Re: PyOpenGL_accelerate / subclassed numpy arrays

Chris Barker - NOAA Federal <[email protected]> Mon, 9 Nov 2015 08:05:22 -0800
Newsgroups gmane.comp.python.opengl.user
Message-ID <7222150143615216347@unknownmsgid>
It looks like accelerate is written in Cython. So you should be able
to refactor it to use memoryviews, which I _think_ will work out of
the box with proper nearest subclasses.

-CHB


> On Nov 9, 2015, at 5:22 AM, Nicolas P. Rougier <[email protected]> wrote:
>
>
> Ok, thanks for the explanation, I'll look into that then.
>
> Nicolas
>
>
>
>> On 08 Nov 2015, at 18:04, Chris Barker <[email protected]> wrote:
>>
>> On Sun, Nov 8, 2015 at 8:20 AM, Rob McMullen - NOAA Affiliate <[email protected]> wrote:
>> I know I've run into that problem when my VBO data was in recarray format. Defining the VBO as a view of the array in dtype=np.float32 format worked.
>>
>> I haven't lloked an ANY of the relevant coce here, but this seems to indicae that  PyOpenGL_accelerate can't handle subclasses of ndarray -- which doesn't surprise me -- it's notoriously difficult to do that.
>>
>> but the error:
>>
>> File "arraydatatype.pyx", line 57, in OpenGL_accelerate.arraydatatype.HandlerRegistry.c_lookup (src/arraydatatype.c:2230)
>> TypeError: No array-type handler for type <class 'glumpy.gloo.buffer.VertexBuffer'> (value: VertexBuffer([151,  10,   1, ...,   0, 128,  63], ) registered
>>
>> makes me think that arraydatatype.HandlerRegistry is where you need to put awareness of a new subclass.
>>
>> or it needs to be fixed to recognize ndarray subclasses -- either by directly checking if it is a proper subclass, or by using the array protocol -- which I think is the __array__ method...
>>
>> HTH,
>>
>> -CHris
>>
>>
>>
>> Rob
>>
>> On Thu, Nov 5, 2015 at 11:08 PM, Nicolas P. Rougier <[email protected]> wrote:
>>
>> I've made a numpy array subclass for a vertex buffer that works great with PyOpenGL but does not work if PyOpenGL_accelerate is installed, it complains about about missing handler:
>>
>> File "/Users/rougier/Development/Public/glumpy/glumpy/gloo/buffer.py", line 75, in _update
>>    gl.glBufferSubData(self.target, offset, nbytes, data)
>>  File "latebind.pyx", line 44, in OpenGL_accelerate.latebind.Curry.__call__ (src/latebind.c:1201)
>>  File "/usr/locaL/Cellar/python3/3.5.0/Frameworks/Python.framework/Versions/3.5/lib/python3.5/site-packages/OpenGL/GL/VERSION/GL_1_5.py", line 122, in glBufferSubData
>>    data = ArrayDatatype.asArray( data )
>>  File "arraydatatype.pyx", line 174, in OpenGL_accelerate.arraydatatype.ArrayDatatype.asArray (src/arraydatatype.c:4221)
>>  File "arraydatatype.pyx", line 57, in OpenGL_accelerate.arraydatatype.HandlerRegistry.c_lookup (src/arraydatatype.c:2230)
>> TypeError: No array-type handler for type <class 'glumpy.gloo.buffer.VertexBuffer'> (value: VertexBuffer([151,  10,   1, ...,   0, 128,  63], ) registered
>>
>>
>> Python: 3.5
>> PyOpenGL: 3.1.0
>> PyOpenGL_accelerate: 3.1.0
>> The package I'm developing: http://glumpy.github.io
>>
>> Any idea ?
>>
>> Nicolas
>>
>>
>>
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> PyOpenGL Homepage
>> http://pyopengl.sourceforge.net
>> _______________________________________________
>> PyOpenGL-Users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/pyopengl-users
>>
>>
>> ------------------------------------------------------------------------------
>>
>> _______________________________________________
>> PyOpenGL Homepage
>> http://pyopengl.sourceforge.net
>> _______________________________________________
>> PyOpenGL-Users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/pyopengl-users
>>
>>
>>
>>
>> --
>>
>> Christopher Barker, Ph.D.
>> Oceanographer
>>
>> Emergency Response Division
>> NOAA/NOS/OR&R            (206) 526-6959   voice
>> 7600 Sand Point Way NE   (206) 526-6329   fax
>> Seattle, WA  98115       (206) 526-6317   main reception
>>
>> [email protected]
>

------------------------------------------------------------------------------
Presto, an open source distributed SQL query engine for big data, initially
developed by Facebook, enables you to easily query your data on Hadoop in a 
more interactive manner. Teradata is also now providing full enterprise
support for Presto. Download a free open source copy now.
http://pubads.g.doubleclick.net/gampad/clk?id=250295911&iu=/4140
_______________________________________________
PyOpenGL Homepage
http://pyopengl.sourceforge.net