Hello,
I finally managed to compile the bm_file program. I am trying to modify
chunk parameters in my NetCDF4 file but I do not clearly understand how
to write the good arguments for doing that. Here is an example:
## file.nc: NetCDF4 file of dimension X[5], Y[10] and Time[12].
ncdump -h file.nc
netcdf file.nc {
dimensions:
X = 5 ;
Y = 10 ;
Time = UNLIMITED ; // (12 currently)
variables:
int X(X) ;
X:units = "meters" ;
X:long_name = "X" ;
int Y(Y) ;
Y:units = "meters" ;
Y:long_name = "Y" ;
int Time(Time) ;
Time:units = "seconds" ;
Time:long_name = "Time" ;
Time:calendar = "gregorian" ;
float var3d(Time, Y, X) ;
var3d:units = "units" ;
var3d:missing_value = 1.e+30f ;
}
## Following the documentation (run_bm_elena.sh), I have tried to modify
chunk parameters, as X[1], Y[5], Time [12] :
bm_file -d -f 3 -o file_out2.nc -c 0:-1:0:1:5:20 file.nc
file_out2.nc is well created but I am not sure the chunking parameters
have been correctly set... How can I know ?
I am not sure I understand the meaning of -c arguments (i.e. [-c
V:Z:S:C:C:C[,V:Z:S:C:C:C, etc.]] Deflate, shuffle, and chunking
parameters for vars). Please could you explain ?
Thanks in advance.
Clement Tisseuil
On 20/10/2010 18:29, clement wrote:
Hello,
I would like to test the bm_file program to modify chunck parameters
from my NetCDF4 files. Unfortunately, my "make check" step fails when
compiling with the "--enable-benchmarks" option. Since my platform
does not allow me to download any external files, I suspect the
problem comes from the sample data files that has to be fetched from
the Unidata ftp . If so, is there a way to build the program without
using these sample data files ?
Thanks in advance for your assistance.
Regards,
Clement Tisseuil
--
Clément Tisseuil
