to convert those tecplot data
> into the ensight format ... unfortunately I was not able to
> do that yet :-(
> 
> 
> Greetings!
> Fabian
> 
> 
> 
>>Octaviz does allow you to use VTK objects and manipulate them.  The 
>>ClipCow.m example uses vtkCutter, so you can look at that if that is 
>>what you want.
>>
>>Jonathan
>>
>>
>>Fabian Braennstroem wrote:
>>
>>>Hi,
>>>
>>>I'm still using it :-)
>>>
>>>Right now, I have some trouble to display some 'tecplot'
>>>formated data using octaviz. One problem is, that their
>>>exist no reader in vtk for it. To get around that I would
>>>load the data into octave and display the data using pcolor,
>>>but that just would work in Matlab nicely. 
>>>
>>>Anyways, to get all simulation results in the same vtk-style
>>>I am looking for a vtk-function, which replaces 'pcolor'. In
>>>my point of view, the vtkCutter is the right modul, but I
>>>don't know, how I can get my data into it!? There does not
>>>exist a 'vtk_cutter' function similar to pcolor in octaviz,
>>>doesn't it? 
>>>
> 
> 

[Octaviz-help] Re: pcolor function

[Fabian B. <f.b...@gm...>]

Hi Jonathan,

* On 30 Nov 2005 * Jonathan Stickel wrote:

> Fabian
> 
> I am not really sure what you are asking, perhaps because I do not know 
> what 'tecplot' is or what its data is for.  The octaviz functions 
> (vtk_something) are attempts to mimic high-level plotting commands that 
> are available in Matlab.  Is there something in Matlab that you like 
> that isn't yet in Octaviz?  Please point that out, and we can try to 
> implement it.

Sorry, I was not clear enough; I try it better:

I have some 3D data (coordinates and velocities) from a CFD simulation (actually in
tecplot format, but since vtk does not have any reader for
tecplot, the format is not important). I load these datas
using 'load' in octave. Now, I would like to display the
velocity field at a certain plane. Using original vtk
commands and a vtk reader I would use the vtkCutter, e.g.:

  cutter = vtkCutter();
  cutter.SetInput(reader.GetOutput());

In matlab I used the function 'pcolor', which does about the
same as the 'cutter'.

As mentioned there exists no reader for my formated data, so
I was asking, if it somehow possible to use the 'loaded'
data aus input for the vtkCutter, e.g.:

  load data;
  
  x=data(:,1);
  y=data(:,2);
  z=data(:,3);
  u=data(:,4);
  
  cutter= vtkCutter();
  cutter.SetInput(x,y,z,u);

Using the pcolor function in matlab I had to 'reshape' the
velocity: 

  U=reshape(u,size(x),size(y));
  pcolor(x,y,U)

Maybe, you have an idea how to do that!?
Probably the best way would be to convert those tecplot data
into the ensight format ... unfortunately I was not able to
do that yet :-(


Greetings!
Fabian


> 
> Octaviz does allow you to use VTK objects and manipulate them.  The 
> ClipCow.m example uses vtkCutter, so you can look at that if that is 
> what you want.
> 
> Jonathan
> 
> 
> Fabian Braennstroem wrote:
> >Hi,
> >
> >I'm still using it :-)
> >
> >Right now, I have some trouble to display some 'tecplot'
> >formated data using octaviz. One problem is, that their
> >exist no reader in vtk for it. To get around that I would
> >load the data into octave and display the data using pcolor,
> >but that just would work in Matlab nicely. 
> >
> >Anyways, to get all simulation results in the same vtk-style
> >I am looking for a vtk-function, which replaces 'pcolor'. In
> >my point of view, the vtkCutter is the right modul, but I
> >don't know, how I can get my data into it!? There does not
> >exist a 'vtk_cutter' function similar to pcolor in octaviz,
> >doesn't it? 
> >

-- 

Re: [Octaviz-help] pcolor function

[Jonathan S. <jjs...@sb...>]

Fabian

I am not really sure what you are asking, perhaps because I do not know 
what 'tecplot' is or what its data is for.  The octaviz functions 
(vtk_something) are attempts to mimic high-level plotting commands that 
are available in Matlab.  Is there something in Matlab that you like 
that isn't yet in Octaviz?  Please point that out, and we can try to 
implement it.

Octaviz does allow you to use VTK objects and manipulate them.  The 
ClipCow.m example uses vtkCutter, so you can look at that if that is 
what you want.

Jonathan


Fabian Braennstroem wrote:
> Hi,
> 
> I'm still using it :-)
> 
> Right now, I have some trouble to display some 'tecplot'
> formated data using octaviz. One problem is, that their
> exist no reader in vtk for it. To get around that I would
> load the data into octave and display the data using pcolor,
> but that just would work in Matlab nicely. 
> 
> Anyways, to get all simulation results in the same vtk-style
> I am looking for a vtk-function, which replaces 'pcolor'. In
> my point of view, the vtkCutter is the right modul, but I
> don't know, how I can get my data into it!? There does not
> exist a 'vtk_cutter' function similar to pcolor in octaviz,
> doesn't it? 
> 
> Best Greetings!
> Fabian
> 
> 

[Octaviz-help] pcolor function

[Fabian B. <f.b...@gm...>]

Hi,

I'm still using it :-)

Right now, I have some trouble to display some 'tecplot'
formated data using octaviz. One problem is, that their
exist no reader in vtk for it. To get around that I would
load the data into octave and display the data using pcolor,
but that just would work in Matlab nicely. 

Anyways, to get all simulation results in the same vtk-style
I am looking for a vtk-function, which replaces 'pcolor'. In
my point of view, the vtkCutter is the right modul, but I
don't know, how I can get my data into it!? There does not
exist a 'vtk_cutter' function similar to pcolor in octaviz,
doesn't it? 

Best Greetings!
Fabian


-- 

Re: [Octaviz-help] How to use the vtk_trisurf function

[Dragan T. <dra...@ul...>]

Apparently, there is a bug in vtk_trisurf that prevents you from rendering
surfaces that contain two triangles only. Try with four triangles (don't mind
overlaping triangles):

tri=[1 2 3; 2 3 4; 1 2 3; 2 3 4;]

Cheers,

Dragan

Selon LUK ShunTim <shu...@po...>, 13.11.2005:

> LUK ShunTim wrote:
> > Hello,
> >
> > I do not quite understand the use of the vtk_trisurf. I tried this simple 2
> > triangle example - the unit square with a diagonal connecting (0,0) and
> (1,1).
>
> Sorry! The diagonal should be between (1,0) and (0,1).
>
> > The coordinates (x, y) and function values f are (columns vectors):
> >
> > x=[0. 1. 0. 1.]'
> > y=[0. 0. 1. 1.]'
> > f=[1. 2. 3. 4.]'
> >
> > and the triangulation is
> > tri=[1 2 3; 2 3 4]
> >
> > Now using
> > vtk_trisurf(tri, x, y, f)
> >
> > doesn't work. I got the error message
> >
> > "error: number of rows must match (2 != 3) near line 149, column 27
> > error: evaluating postfix operator `'' near line 149, column 28
> > error: evaluating assignment expression near line 149, column 5
> > error: called from `vtk_trisurf' in file
> > `/usr/share/octave/site/m/octaviz/vtk_trisurf.m'"
> >
> > I think I'm using the "4-parameter" method outlined in the online manual
> > "trisurf can be called with 2(3) or 4(5) parameters as trisurf(t,[x1 y1 z1;
> ...;
> > xn, yn, zn],(c)) or trisurf(t,[x1; ...; xn],[y1; ...; yn], [z1; ...;
> zn],(c))"
> >
> > (It appears this is also the M**lab way to invoke trisurf according to
> their
> > online documentation.)
> >
> > I'd be grateful if someone can provide a working snippet with this simple
> > example. IMHO, it'd be great if it's added to the online documentation.
> >
> > Regards,
> > ST
> > --
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App Server. Download
> it for free - -and be entered to win a 42" plasma tv or your very own
> Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
> _______________________________________________
> Octaviz-help mailing list
> Oct...@li...
> https://lists.sourceforge.net/lists/listinfo/octaviz-help
>
>

Re: [Octaviz-help] How to use the vtk_trisurf function

[LUK S. <shu...@po...>]

LUK ShunTim wrote:
> Hello,
> 
> I do not quite understand the use of the vtk_trisurf. I tried this simple 2
> triangle example - the unit square with a diagonal connecting (0,0) and (1,1).

Sorry! The diagonal should be between (1,0) and (0,1).

> The coordinates (x, y) and function values f are (columns vectors):
> 
> x=[0. 1. 0. 1.]'
> y=[0. 0. 1. 1.]'
> f=[1. 2. 3. 4.]'
> 
> and the triangulation is
> tri=[1 2 3; 2 3 4]
> 
> Now using
> vtk_trisurf(tri, x, y, f)
> 
> doesn't work. I got the error message
> 
> "error: number of rows must match (2 != 3) near line 149, column 27
> error: evaluating postfix operator `'' near line 149, column 28
> error: evaluating assignment expression near line 149, column 5
> error: called from `vtk_trisurf' in file
> `/usr/share/octave/site/m/octaviz/vtk_trisurf.m'"
> 
> I think I'm using the "4-parameter" method outlined in the online manual
> "trisurf can be called with 2(3) or 4(5) parameters as trisurf(t,[x1 y1 z1; ...;
> xn, yn, zn],(c)) or trisurf(t,[x1; ...; xn],[y1; ...; yn], [z1; ...; zn],(c))"
> 
> (It appears this is also the M**lab way to invoke trisurf according to their
> online documentation.)
> 
> I'd be grateful if someone can provide a working snippet with this simple
> example. IMHO, it'd be great if it's added to the online documentation.
> 
> Regards,
> ST
> --

[Octaviz-help] How to use the vtk_trisurf function

[LUK S. <shu...@po...>]

Hello,

I do not quite understand the use of the vtk_trisurf. I tried this simple 2
triangle example - the unit square with a diagonal connecting (0,0) and (1,1).
The coordinates (x, y) and function values f are (columns vectors):

x=[0. 1. 0. 1.]'
y=[0. 0. 1. 1.]'
f=[1. 2. 3. 4.]'

and the triangulation is
tri=[1 2 3; 2 3 4]

Now using
vtk_trisurf(tri, x, y, f)

doesn't work. I got the error message

"error: number of rows must match (2 != 3) near line 149, column 27
error: evaluating postfix operator `'' near line 149, column 28
error: evaluating assignment expression near line 149, column 5
error: called from `vtk_trisurf' in file
`/usr/share/octave/site/m/octaviz/vtk_trisurf.m'"

I think I'm using the "4-parameter" method outlined in the online manual
"trisurf can be called with 2(3) or 4(5) parameters as trisurf(t,[x1 y1 z1; ...;
xn, yn, zn],(c)) or trisurf(t,[x1; ...; xn],[y1; ...; yn], [z1; ...; zn],(c))"

(It appears this is also the M**lab way to invoke trisurf according to their
online documentation.)

I'd be grateful if someone can provide a working snippet with this simple
example. IMHO, it'd be great if it's added to the online documentation.

Regards,
ST
--