Hi Michael, Sorry to have taken so long to answer your question. > I am trying to read some data from a file called TS.nc using some examples I > found (http://www.unidata.ucar.edu/software/netcdf/examples/programs/), > however I am having some issues. The code I am using is copied below. > Below is the result of an ncdump -v TS TS.nc command: > fr0103en: /ptmp/arndtma/Comp_out(=0)$ ncdump -v TS TS.nc | more > netcdf TS { > dimensions: > lat = 384 ; > lon = 576 ; > time = UNLIMITED ; // (133 currently) > variables: > double lat(lat) ; > lat:long_name = "latitude" ; > lat:units = "degrees_north" ; > double lon(lon) ; > lon:long_name = "longitude" ; > lon:units = "degrees_east" ; > double time(time) ; > time:long_name = "time" ; > time:units = "days since 0001-01-01 00:00:00" ; > time:calendar = "noleap" ; > time:bounds = "time_bnds" ; > float TS(time, lat, lon) ; > TS:units = "K" ; > TS:long_name = "Surface temperature (radiative)" ; > TS:cell_method = "time: mean" ; > > // global attributes: > :Conventions = "CF-1.0" ; > :source = "CAM" ; > :case = "LRC01" ; > :title = "UNSET" ; > :logname = "" ; > :host = "" ; > :Version = "$Name$" ; > :revision_Id = "$Id$" ; > :initial_file = > "/lustre/scratch/proj/ccsm/inputdata/atm/cam/inic/fv/cami_1980-01-01_0.47x0.63_L26_c071226.nc" > ; > :topography_file = > "/lustre/scratch/proj/ccsm/inputdata/atm/cam/topo/USGS_gtopo30_0.47x0.63_remap_c061106.nc" > ; > :history = "Mon Oct 17 14:30:23 2011: ncrcat -v TS > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0079-12.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.c > am2.h0.0080-01.nc /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-02.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-03.nc > /ptmp/dennis/LRC01/atm/hist/LRC > 01.cam2.h0.0080-04.nc /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-05.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-06.nc > /ptmp/dennis/LRC01/atm/hist > /LRC01.cam2.h0.0080-07.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-08.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-09.nc /ptmp/dennis/LRC01/atm/ > hist/LRC01.cam2.h0.0080-10.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-11.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0080-12.nc /ptmp/dennis/LRC01/ > atm/hist/LRC01.cam2.h0.0081-01.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0081-02.nc > /ptmp/dennis/LRC01/atm/hist/LRC01.cam2.h0.0081-03.nc /ptmp/dennis/LR > C01/atm/hist/LRC01.cam2.h0.0081-04.nc /ptmp/dennis/LRC01/atm/hist/ > ... > As you can see I created this file by extracting each variable from the > original nc files using the ncrcat command. > > > The error (Segmentation fault) comes when the code tries to read in the > actual data and therefore I am convinced I am missing something in the length > or structure of the data itself. The nc_open work fine, the nc_inq_varid > works fine, it is only when I try to use the nc_get_var_float that the seg > fault occurs. You're correct, the problem is that you're trying to read all of the TS variable (133 * 384 * 576 values) into an array that can only hold 6 *12 = 72 values. The call to nc_get_var_float(ncid, varid, &data_in[0][0]))) reads all the data from the variable identified by varid into the location you've specified, the address of the data_in array. If you really want to try to read all the data for TS into memory at once, you'll have to declare data_in to be a much larger array, e.g. #define NT 133 float data_in[NT][NX][NY]; But you might only need one slice at a time of the TS data, for example all the values at a specific time. For this, you'll need to use the function nc_get_vara_float() instead of nc_get_var_float(), and you'll need to provide 2 more arguments specifying where the start of the slab of data is (e.g. the 10th time would start at indices [9][0][0] so the "start" argument would be {9, 0, 0}) and how large the slab is on each side (e.g the whole 384 by 576 slice for 1 time would mean an "edges" argument of {1,384,576}). There are other nc_get_var* functions you can use to get just one value at a time, or various kinds of subsets of the data from the disk into memory where your program can do something with it, e.g. nc_get_var1_float(), nc_get_vars_float(), etc., but I expect nc_get_vara_float() is what you want to use. --Russ > Thanks in advance, > > Michael > > > ===== > 1 #include <stdlib.h> > 2 #include <stdio.h> > 3 #include <netcdf.h> > 4 > 5 > 6 /* This is the name of the data file we will read. */ > 7 > 8 #define FILE_NAME "simple_xy.nc" > 9 > 10 /* We are reading 2D data, a 6 x 12 grid. */ > 11 > 12 #define NX 6 > 13 #define NY 12 > 14 > 15 /* Handle errors by printing an error message and exiting with a > 16 * non-zero status. */ > 17 > 18 #define ERRCODE 2 > 19 #define ERR(e) {printf("Error: %s\n", nc_strerror(e)); > exit(ERRCODE);} > 20 > 21 int main() > 22 > 23 { > 24 /* This will be the netCDF ID for the file and data variable. */ > 25 int ncid, varid; > 26 > 27 float data_in[NX][NY]; > 28 > 29 /* Loop indexes, and error handling. */ > 30 int x, y, retval; > 31 > 32 // /* Open the file. NC_NOWRITE tells netCDF we want read-only access > 33 // * to the file.*/ > 34 // > 35 if ((retval = nc_open(FILE_NAME, NC_NOWRITE, &ncid))) > 36 ERR(retval); > 37 > 38 // /* Get the varid of the data variable, based on its name. */ > 39 if ((retval = nc_inq_varid(ncid, "TS", &varid))) > 40 ERR(retval); > 41 printf("varid is : %i \n",varid); > 42 // /* Read the data. */ > 43 if ((retval = nc_get_var_float(ncid, varid, &data_in[0][0]))) > 44 ERR(retval); > 45 > 46 > 47 // /* Close the file, freeing all resources. */ > 48 // if ((retval = nc_close(ncid))) > 49 // ERR(retval); > 50 > 51 printf("*** SUCCESS reading example file %s!\n", FILE_NAME); > 52 return 0; > 53 } > ===== > > > Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: ZBT-271812 Department: Support netCDF Priority: Critical Status: Closed
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.