I have to mention that I am running your script using ipython notebook,
apparently it gives better exception information.

Here is the notebook with the full exception.

http://nbviewer.ipython.org/urls/raw.githubusercontent.com/guziy/PyNotebooks/master/shapely_test.ipynb?create=1

Cheers


2014-04-28 15:31 GMT-04:00 Oleksandr Huziy <guziy.sa...@gmail.com>:

> I am using Python 2.7.6 and shapely 1.3.1.
>
> What I also noticed that if you increase the distance to 10000 the error
> also disappears...
>
> Cheers
>
>
> 2014-04-28 15:24 GMT-04:00 Alex Flint <alex.fl...@gmail.com>:
>
> Sorry, python 2.7.6.
>>
>>
>> On Mon, Apr 28, 2014 at 3:23 PM, Alex Flint <alex.fl...@gmail.com> wrote:
>>
>>> Hmm that's strange, I haven't seen that error. What version of shapely
>>> are you using (I'm using 1.3.0). Also I'm on python 1.7.6.
>>>
>>> Attaching the exact python script that I am running at my end.
>>>
>>> I have also found that if I add small random numbers to each element of
>>> the data array (gaussian random numbers of order 1e-2) then I do not get
>>> the crash (and in fact I get the correct result), which suggests to me that
>>> this is some instability/round-off issue.
>>>
>>> Alex
>>>
>>>
>>>
>>> On Mon, Apr 28, 2014 at 3:12 PM, Oleksandr Huziy 
>>> <guziy.sa...@gmail.com>wrote:
>>>
>>>> I get this error message when executing your code:
>>>>
>>>> ERROR:shapely.geos:TopologyException: depth mismatch at  at 
>>>> 581037.39250326285 4509301.9737485535
>>>>
>>>>
>>>> Have no idea what this means, but maybe you know))
>>>>
>>>> Cheers
>>>>
>>>>
>>>> 2014-04-28 14:55 GMT-04:00 Alex Flint <alex.fl...@gmail.com>:
>>>>
>>>>> The following code produces an exception under shapely (python
>>>>> version) 1.3.0
>>>>>
>>>>>     import shapely.geometry
>>>>>     path = shapely.geometry.LineString([
>>>>>         [581037.19575668662, 4509299.98342364],
>>>>>         [580996.61134201661, 4509304.202782942],
>>>>>         [581037.19575668662, 4509299.98342364],
>>>>>         [581072.19594137045, 4509296.7024005251],
>>>>>         [581072.19594137045, 4509296.7024005251],
>>>>>         [581037.19575668662, 4509299.98342364]
>>>>>     ])
>>>>>     path.parallel_offset(2, shapely.geometry.JOIN_STYLE.mitre)
>>>>>
>>>>> The exception is:
>>>>>
>>>>>     File
>>>>> "/Users/alexflint/Code/wifi-localization/tools/localization/estimation/scripts/shapely_bug.py",
>>>>> line 39, in <module>
>>>>>       linestring.parallel_offset(2, shapely.geometry.JOIN_STYLE.mitre)
>>>>>     File
>>>>> "/usr/local/Cellar/python/2.7.6/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/shapely/geometry/linestring.py",
>>>>> line 114, in parallel_offset
>>>>>       bool(side == 'left')))
>>>>>     File
>>>>> "/usr/local/Cellar/python/2.7.6/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/shapely/geometry/base.py",
>>>>> line 55, in geom_factory
>>>>>       raise ValueError("No Shapely geometry can be created from null
>>>>> value")
>>>>>     ValueError: No Shapely geometry can be created from null value
>>>>>
>>>>> The points on the path above are actually very close to collinear, but
>>>>> my understanding was that parallel_offset should work regardless. I have
>>>>> attached a plot of the path itself.
>>>>>
>>>>> Any suggestions for what's going on here?
>>>>>
>>>>> Alex
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Community mailing list
>>>>> Community@lists.gispython.org
>>>>> http://lists.gispython.org/mailman/listinfo/community
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Sasha
>>>>
>>>> _______________________________________________
>>>> Community mailing list
>>>> Community@lists.gispython.org
>>>> http://lists.gispython.org/mailman/listinfo/community
>>>>
>>>>
>>>
>>
>> _______________________________________________
>> Community mailing list
>> Community@lists.gispython.org
>> http://lists.gispython.org/mailman/listinfo/community
>>
>>
>
>
> --
> Sasha
>



-- 
Sasha
_______________________________________________
Community mailing list
Community@lists.gispython.org
http://lists.gispython.org/mailman/listinfo/community

Reply via email to