Solve for point value.

I'm reading a number of values from external sensors which correlate to mesh vertices as shown on the attached image, I now wish to approximate or solve the values at the other vertices. My gut instinct is that this can be solved with the point charge field component however I'm unsure as to how to utilize the decay function. Any pointers on how to figure this out would be great. Thanks

Replies to This Discussion

Permalink Reply by Daniel Piker on October 24, 2013 at 6:03am

I've found that inverse distance weighting, or Shepard's method works nicely for this sort of thing.

(see here for one example http://www.grasshopper3d.com/forum/topics/packing-circles-on-surface)

You'd need to tweak the above slightly for the case where the sample values are exactly on existing vertices of the mesh, to avoid division by zero.

Changing the exponent affects how smooth the interpolation is, with low values being a more gradual blend, and as the value goes up it approaches a voronoi diagram.

Permalink Reply by Robert Harvey on October 24, 2013 at 6:03am

Thank you Daniel, I've just been looking at weighted averages, I'll investigate this method.

Permalink Reply by Robert Harvey on November 9, 2013 at 9:53am

As I'm now working with values that occur at regular intervals I've been looking at bilinear interpolation see attached .gh file for more info. Hope this is useful to someone.

Attachments: