Loading ...
Sorry, an error occurred while loading the content.

GEOSTATS: SUM: Post processing of simulation results in GSLIB

Expand Messages
  • Andrew Lister
    Hello, Many thanks to all who replied. It seems like the consensus is that writing some fortran or c code to do it is the best bet. Christopher LLoyd is
    Message 1 of 1 , Nov 5 5:59 AM
    • 0 Attachment
      Hello, Many thanks to all who replied.  It seems like the consensus is that writing some fortran or c code to do it is the best bet.  Christopher LLoyd is doing just that.  I should really sit down with a manual myself!
      Here are some responses from some of the people who responded:
      Susan Srivkin rewrote some GSLIB kriging code that flips gslib kriging output over.  I also found a command in arc info grid called gridflip, which does the same thing, as far as I can tell.
      I use RSI's IDL (Interactive Data Language), which makes
      displaying and analyzing output from gslib very convenient.
      Another excellent choice is MatLab.  I also use Splus, but
      the image-processing capabilities are much more limited than
      the other two.
      Jennifer Dungan  
      Both G0cad and IRAP/RMS have interfaces and better viewers
      Glenn L. Krum
      I've been dumping them into a mining package, but have only started to play
      here.  I'll be most interested in the results of your survey if you collate
      them.  Could I have a copy of your responses please?
      Ivor Jones
      GSLIB has a little program called AddCoords which works quite well then
      I suggest you load it into a database to get your stats using GROUP BY functions (ie realisatin number and or layer).  Depends how big the data set are and how patient you are.  If you use MS Access you can link direct by ODBC to MapInfor and I'd be surprised if ArcInfo does not have the same facility. You can get a compile Fortran90 DOS exe file for Addcoords from
      Mark Murphy
      Andrew, I ve found the vario and variofit packages in UNCERT to be the most efficient means of calculating variograms on multiple realizations and comparing the ensemble of variograms to the model variogram.  I have not had as much luck with the UNCERT package for calculating summary statistics, but as you say, it is very easy to write a code in Fortran or C to do just that.
      Sean McKenna
      It wouldn't solve all your problems but it is quite easy to insert a couple
      of "write" statements into the GSLIB code so that position coordinates are
      associated with each kriged or simulated value.
      GSLIB provides a utility for doing this after the fact.
      Donald Myers
    Your message has been successfully submitted and would be delivered to recipients shortly.