muhammad.rahiz at OUCE.OX.AC.UK
Sat Jun 20 00:12:50 EDT 2009
Perhaps I should brief you on what I'd like to do.
I have a set of RegCM data (1960-1990 and 2070-2010) in .bin/.ctl
format. I'm trying to extract only selected variables i.e. temperature
and precipitation from the output file. In CDO, I can issue the command
'selvars' to extract a variable but I'm new to GrADS and there is no
equivalent function of extracting variable(s). A description of a sample
output file is as follows;
title RegCM normal output variables
pdef 118 109 eta.u 80.000 21.000 0.45045 0.42900
xdef 303 linear 46.25 0.2252
ydef 226 linear -5.07 0.2252
zdef 18 levels 1008.23 998.20 983.15 958.07 922.96 877.81 822.63 757.42
682.18 601.92 521.66 441.40 361.14 280.88 205.63 140.42 85.24 35.08
tdef 124 linear 6z01jul1988 6hr
u 18 0 westerly wind (m/s)
v 18 0 southerly wind (m/s)
w 18 0 omega (hPa/s) p-velocity
t 18 0 air temperature (degree)
qv 18 0 air specific humidity
qc 18 0 cloud water mixing ratio
psa 0 99 surface pressure (hPa)
tpr 0 99 total precipitation(mm/day
tgb 0 99 groud temperature in BATS
swt 0 99 total soil water in mm H2O
rno 0 99 accumulated infiltration
I have a script (save.gs) that someone (Bin Guan) wrote that enables me
to extract the variable(s) I want. So far, running the script for one
ATM.xxx.ctl file does the job of extracting temperature. But I have 1080
(there are 360 ATM, 360 SRF and 360 RAD data) x 2 sets of files (present
and future) to extract so the script writer recommended doing an
aggregate and running the save.gs file just once.
Previously, I did this in a simple script.
'set t 1 124'
'save -v t -n t -u -1.e34 -o atm.1988070100 -p ./output/'
This works for one ctl file but I want Grads to automatically choose the
input file and prints the output file. Any idea of how I can do that?.
In shell scripting, I can issue a wildcard (*). For example,
$ for input in $(ls *.tgz);do
tar xvfz $input
Obviously, wildcards do not work in Grads and it requires the specific
name for the descriptor file as input.
While Grads can take shell commands, can shell take grad commands?
In short, I'd like to do either (but I'd like to know how to do both),
1. aggregate of all data files and run 'save.gs' once or
2. auto select input and generate corresponding output and run 'save.gs'
on every file.
If you have any suggestions, please advise.
Muhammad Rahiz | Doctoral Student in Regional Climate Modeling
Climate Research Laboratory, School of Geography & the Environment
Oxford University Centre for the Environment, University of Oxford
South Parks Road, Oxford, OX1 3QY, United Kingdom
Tel: +44 (0)1865-285194 Mobile: +44 (0)7854-625974
Email: muhammad.rahiz at ouce.ox.ac.uk
See Hai Ooi wrote:
> Hello Muhammad,
> Without knowing the nature/characteristics of your data file and the full data description file, it is difficult for me to offer any help, if possible.
> --- On Sat, 20/6/09, Muhammad Rahiz <muhammad.rahiz at OUCE.OX.AC.UK> wrote:
>> From: Muhammad Rahiz <muhammad.rahiz at OUCE.OX.AC.UK>
>> Subject: Re: Template usage
>> To: GRADSUSR at LIST.CINECA.IT
>> Date: Saturday, 20 June, 2009, 8:29 AM
>> Hello Ooi,
>> Thanks for the reply. I've made the changes but the problem still persists. Any further recommendations?
>> Muhammad Rahiz | Doctoral Student in Regional Climate Modeling
>> Climate Research Laboratory, School of Geography & the Environment
>> Oxford University Centre for the Environment, University of Oxford
>> South Parks Road, Oxford, OX1 3QY, United Kingdom
>> Tel: +44 (0)1865-285194 Mobile: +44 (0)7854-625974
>> Email: muhammad.rahiz at ouce.ox.ac.uk
> New Email names for you!
> Get the Email name you've always wanted on the new @ymail and @rocketmail.
> Hurry before someone else does!
More information about the gradsusr