The vertical integration routine "vibeta" seems to be causing a floating point
error on our COMPAQ Alpha machines running Red Hat Linux. The same code works
fine on our Solaris machines.
How can I further debug this?
Are there any other routines that do vertical integrals in pressure
coordinates?
Is it possible to get the (FORTRAN? C?) code to vibeta so that I can wrap it
myself and see what's going on?
Joe Barsugli :: jjb AT cdc.noaa.gov :: 303-497-6042
Climate Diagnostics Center :: University of Colorado
-- Example
Here is the screen output:
[jjb@cr01 ~/output]$ ncl < daily_psfc.ncl
Copyright (C) 1995-2002 - All Rights Reserved
University Corporation for Atmospheric Research
NCAR Command Language Version 4.2.0.a023
The use of this software is governed by a License Agreement.
See http://ngwww.ucar.edu/ncl/ for more details.
Floating exception (core dumped)
Here is a test program bombs:
;
load "/usr/local/ncarg/lib/ncarg/nclscripts/csm/gsn_code.ncl"
load "/usr/local/ncarg/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
load "/usr/local/ncarg/lib/ncarg/nclscripts/csm/shea_util.ncl"
;
begin
ccm3 = addfile("daily_debug.nc","r" )
pm = ccm3->pm
hr = ccm3->hr
ps = ccm3->ps
interpolation = ccm3->interp
Top = 150.
vhr = vibeta(pm(time|:,lat|:,lon|:,lev|:),hr(time|:,lat|:,lon|:,lev|:),interpolation,ps,110000.,Top*100.)
end
I'll put the input file at http://www.cdc.noaa.gov/~jjb/ncl/daily_debug.nc if
anyone wants to test this.
_______________________________________________
ncl-talk mailing list
ncl-talk AT ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2b29 : Thu Aug 01 2002 - 07:48:07 MDT