[Scilab-users] mesh2d - finding identical points

Fri Nov 20 13:05:36 CET 2020

I guess I do have such an u-v-plane.

The background of the background :

The task is to map an image onto a 3D geometry.

The image includes reference points, which correspond to known x-y-z
coordinates.

So the image plane is (my guess) the u-v-plane.

BR
Philipp

Am Fr., 20. Nov. 2020 um 12:16 Uhr schrieb Stéphane Mottelet <
stephane.mottelet at utc.fr>:

> In order to use mesh2d with your data, there must exist a plane (u,v) on
> which you could project your (x,y,z) data. The mesh obtained on projected
> data (u,v) would give you the triangulation of the (x,z,y) data. This won't
> be possible with mesh2d if you take for example scattered points on a
> sphere. However, any non-linear transformation "unfolding" the data could
> also help to use pure 2d tools like mesh2d.
>
> S.
> Le 20/11/2020 à 12:05, P M a écrit :
>
>
>
> Am Fr., 20. Nov. 2020 um 11:47 Uhr schrieb Stéphane Mottelet <
> stephane.mottelet at utc.fr>:
>
>> Hi again,
>> Le 20/11/2020 à 11:30, P M a écrit :
>>
>> OK,
>> some more background:
>>
>> I actually want to perform a delaunay triangulation on a set of
>> X-Y-Z-coordinates.
>>
>> I suppose you mean a triangulation of scattered X-Y-Z but you call mesh2d
>> with X-Y only ?
>>
>
> Yes, indeed. And there is the bug in my data.
> As x-y coordinates repeat, they do have a different z in the original data.
> But since I call mesh2d with x-y only there are indeed identical points.
>
>
>
>>
>> I know of CGLAB, but it seems only available to Scilab 6.0.x, while I am
>> on Scilab 6.1.0
>>
>> CGLAB is based on a very old version of CGAL and thus is very hard  to
>> maintain as is. There is a lot of work...
>>
>
> I am currently installing Scilab 6.0 to try CGLAB...
>
>>
>> So the nearest thing to a delaunay triangulation I found is mesh2d...
>>
>> Yeah. Suboptimal choice indeed...
>>
>> Because of the delaunay triangulation, I am not sure if the order of
>> points may change.
>>
>> What do you mean ?
>>
>
> The scattered points represent a 3D surface geometry.
> If the points will be resorted, I could imagine that the triangulation
> does not fit to the surface geometry anymore.
>
>>
>> About unique:
>>
>> The result I get is actual of same size as the input matrix...I therefore
>> do assume that each point is unique in the first place (which they actually
>> should).
>>
>> What does repeat are individual x, y, z coordinates, but in different
>> combinations.
>> So the actual point is different.
>>
>>
>> e.g.:    x may repeat, but than the y-z-coordinates are different from
>> point to point
>>
>> Could you share the data (you can send me a private message) ?
>>
>> S.
>>
>>
>> Thanks,
>> Philipp
>>
>>
>>
>> Am Fr., 20. Nov. 2020 um 10:57 Uhr schrieb Stéphane Mottelet <
>> stephane.mottelet at utc.fr>:
>>
>>> Hi Philipp,
>>>
>>> Glad to hear that some people use mesh2d. I know we have to improve the
>>> error message by getting the numbers at the Fortran level... There is also
>>> other errors which are not easy to fix without knowing the number of the
>>> problematic point (point too close to the boundary, for example). Can you
>>> file a bug on BZ ?
>>>
>>> S.
>>> Le 20/11/2020 à 10:03, P M a écrit :
>>>
>>> Dear all,
>>>
>>> using the mesh2d - function I get an error:     mesh2di: some points are
>>> identical.
>>>
>>>
>>> How to identify these points and remove the duplicates?
>>>
>>> There are roughly 500'000 coordinates....
>>>
>>> Thank you,
>>> Philipp
>>>
>>> _______________________________________________
>>> users mailing listusers at lists.scilab.orghttps://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users <https://antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users>
>>>
>>> --
>>> Stéphane Mottelet
>>> Ingénieur de recherche
>>> EA 4297 Transformations Intégrées de la Matière Renouvelable
>>> Département Génie des Procédés Industriels
>>> Sorbonne Universités - Université de Technologie de Compiègne
>>> CS 60319, 60203 Compiègne cedex
>>> Tel : +33(0)344234688http://www.utc.fr/~mottelet <https://antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/www.utc.fr/~mottelet>
>>>
>>> _______________________________________________
>>> users mailing list
>>> users at lists.scilab.org
>>> http://lists.scilab.org/mailman/listinfo/users
>>> <https://antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users>
>>>
>>
>> _______________________________________________
>> users mailing listusers at lists.scilab.orghttps://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users <https://antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users>
>>
>> --
>> Stéphane Mottelet
>> Ingénieur de recherche
>> EA 4297 Transformations Intégrées de la Matière Renouvelable
>> Département Génie des Procédés Industriels
>> Sorbonne Universités - Université de Technologie de Compiègne
>> CS 60319, 60203 Compiègne cedex
>> Tel : +33(0)344234688http://www.utc.fr/~mottelet <https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/www.utc.fr/~mottelet>
>>
>> _______________________________________________
>> users mailing list
>> users at lists.scilab.org
>> http://lists.scilab.org/mailman/listinfo/users
>> <https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users>
>>
>
> _______________________________________________
> users mailing listusers at lists.scilab.orghttps://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users
>
> --
> Stéphane Mottelet
> Ingénieur de recherche
> EA 4297 Transformations Intégrées de la Matière Renouvelable
> Département Génie des Procédés Industriels
> Sorbonne Universités - Université de Technologie de Compiègne
> CS 60319, 60203 Compiègne cedex
> Tel : +33(0)344234688http://www.utc.fr/~mottelet
>
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20201120/63dc1914/attachment.htm>