| nearest node of Delauny tesselation [message #84060] |
Thu, 25 April 2013 14:35  |
Jeremy Bailin
Messages: 618
Registered: April 2008
|
Senior Member |
|
|
Under the category of "this must be easy, but I can't seem to figure out
the right function":
If I have created a Delauny tesselation using TRIANGULATE, how can I
easily find which nodes form the triangle that contains an arbitrary
point in the space?
(more specifically, I am using /NATURAL_NEIGHB interpolation in GRIDDATA
and it's going horribly wrong for one point, so I'm trying to figure out
what nodes it's actually using in the interpolation for that point)
Thanks,
-Jeremy.
|
|
|
|
| Re: nearest node of Delauny tesselation [message #84125 is a reply to message #84060] |
Fri, 26 April 2013 19:32  |
ben.bighair
Messages: 221
Registered: April 2007
|
Senior Member |
|
|
On Friday, April 26, 2013 7:46:33 PM UTC-4, Jeremy Bailin wrote:
> On 4/26/13 3:37 PM, ben.bighair wrote:
>
>> On Thursday, April 25, 2013 5:35:40 PM UTC-4, Jeremy Bailin wrote:
>
>>> Under the category of "this must be easy, but I can't seem to figure out
>
>>>
>
>>> the right function":
>
>>>
>
>>>
>
>>>
>
>>> If I have created a Delauny tesselation using TRIANGULATE, how can I
>
>>>
>
>>> easily find which nodes form the triangle that contains an arbitrary
>
>>>
>
>>> point in the space?
>
>>>
>
>>>
>
>>>
>
>>> (more specifically, I am using /NATURAL_NEIGHB interpolation in GRIDDATA
>
>>>
>
>>> and it's going horribly wrong for one point, so I'm trying to figure out
>
>>>
>
>>> what nodes it's actually using in the interpolation for that point)
>
>>>
>
>>
>
>> Hi,
>
>>
>
>> COuld you use IDLanROI::ContainsPoint()? You would have to convert the triangulations to ROIs first.
>
>>
>
>> Cheers,
>
>> Ben
>
>>
>
>
>
> Yes, that would probably work... although converting each one separately
>
> to an ROI sounds like overkill to just figure out where one point lies.
>
> :) But that's the best suggestion, so I suspect that's what's going to
>
> happen.
>
Well, you could at least narrow it down using the grid coordinates, the input point coordinates and VALUE_LOCATE. Seems like "horribly wrong" would stick out like a sore thumb if you could reduce it to just a handful of candidates.
If you go the IDLanROI route then be sure to stuff them into IDLanROIGroup and let it do the fussing with ContainsPoint().
TRIANGULATE, x, y, tri, bounds
d = SIZE(tri, /dim)
rois = OBJ_NEW("IDLanROIGroup")
for i = 0, d[1]-1 do rois->Add, OBJ_NEW("IDLanROI", x[tri[*,j]], y[tri[j]])
inout = rois->ContainsPoints(badGridX, badGridY)
badTris = which(inout GT 0, nBadTris)
|
|
|
|
| Re: nearest node of Delauny tesselation [message #84126 is a reply to message #84060] |
Fri, 26 April 2013 16:46 |
Jeremy Bailin
Messages: 618
Registered: April 2008
|
Senior Member |
|
|
On 4/26/13 3:37 PM, ben.bighair wrote:
> On Thursday, April 25, 2013 5:35:40 PM UTC-4, Jeremy Bailin wrote:
>> Under the category of "this must be easy, but I can't seem to figure out
>>
>> the right function":
>>
>>
>>
>> If I have created a Delauny tesselation using TRIANGULATE, how can I
>>
>> easily find which nodes form the triangle that contains an arbitrary
>>
>> point in the space?
>>
>>
>>
>> (more specifically, I am using /NATURAL_NEIGHB interpolation in GRIDDATA
>>
>> and it's going horribly wrong for one point, so I'm trying to figure out
>>
>> what nodes it's actually using in the interpolation for that point)
>>
>
> Hi,
>
> COuld you use IDLanROI::ContainsPoint()? You would have to convert the triangulations to ROIs first.
>
> Cheers,
> Ben
>
Yes, that would probably work... although converting each one separately
to an ROI sounds like overkill to just figure out where one point lies.
:) But that's the best suggestion, so I suspect that's what's going to
happen.
-Jeremy.
|
|
|
|
| Re: nearest node of Delauny tesselation [message #84128 is a reply to message #84060] |
Fri, 26 April 2013 13:37 |
ben.bighair
Messages: 221
Registered: April 2007
|
Senior Member |
|
|
On Thursday, April 25, 2013 5:35:40 PM UTC-4, Jeremy Bailin wrote:
> Under the category of "this must be easy, but I can't seem to figure out
>
> the right function":
>
>
>
> If I have created a Delauny tesselation using TRIANGULATE, how can I
>
> easily find which nodes form the triangle that contains an arbitrary
>
> point in the space?
>
>
>
> (more specifically, I am using /NATURAL_NEIGHB interpolation in GRIDDATA
>
> and it's going horribly wrong for one point, so I'm trying to figure out
>
> what nodes it's actually using in the interpolation for that point)
>
Hi,
COuld you use IDLanROI::ContainsPoint()? You would have to convert the triangulations to ROIs first.
Cheers,
Ben
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
