Re: Data interpolation
- To: mathgroup at smc.vnet.net
- Subject: [mg96301] Re: Data interpolation
- From: "Sjoerd C. de Vries" <sjoerd.c.devries at gmail.com>
- Date: Wed, 11 Feb 2009 05:20:01 -0500 (EST)
- References: <gmp11b$c6r$1@smc.vnet.net> <gmrm4r$9mr$1@smc.vnet.net>
Come to think of it, I actually meant Interpolation. I thought it
could do irregular sampled data points, and the documentation does
suggest it. However, this feature seems to be restricted to 1D data
sets...
Cheers -- Sjoerd
On Feb 10, 12:50 pm, "Sjoerd C. de Vries" <sjoerd.c.devr... at gmail.com>
wrote:
> hemanth,
>
> ListInterpolation will do what you want nicely.
>
> Cheers -- Sjoerd
>
> On Feb 9, 12:37 pm, hemanth korrapati <hemanthkorrap... at gmail.com>
> wrote:
>
> > Hi,
>
> > I have some 3D point data acquired through a laser range finder. The
> > data is sparse. I need to construct a height field out of this data
> > and as a part of this i have to interpolate heights at some locations
> > {x,y} where the actual data is not available.
> > I want to use B-splines. But it seems that B-splines require uniformly
> > distributed control points, which is not the case with my data.
> > My data cannot form a uniform table which is required by the
> > BSplineSurface function if I sort my data.
>
> > for some data i got the following error:
> > Interpolation::indim: The coordinates do not lie on a structured
> > tensor product grid. >>
>
> > So I started using the "ListDensityPlot" function which does a linear
> > interpolation of the data. The interpolation is sufficiently good. But
> > I would like to get an output consisting of list of tuples like
> > {x,y,InterpolatedHeight} at the desired {x,y} locations. Right now I
> > can only see the plot but cannot get the interpolated heights at
> > desired locations({x,y}).
>
> > Now I have three questions:
> > 1) how do i make my data usable for splines in mathematica ?
> > 2) how do i get the interpolated values at desired {x,y} using
> > "ListDensityPlot" function ?
> > 3) Is there a better way other than the above two methods to
> > interpolate the missing data ?
>
> > Thank you