Hi Ansley,
Thanks for
you helpful comments. “use/order=xyt”
works well to reshape the grid.
yes?
use/order=xyt HADISST_1_1_SST_2004.nc
yes? show
grid sst
Default
grid for DEFINE VARIABLE is
ABSTRACT
Last
successful data access was on
grid GBR4
GRID GBR4
name
axis
# pts
start
end
NI
X
360 r
1
360
NJ
Y
180 r
1
180
normal Z
NR
T
12 r
1
12
However, I
cannot create the correct
figures using Fill and plot commands for the regions including missing
values
(nan) in the data. The errors in the graphs contain many lines
pointing to
the outside of plot area, and metafile.plt
created by ferret cannot be changed to .ps file
by gksm2ps. Is there
a way to
fix this nan problem?
yes?
list/i=1/J=1:5 SST
VARIABLE : SST (deg)
FILENAME : HADISST_1_1_SST_2004.nc
SUBSET : 5 by 12 points (Y-T)
X : 1
1 2
3 4
5
1 2 3
4 5
1 /
1: nan -18.00 -18.00 -18.00 -18.00
2 /
2: nan nan
nan nan nan
3 /
3: nan nan
nan nan nan
4 /
4: nan nan
nan nan nan
5 / 5:
-18.00 -18.00 -18.00 nan nan
6 /
6: nan nan
nan nan nan
7 / 7:
-18.00 -18.00 -18.00 -18.00 -18.00
8 / 8:
-18.00 -18.00 -18.00 -18.00 -18.00
9 / 9:
-18.00 -18.00 -18.00 -18.00 -18.00
10 / 10:
-18.00 -18.00
-18.00 -18.00 -18.00
11 /
11: nan -18.00 -18.00 -18.00 -18.00
12 / 12:
-18.00 -18.00
-18.00 -18.00 -18.00
Regards,
Yun
Hi Yun,
Yesterday you got some ideas about how to reshape the grid of the
variable SST
to be able to plot it. Another method would be to tell Ferret the
order
of the axes when you open the file. I think this is what you need
yes? use/order=xyt HADISST_1_1_sst_2004.nc
yes? show grid sst
If there are missing SST values this should not be a problem. What plot
command
are you using? What error message are you seeing?
Yun.Li@xxxxxxxx wrote:
Hi
Ansley,
I am using
ferret/5.81. I have
another problem to using “Fill/L=7 SST” to make ferret plots
because there are missing SSTs at some grids in the
HADISST_1_1_SST_2004.nc. Can you please let me know how to plot SST
image
when there are missing values on some grid.
The output
of ncdump of
HADISST_1_1_SST_2004.nc
are as follows.
sst>
ncdump -h HADISST_1_1_SST_2004.nc
netcdf
HADISST_1_1_SST_2004 {
dimensions:
nr = UNLIMITED ; // (12 currently)
in = 360 ;
no = 180 ;
variables:
double t(nr) ;
tunics = "years" ;
t:long_name = "Fake time" ;
t:coordinate_type = "time" ;
float lon(in) ;
lon:units = "degrees_east" ;
lon:long_name = "Longitude" ;
lon:coordinate_type = "longitude" ;
float lat(no) ;
lat:units = "degrees_north" ;
lat:long_name = "Latitude" ;
lat:coordinate_type = "latitude" ;
float sst(nr, no, in) ;
sst:long_name
= "SST" ;
sst:units = "deg" ;
sst:missing_value = -999.f ;
sst:coordinates = "t, lat, lon" ;
// global
attributes:
:Conventions = "CMR/Timeseries" ;
}
Yun Li
Hi Yun Li,
If the name of the "Fake Time" axis is T, then Ferret should detect
that axis as an axis in the T direction. Could you tell me what
version
of Ferret are you using, and could you send me the output of the Unix
ncdump
command
ncdump -h HADISST_1_1_SST_2004.nc
Ansley
Yun.Li@xxxxxxxx wrote:
Dear All,
I have a copy of monthly HADISST_1_1_SST_2004.nc as follows.
yes? use HADISST_1_1_SST_2004.nc
yes? show data
currently SET data sets:
1> ./HADISST_1_1_SST_2004.nc (default)
name title I J K
L
T Fake time 1:12 ... ...
...
LON Longitude 1:360 ... ...
...
LAT Latitude 1:180 ... ...
...
SST SST 1:360 1:180 1:12
...
Ferret cannot draw the plots properly due to the incorrect grid settings
(K =1:12). Could you please help me to transform the grid as the correct
format (L=1:12)? That is,
show data
currently SET data sets:
1> ./HADISST_1_1_SST_2004.nc (default)
name title I J K
L
SST SST 1:360 1:180 ...
1:12
Regards,
Yun Li
CSIRO, Australia