- sounds a lot  like course-work to me...  addressed to whom?

On 15 Jul 2002 03:38:33 -0700, [EMAIL PROTECTED] (Ashraf A
Ahmed) wrote:

> Dear Sir,
> 
> I have a statistical problem and would appreciate your guidance.
> 
> I am doing my research in Civil and Environmental Engineering.
> I have a field site 155 ft length and 60 ft wide, where there are 4 
> boreholes. I know the soil properties only at these boreholes. I 
> need to generate these properties for the whole site.
> I have read that many interpolation options like Kriging, natural 
> neighbor or inverse distance weighted can do that.

With merely four points, you potentially have much more 
OUTSIDE extrapolation instead of INSIDE  interpolation.   

Are those estimators  equally robust that way?

> But I don't know which method of those mentioned above gives me 
> the best results. I would appreciate very much your hint in that.

If one method was *uniformly best*,  would you be likely to read 
much about the others?  Of course, you can't trust my comments here;
I hardly know anything about this subject, at all.

> Also, is there any program which could be downloaded free from 
> your website or any other website that can do that?.

 - whose website?

-- 
Rich Ulrich, [EMAIL PROTECTED]  
http://www.pitt.edu/~wpilib/index.html
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to