[Shake-dev] About sm_nearneighbor

[Bruce Worden]

Hi WSC,

The -w option is something I added to sm_nearneighbor to allow for the  
output of the sum of the weights of the data at each grid point in the  
output. This is needed to both normalize the weighted average and to  
produce the map of total uncertainty. $nnwgto is the weights of the  
native observations, and $nnwgtc is the weights of the converted  
observations. The weight of the GMPE is derived elsewhere.

To see the relative weight of the GMPE, and the native and converted  
observations, run grind with the -rcg option and then run the program  
view_rcg to plot the results. The output will be in PostScript files  
in the event's richter directory with names  
<param>_<type>_contrib.grd, where <param> is one of "mi" (macroseismic  
intensity), "pga" or "pgv," and <type> is one of "obs," "conv," or  
"ests" for the weights of the native observations, the converted  
observations, or the GMPE estimates, respectively.

Bruce


On Dec 7, 2010, at 7:51 PM, WSC wrote:

> Hi,
>
> I konw the sm_nearneighbor is base on GMT`s nearneighbor,
> but the "nearneighbor" in GMT do not  have a -w(in lower case)  
> option ,
> I just know it is used to compute the pga_rock.grd(or pgv_rock.grd)  
> with sm_nearneighbor`s
> -G ouput,but I dont know what it is.
>  so what`s  the output of SM`s sm_nearneighbor with -w(in lower  
> case)option??
>
> codes in grind maybe at line  1340 :
> $cmd = "$cfname -G$nnoutc -W -w$nnwgtc $grdpars -S$gmroi "
>            . "-D$gmdecay -E0 -N16/1";
>       print STDERR "Running 'sm_nearneighbor $cmd'\n" if $debug;
>       `$shakebin/sm_nearneighbor$v $cmd`;
>
> Yes, I want to know what $nnwgtc is here!
>
>
> Thanks,
> WSC
>
>
> 网易163/126邮箱百分百兼容iphone ipad邮件收发  
> _______________________________________________
> Shake-dev mailing list
> Shake-dev at geohazards.usgs.gov
> https://geohazards.usgs.gov/mailman/listinfo/shake-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://geohazards.usgs.gov/pipermail/shake-dev/attachments/20101209/1dc2b706/attachment.html>