[GRASS-dev] loess/lowess smoothing?

[Michael Barton]

I'm messing with some 3D point data (xyz locations of artifacts at a Pleistocene site in Europe) from far in my past that I've always wanted to revisit. I've worked with them in GRASS and some other applications. The R function scatterplot3D seems to be coming closest to what I wanted to do. I want to estimate one or more paleosurfaces on the basis of buried artifact locations, given that postdepositional processes will have moved the artifacts vertically by varying amounts--but amounts that probably vary randomly according to some simple statistical function. This has been examined in 2D profiles, but I'd like to do it in 3D.

scattplot3d allows you to create statistical surfaces from the points and display them above and below the surfaces. This is all pretty straightforward in most 3D visualization/analysis software. The loess smoother seems to pass the initial LPG test ('looks pretty good'). This got me to thinking that while GRASS works with 3D raster and vector data, it doesn't have a loess smoother AFAIK. In GRASS, I've used the neighborhood function and r.in.xyz to estimate some surfaces (e.g., median surface) that also look pretty good. But is there a way to create a 3D loess surface like scatterplot3d in R?

Michael
_____________________
C. Michael Barton
Visiting Scientist, Integrated Science Program
National Center for Atmospheric Research &
University Consortium for Atmospheric Research
303-497-2889 (voice)

Director, Center for Social Dynamics & Complexity 
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University
www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-dev/attachments/20120716/1e3b0c8b/attachment.html>