Welcome, Guest
Username: Password: Remember me

TOPIC: Strange values after interpolation

Strange values after interpolation 8 years 11 months ago #19013

  • c.coulet
  • c.coulet's Avatar
  • OFFLINE
  • Moderator
  • Posts: 3722
  • Thank you received: 1031
Hello

I want to interpolate some topographic values on a mesh but the result seems to be slightly different to what I expected.
I start with the mesh

File Attachment:

File Name: mesh.slf
File Size: 37 KB

I want to apply the values taken from this topographic file

File Attachment:

File Name: topo_z1.xyz.txt
File Size: 49 KB


as I work in a larger mesh, I use a constraint outline for the interpolator

File Attachment:

File Name: z1.i2s.txt
File Size: 1 KB



The bottom left point value is 6.000 but when I map the interpolator2D on the mesh, the mesh node at the same location has the value 6.025!
I can't understand the reason behind this strange behaviour

Martin, could you help me or explain this behaviour if it's logical

Thanks in advance
Christophe
The administrator has disabled public write access.

Strange values after interpolation 8 years 11 months ago #19023

  • Lux
  • Lux's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 96
  • Thank you received: 39
I can reproduce this bug with a 2D Interpolator using the nearest neighbour method.
The bathymetry located at the border is not always taken into account. It seems that only bathymetric points strictly included in the footprint of the mesh are used in the interpolator.
I hope there is (or at least will be) a solution/correction to fix this.
Luc
The administrator has disabled public write access.

Strange values after interpolation 8 years 11 months ago #19039

  • Serrer
  • Serrer's Avatar
  • OFFLINE
  • Moderator
  • Posts: 720
  • Thank you received: 304
Hi Christophe,

You uncovered a bug :-(
This occurs under certain conditions when source data points in the interpolator are concident with nodes in the target mesh AND multiple source data points share the same X coordinates.
You can verify this by simply shifting either the mesh or the xyz points by a VERY small amount in X. (say 10e-6 in this case)

I will pass this on to Alan at NRC.
My apologies for any inconvenience this may have caused.

Cheers... Martin
The administrator has disabled public write access.
Moderators: Serrer, a.barton

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.