[Python Wpg] Fwd: Cannot figure out this ZeroDivisionError exception

[Mike pfaiffer]

On 11-03-11 2:22 PM, Kevin Veroneau wrote:
> ---------- Forwarded message ----------
> From: Kevin Veroneau<kveroneau at gmail.com>
> Date: Fri, Mar 11, 2011 at 2:22 PM
> Subject: Re: [Python Wpg] Cannot figure out this ZeroDivisionError exception
> To: Stuart Williams<stuart at swilliams.ca>
>
>
> Thank you the help.
>
> I updated the code with the added brackets.  The exception is still firing.
>
> camz is set on line 17:  camx, camy, camz = 0, 0, 300
>
> scale is set to 1, and when the mode is changed to from 0 to 1, the
> following is executed:
> scale = scale/camz
>
> I suspect that I have the scale set too low, it is current 1.  In the end,
> it would be doing:
> scale = 1/300 which then equals a zero-based float point number:  0.003333
>
> If I set the scale to 1, the objects in 3D space are too far to see and
> defeats the purpose of the demo.  Although I do suspect it is more related
> to the math in dividing the scale than it is in the math for the drawing
> routine.
>
> This helps a lot as now the problem is more pinpointed and I know where to
> look.
>
> Thanks!

	Just a thought. Would scale = 1/300 not produce 0 instead of the 
floating point 0.00333? It looks like you may have a data type 
conversion issue. If the numerator is an integer then the result of the 
division will be an integer. I was caught by this a couple of times (and 
will no doubt be caught by it again in the future).

				Later
				Mike



> On Fri, Mar 11, 2011 at 1:58 PM, Stuart Williams<stuart at swilliams.ca>wrote:
>
>> I briefly looked through your code, and the print of scale + movex makes me
>> wonder if you're assuming different operator precedence than Python gives
>> you:
>>
>>    nx = (x[n]+xrotoffset+camx)/nz/scale+movex
>>
>> is interpreted as (formatted to match PEP8):
>>
>>    nx = (((x[n] + xrotoffset + camx) / nz) / scale) + movex
>>
>> I suspect scale is zero, even if scale+movex is not.
>>
>> camz starts at zero, and if line 68 execute then scale becomes zero.
>>
>> The other thing I noticed is your code has the loop over 8 values later in
>> the loop than the code that you're following.
>>
>> Perhaps that will help.
>>
>> Stuart.
>>
>> On Thu, Mar 10, 2011 at 4:33 PM, Kevin Veroneau<kveroneau at gmail.com>wrote:
>>
>>> Hello,
>>>
>>>    I am in the process of learning the basics of 3D math and programming,
>>> and thought using Python and Pygame would be a good choice to prototype in.
>>> The 2D example works great, the isometric 3D example also works.  But in the
>>> 3rd example, when I press Space to enter into full 3D, it uses a flow
>>> control statement to change a calculation.  Even though the calculation is
>>> very similar in both isometric and full 3D, it just does not work and I am
>>> very confused why.
>>>
>>>    You can view the source code online in my public Subversion Repository:
>>> http://django.kveroneau.info/svn/node1/python/pygame/3danim2.py/cat
>>>
>>>    I am following the tutorial step-by-step on this website:
>>> http://skytopia.com/project/cube/cube.html
>>>
>>>    That website explains 3D using pseudo-code, the author made it resemble
>>> BASIC-like syntax for clarity.  As I read through the tutorial, I coded it
>>> in Python, and thus is why there is three python files, one for each phase
>>> of the example.
>>>
>>>    I thought I would ask here, as I noticed many of you on this mailing
>>> list using Python for mathematical purposes, and this exception is related
>>> to a math operation in a calculation.  This is my first time going in depth
>>> with 3D programming and how the math behind it works.  I would of course
>>> move to OpenGL once I understand the fundamentals.
>>>
>>>    On the topic of OpenGL, which OpenGL Python library would you recommend
>>> someone who is just diving into GL.  I feel my Python skills are pretty
>>> strong at this point.  I understand that Blender can be scripted using
>>> Python, would this be a good place to start or should I use a stand-alone
>>> python module?
>>>
>>> Thanks,
>>>    Kevin Veroneau
>>>
>>>
>>> _______________________________________________
>>> Winnipeg Python Users Group mailing list
>>> http://WinniPUG.ca
>>> Winnipeg at python.org
>>> http://mail.python.org/mailman/listinfo/winnipeg
>>>
>>>
>>
>
>
>
> _______________________________________________
> Winnipeg Python Users Group mailing list
> http://WinniPUG.ca
> Winnipeg at python.org
> http://mail.python.org/mailman/listinfo/winnipeg