Re: resampling methods

• To: Stuart Wier <wier@xxxxxxxxxxxxxxxx>
• Subject: Re: resampling methods
• From: "From: Bill Hibbard <hibbard@xxxxxxxxxxxxxxxxx>" <billh@xxxxxxxxxxxxx>
• Date: Thu, 24 Aug 2000 15:14:03 -0500 (CDT)

Hi Stu,

This is the bug I fixed yesterday AM - the fix is on the
ftp server.

Cheers,
Bill
----------------------------------------------------------
Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
hibbard@xxxxxxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
http://www.ssec.wisc.edu/~billh/vis.html


On Thu, 24 Aug 2000, Stuart Wier wrote:

> 
> While resampling a 3D data set (a FlatField) "g3ds" into
> another FlatField which is a 2D slice across the 3D data,
> I get some unexpected values produced.
> 
> The max/min values of the 3D grid are 52.45282, -23.908264
> 
> use of "resample" without specifying sampling mode:
> 
>   newffld = (FlatField)ffld.resample(g3ds);
> 
> or using 
>   newffld = (FlatField)ffld.resample(g3ds,
>                                      Data.WEIGHTED_AVERAGE, 
>                                    Data.NO_ERRORS);
> 
> gets a new FlatField whose  max, min are 320.39398, -135.35602 
> -- larger than the original grid's max and min. Since this
> is humidity data it is also surprizing to see a value over 100.
> 
> However use of 
> 
>  newffld = (FlatField)ffld.resample(g3ds, Data.NEAREST_NEIGHBOR, NO_ERRORS);
> 
> gives a reasonable 2D Field whose max, min are 39.018806, -11.9811945  
> 
> What's going on? 
> 
> 
> Stu Wier     
> 
> Unidata Program Center
> UCAR
> P.O. Box 3000
> Boulder, CO 80307
> Unidata home page: http://www.unidata.ucar.edu/
> 

• References:
  • resampling methods
    • From: Stuart Wier