Hi Jean,
If you were able to do the computation for one single time and level,
then doing the computation time by time, level by level should work.
For appending time steps to a netCDF file, you can simply use
SAVE/APPEND, but to append Z levels to a netCDF file, you will need to
set up the file to begin with to accept more levels. To see how this
works, look up "append, slab to netCDF, example" in the Ferret
Users Guide Index. Also, in your loop, after each append, try
putting a CANCEL MEMORY/ALL command to recover the memory used by the
previous computation.
Ansley
jean oliveira wrote:
Hi Ronaldo,
Yes, I tried time by time, level by level using repeat and append
commands, but the problem is the same (see the script below).
NOAA/PMEL TMAP
FERRET v6.1
Linux(g77) 2.6.9-22. -
03/26/08
use "http://hycom.coaps.fsu.edu:8080/thredds/dodsC/glb_analysis"
use curv_map_05.nc
repeat/range=1504:1869:5/name=a (repeat/range=1:33/name=b (let
out_temp1 =
curv_to_rect(temperature[d=1,i=2699:3850,j=425:1631,l=`a`,k=`b`],map[d=2]);
\
save/file=out_temp1.nc/append out_temp1))
exit
Date: Sun, 14 Jun 2009 15:23:05 -0300
From: palmeira@xxxxxxxxx
Subject: Re: [ferret_users] Curv_to_rect with a big data set
To: jeanbrazil@xxxxxxxxxxx
CC: oar.pmel.ferret_users@xxxxxxxx
Hi Jean,
Did you tried save interpoled field time by time, level by level using
repeat and append commands?
regards
Ronaldo Palmeira.
On Sun, Jun 14, 2009 at 1:15 PM, jean
oliveira <jeanbrazil@xxxxxxxxxxx>
wrote:
Hi all,
I´m trying to regridding
a big data set from a curvilinear grid to a rectilinear
latitude-longitude grid (the area can be seen in the ssh1.gif). I´ve
already done the first step (to create curv_map.nc). But when I try “curv_to_rect”
it is occurring a error (see below). I´m working in a cluster.
Does Anyone have any idea?
Thanks a lot
Jean
PS: I´ve already tried with different numbers for "set mode
desperate" and "set memory/size", and the result is the same.
NOAA/PMEL TMAP
FERRET v6.1
Linux(g77) 2.6.9-22. -
03/26/08
yes? set memory/size=1900
Cached data cleared from memory
yes? use "http://hycom.coaps.fsu.edu:8080/thredds/dodsC/glb_analysis"
yes? use curv_map.nc
yes? sho dat
currently SET data sets:
1> http://hycom.coaps.fsu.edu:8080/thredds/dodsC/glb_analysis
name title
I J
K L
LATITUDE 1:4500
1:3298 ...
...
LONGITUDE 1:4500
1:3298 ... ...
DATE date
...
... ... 1:2039
QTOT surf.
heat flux [60.5H] 1:4500 1:3298
... 1:2039
EMP surf. water flux
[60.5H] 1:4500 1:3298
... 1:2039
SURFACE_TEMPERATURE_TREND
surf. temp. trend [60.5H] 1:4500
1:3298 ... 1:2039
SURFACE_SALINITY_TREND
surf. saln. trend
[60.5H] 1:4500 1:3298
... 1:2039
SSH sea surf.
height [60.5H] 1:4500
1:3298 ... 1:2039
MIXED_LAYER_U_VELOCITY
mix.l. u-velocity [60.5H]
1:4500 1:3298 ...
1:2039
MIXED_LAYER_V_VELOCITY
mixl. v-velocity
[60.5H] 1:4500
1:3298 ... 1:2039
SURFACE_BOUNDARY_LAYER_THICKNESS
bnd.layr.thickness
[60.5H] 1:4500
1:3298 ... 1:2039
MIXED_LAYER_THICKNESS
mix.layr.thickness
[60.5H] 1:4500
1:3298 ... 1:2039
MIXED_LAYER_TEMPERATURE
mix.layr.temp
[60.5H] 1:4500
1:3298 ... 1:2039
MIXED_LAYER_SALINITY
mix.layr.saln
[60.5H] 1:4500
1:3298 ... 1:2039
MIXED_LAYER_DENSITY
mix.layr.dens [60.5H]
1:4500 1:3298 ...
1:2039
U
u-veloc. [60.5H] 1:4500
1:3298 1:33 1:2039
V
v-veloc. [60.5H] 1:4500
1:3298 1:33 1:2039
TEMPERATURE
temp [90.3H] 1:4500
1:3298 1:33 1:2039
SALINITY salinity
[60.5H] 1:4500
1:3298 1:33 1:2039
2> ./curv_map.nc
(default)
name title
I
J K
L
MAP CURV_TO_RECT_MAP(LONIN,LATIN,LO
1:369 1:301 1:4
1:3
yes? set mode diagnostic
yes? set mode desperate:5000
yes? let out_temp1 =
curv_to_rect(temperature[d=1,i=2699:3850,j=425:1631,l=1504:1869],map[d=2])
yes? save/file=out_temp1.nc out_temp1
eval
EX#1 C: 4 I: 1
369 J: 1 301
K: 1 33 L:
-999 -999
eval OUT-EMP1 C: 6 I:
1 369 J: 1
301 K: 1 33
L: -999 -999
eval (C01,V01 C:
8 I: 1 369 J:
1 301 K: 1
33 L: -999 -999
eval (C14,V01 C:
10 I: -999 -999 J: -999 -999 K: 1
33 L: -999 -999
**ERROR: insufficient memory: A
negative number of words were requested.
*** NOTE: The current grid is
most likely too large
Novo Internet Explorer 8: mais rápido e muito mais seguro. Baixe
agora, é grátis!
Novo Internet Explorer 8: mais rápido e muito mais seguro. Baixe agora, é grátis!
|