Hi Alberto,
As I explained several times, the overlap checker is able to find
overlaps ONLY if some vertices of one node's mesh is inside (not on the
surface) of the other node. In your example this is not happening, so
you have to use the sampling method ("s") as Andreas explained. When
sampling is done, the number of overlapping candidates cannot be
computed nor put in the list of overlaps. The result is given in a
graphical manner but you should not draw in "ogl" mode. See the atached
macro for the output.
Andrei
Alberto Pulvirenti wrote:
> So this means that there is an overlap?
> Then there must be a bug in the code, because as you notice, at the end
> it tells me that there are no overlaps, and if I open a TBrowser to look
> at these overlaps, the folder named "Illegal overlaps" is empty.
> Then, the final output of this method is misleading.
>
> Thanks anyway,
> best regards.
>
> Alberto
>
> Andreas Morsch wrote:
>> Dear Alberto, I get
>> Info in <TGeoChecker::CheckOverlapsBySampling>: Overlap region(s)
>> adding-up to 11844 +/- 108.83 [cm3] for daughters of CONT
>>
>> as it should be.
>> Best regards - Andreas.
>>
>>
>>> Dear Andreas and all,
>>>
>>> here in attachment there is a test macro I created to check this.
>>> This macro simply creates two boxes and displaces them into a
>>> container in
>>> such a way that they overlap by a quite large region.
>>> Nevertheless, when I do the overlap check there, it finds no overlaps.
>>>
>>> I am working with the Aliroot head of last friday and ROOT v5.16.00
>>>
>>> I don't know what else I should do to make it tell me that there "are"
>>> overlaps, but on the other side I cannot work if I cannot rely on the
>>> overlap
>>> finder.
>>>
>>> Any help will be extremely useful to me.
>>> Thanks in advance.
>>>
>>> Cheers,
>>>
>>> Alberto
>>>
>>> Andreas Morsch wrote:
>>>
>>>> Dear Alberto, indeed not all overlaps can be detected in an
>>>> analytical way.
>>>> Please use the overlap checking with the option "s" = sampling. Best
>>>> regards
>>>> - Andreas.
>>>>
>>>>
>>>>> Dear all,
>>>>>
>>>>> I wrote this email to both ROOT and AliROOT mailing lists, because
>>>>> I don't
>>>>> know at what level the problem could be.
>>>>> I have implemented something with TGeo, and when I try to draw it
>>>>> with the
>>>>> OpenGL viewer, I get a picture like the one I attaches (it is a
>>>>> detail of
>>>>> the
>>>>> whole).
>>>>> As it is quite evident, there IS an overlap, because I didn't compute
>>>>> correctly the translation of the blue box in the picture, and then
>>>>> it is
>>>>> not
>>>>> placed exactly in the hole I prepared for it, and something
>>>>> "enters" the
>>>>> yellow volume before it. Nevertheless, when I run the "CheckOverlaps"
>>>>> method
>>>>> of gGeoManager, it tells me that there are NO overlaps.
>>>>> What's wrong in my test macro? I attached it also to the email, but
>>>>> the
>>>>> geometry object I created is not in CVS, so you can just look at
>>>>> the code
>>>>> and
>>>>> tell me if there is something wrong in how I use the geometry objects.
>>>>>
>>>>> Thanks for any help
>>>>>
>>>>> Best regards,
>>>>>
>>>>> Alberto
>>>>>
>>>>>
>>>>
>>>
>>
>>
>
Received on Mon Jul 23 2007 - 13:18:29 CEST