[STATSGRASS] Re: R/GRASS Gaussian simulation issue

[rsadler at cyllene.uwa.edu.au]

Hi Thomas,

Does z<-qnorm(ppoints(x))$y[rank(x)] work?

Rohan Sadler
Ecosystems Research Group
The University of Western Australia


In reply to:

Date: Fri, 28 Jul 2006 12:29:22 -0400
  From: Thomas Adams <Thomas.Adams at noaa.gov>
  Subject: [STATSGRASS] R/GRASS Gaussian simulation issue
  To: statsgrass-bounces at grass.itc.it, STATSGRASS
          <statsgrass at grass.itc.it>
  Message-ID: <44CA3B62.5060500 at noaa.gov>
  Content-Type: text/plain; charset=ISO-8859-1; format=flowed

  List:

  I am attempting Gaussian simulation of modeled precipitation fields
  which are, of course, like all precipitation data decidedly non-normal.
  In order to do the simulations, the data must be transformed using a
  normal score transform. I can do this in R using:

  z<-qnorm(ppoints(x)), where x is the precipitation field data.

  This works and I can transform the data back to precipitation values.
  The problem I have is how do I do this maintaining the spatial patterns.
  That is, I read the data into R from GRASS:

  precip<-readFLOAT6sp("data")

  After I perform the Gaussian simulation of modeled *transformed*
  precipitation fields, I need to transform the data back to precipitation
  values maintaining geographic/spatial patterns. How do I do this?

  Help is appreciated.

  Regards,
  Tom

  --
  Thomas E Adams
  National Weather Service
  Ohio River Forecast Center
  1901 South State Route 134
  Wilmington, OH 45177

  EMAIL:        thomas.adams at noaa.gov

  VOICE:        937-383-0528
  FAX:        937-383-0033