Re: [netcdfgroup] stupid question: how do I dump just the contents of a variable from a netCDF file?

• To: Dave Allured <dave.allured@xxxxxxxx>
• Subject: Re: [netcdfgroup] stupid question: how do I dump just the contents of a variable from a netCDF file?
• From: "Lynnes, Christopher S. (GSFC-6102)" <christopher.s.lynnes@xxxxxxxx>
• Date: Thu, 20 Sep 2012 07:09:51 -0500

On Sep 19, 2012, at 6:45 PM, Dave Allured wrote:

> Not a stupid question, but a frequent one.  To get just ascii for one
> variable, one datum per line, no indices or labels (following Pedro's
> suggestion to use ncks):
> 
> ncks -s '%13.9f\n' -C -H -d time,0,3 -v precip precip.ens01.nc

Thx!  Would have taken me some time to arrive at that…


>  0.000002705
>  0.000002711
>  0.000002716
>  0.000002711
>  0.000002707
> <snip>
> 
> This follows one of the lower examples in the ncks on-line
> documentation ( http://nco.sourceforge.net/nco.html ).  Notes:
> 
> 1.  Behavior may change with NCO versions.  I had to add -H to get
> output matching the example on the website.  My NCO/ncks version is
> 4.2.0.
> 
> 2.  One advantage of ncks is that you can specify a printf-style
> format string to get the desired format and significant digits, as
> shown here.
> 
> 3.  Another advantage is that you can use NCO's -d option to subscript
> along any dimensions needed (hyperslabs).  Also as shown above.
> 
> 4.  Careful, the \n in the format string is critical when using -s.
> This also results in an extra blank line at the end of the output.
> 
> --Dave
> 
> On Wed, Sep 19, 2012 at 3:09 PM, Pedro Vicente
> <pedro.vicente@xxxxxxxxxxxxxxxxxx> wrote:
>>>> I just want either a single-column ascii output of the data values or a
>>>> binary dump that I can run od on to get same.
>> 
>> 
>> $ ncdump -v data example_000.nc
>> $ ncks -v data example_000.nc
>> $ h5dump -d data example_000.nc
>> 
>> 
>> below outputs for each, it seems ncks is what you want
>> 
>> 
>> $ ncdump -v data example_000.nc
>> 
>> netcdf example_000 {
>> dimensions:
>>       x = 4 ;
>>       y = 6 ;
>> variables:
>>       float data(x, y) ;
>>       float x(x) ;
>>       float y(y) ;
>> data:
>> 
>> data =
>> 0, 1, 2, 3, 4, 5,
>> 6, 7, 8, 9, 10, 11,
>> 12, 13, 14, 15, 16, 17,
>> 18, 19, 20, 21, 22, 23 ;
>> }
>> 
>> 
>> $ ncks -v data example_000.nc
>> 
>> data: type NC_FLOAT, 2 dimensions, 0 attributes, chunked? no, compressed?
>> no, packed? no, ID = 0
>> data size (in RAM) is 4*6*sizeof(NC_FLOAT) = 24*4 = 96 bytes
>> data dimension 0: x, size = 4 NC_FLOAT, dim. ID = 0 (Coordinate dimension)
>> data dimension 1: y, size = 6 NC_FLOAT, dim. ID = 1 (Coordinate dimension)
>> 
>> x: type NC_FLOAT, 1 dimension, 0 attributes, chunked? no, compressed? no,
>> packed? no, ID = 1
>> x size (in RAM) is 4*sizeof(NC_FLOAT) = 4*4 = 16 bytes
>> x dimension 0: x, size = 4 NC_FLOAT, dim. ID = 0 (Coordinate dimension)
>> 
>> y: type NC_FLOAT, 1 dimension, 0 attributes, chunked? no, compressed? no,
>> packed? no, ID = 2
>> y size (in RAM) is 6*sizeof(NC_FLOAT) = 6*4 = 24 bytes
>> y dimension 0: y, size = 6 NC_FLOAT, dim. ID = 1 (Coordinate dimension)
>> 
>> x[0]=0 y[0]=0 data[0]=0
>> x[0]=0 y[1]=10 data[1]=1
>> x[0]=0 y[2]=20 data[2]=2
>> x[0]=0 y[3]=30 data[3]=3
>> x[0]=0 y[4]=40 data[4]=4
>> x[0]=0 y[5]=50 data[5]=5
>> x[1]=10 y[0]=0 data[6]=6
>> x[1]=10 y[1]=10 data[7]=7
>> x[1]=10 y[2]=20 data[8]=8
>> x[1]=10 y[3]=30 data[9]=9
>> x[1]=10 y[4]=40 data[10]=10
>> x[1]=10 y[5]=50 data[11]=11
>> x[2]=20 y[0]=0 data[12]=12
>> x[2]=20 y[1]=10 data[13]=13
>> x[2]=20 y[2]=20 data[14]=14
>> x[2]=20 y[3]=30 data[15]=15
>> x[2]=20 y[4]=40 data[16]=16
>> x[2]=20 y[5]=50 data[17]=17
>> x[3]=30 y[0]=0 data[18]=18
>> x[3]=30 y[1]=10 data[19]=19
>> x[3]=30 y[2]=20 data[20]=20
>> x[3]=30 y[3]=30 data[21]=21
>> x[3]=30 y[4]=40 data[22]=22
>> x[3]=30 y[5]=50 data[23]=23
>> 
>> x[0]=0
>> x[1]=10
>> x[2]=20
>> x[3]=30
>> 
>> y[0]=0
>> y[1]=10
>> y[2]=20
>> y[3]=30
>> y[4]=40
>> y[5]=50
>> 
>> 
>> 
>> $ h5dump -d data example_000.nc
>> 
>> HDF5 "example_000.nc" {
>> DATASET "data" {
>>  DATATYPE  H5T_IEEE_F32LE
>>  DATASPACE  SIMPLE { ( 4, 6 ) / ( 4, 6 ) }
>>  DATA {
>>  (0,0): 0, 1, 2, 3, 4, 5,
>>  (1,0): 6, 7, 8, 9, 10, 11,
>>  (2,0): 12, 13, 14, 15, 16, 17,
>>  (3,0): 18, 19, 20, 21, 22, 23
>>  }
>>  ATTRIBUTE "DIMENSION_LIST" {
>>     DATATYPE  H5T_VLEN { H5T_REFERENCE { H5T_STD_REF_OBJECT }}
>>     DATASPACE  SIMPLE { ( 2 ) / ( 2 ) }
>>     DATA {
>>     (0): (DATASET 539 /x ), (DATASET 823 /y )
>>     }
>>  }
>> }
>> }
>> 
>> 
>> 
>> ----------------------
>> Pedro Vicente
>> pedro.vicente@xxxxxxxxxxxxxxxxxx
>> http://www.space-research.org/
>> 
>> 
>> ----- Original Message ----- From: "Lynnes, Christopher S. (GSFC-6102)"
>> <christopher.s.lynnes@xxxxxxxx>
>> To: "NetCDF Group List" <netcdfgroup@xxxxxxxxxxxxxxxx>
>> Sent: Wednesday, September 19, 2012 3:35 PM
>> Subject: [netcdfgroup] stupid question: how do I dump just the contents of a
>> variable from a netCDF file?
>> 
>> 
>>> I mean, no header info, no data attributes, no coordinate indices. I just
>>> want either a single-column ascii output of the data values or a binary dump
>>> that I can run od on to get same.
>>> 
>>> Can't seem to find the right combo of args in either ncdump or h5dump.. .
>>> --
>>> Dr. Christopher Lynnes     NASA/GSFC, Code 610.2    phone: 301-614-5185

--
Dr. Christopher Lynnes     NASA/GSFC, Code 610.2    phone: 301-614-5185

• References:
  • [netcdfgroup] stupid question: how do I dump just the contents of a variable from a netCDF file?
    • From: Lynnes, Christopher S. (GSFC-6102)
  • Re: [netcdfgroup] stupid question: how do I dump just the contents of a variable from a netCDF file?
    • From: Pedro Vicente
  • Re: [netcdfgroup] stupid question: how do I dump just the contents of a variable from a netCDF file?
    • From: Dave Allured