Issue in inputs reding for the ENSSTAT2 and ENSVAL functions
Added by Lorenzo Sangelantoni over 3 years ago
Hi all,
I am experiencing issues with enscrps, ensrkhisttime and ensrkhispace functions in CDO versions 1.9.3 and 1.9.8.
e.g., for the enscrps function (valid also for the ensrkhisttime and ensrkhispace functions):
the inputs provided to the functions are represented by gridded observational (reference product, "rvar") and simulated (members of an ensemble, "regcm_d2_M*_srf_yearmeanAnomaly_ivar") [.nc] files having the same dimensions (time,lon,lat) and same variable name (t2m) on a single surface level.
running:
<< cdo enscrps rvar regcm_d2_M*_srf_yearmeanAnomaly_ivar ofile >>
I get:
"[bt] Execution path:
[bt] cdo() [0x74bf3d]
[bt] cdo() [0x74c29f]
[bt] cdo() [0x74c959]
[bt] cdo() [0x7438b9]
[bt] cdo() [0x4beaaa]
[bt] cdo() [0x6d69d3]
[bt] cdo() [0x4450f0]
[bt] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0) [0x7f9ac5318840]
[bt] cdo() [0x449f09]
ERROR, reshGetElem, resource_handle.c, line 406, called from grid_to_pointer
errorString: "Error while trying to resolve the ID "gridID" in `grid_to_pointer()`: the value is CDI_UNDEFID (= -1).
This is most likely the result of a failed earlier call. Please check the IDs returned by CDI."
The same syntax successfully run with previous CDO versions e.g., CDO/1.7.2-intel-2016a.
Best regards,
Lorenzo
Replies (4)
RE: Issue in inputs reding for the ENSSTAT2 and ENSVAL functions - Added by Ralf Mueller over 3 years ago
hi!
the rvar file has a 3036 point grid whereas the member files on have a 513 point grid. With cdo-1.9.8 I get:
cdo enscrps rvar regcm_d2_M*_srf_yearmeanAnomaly_ivar ofile cdo enscrps (Abort): Grid size of the input parameter tg do not match!
could be that cdo-1.7.2 is lacking a proper error handling for this case. But I am sure the files don't fit to the operation to be done.
hth
ralf
RE: Issue in inputs reding for the ENSSTAT2 and ENSVAL functions - Added by Lorenzo Sangelantoni over 3 years ago
Many thanks Ralf,
excuse me, I attached the wrong rvar file.
However, assigning the same grid points number (513) to the rvar the function enscrps does not work in my case.
This is can be easily seen considering one of the ensemble member as a reference variable (rvar). Maybe I am missing something.
Bests,
Lorenzo
RE: Issue in inputs reding for the ENSSTAT2 and ENSVAL functions - Added by Ralf Mueller over 3 years ago
ah, ok!
I tested your call without rvar with a CDO binary build with clang - and apparently it worked. even a build with gcc10 forked, but not with gcc-9.3. Seems to be related to certain optimizations used for compiling.
can you uploads the output of 'cdo -V'?
RE: Issue in inputs reding for the ENSSTAT2 and ENSVAL functions - Added by Lorenzo Sangelantoni over 3 years ago
Here it is the 'cdo -V' output
Climate Data Operators version 1.9.8 (https://mpimet.mpg.de/cdo)
System: x86_64-pc-linux-gnu
CXX Compiler: g++ -std=gnu++11 -g -O2 -fopenmp
CXX version : g++ (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
C Compiler: gcc -g -O2 -fopenmp
C version : gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
F77 Compiler: f77 -g -O2
F77 version : unknown
Features: 7GB 4threads C++11 Fortran DATA PTHREADS OpenMP HDF5 NC4/HDF5/threadsafe OPeNDAP UDUNITS2 SSE2
Libraries: HDF5/1.12.0
Filetypes: srv ext ieg grb1 grb2 nc1 nc2 nc4 nc4c nc5
CDI library version : 1.9.8
cgribex library version : 1.9.4
ecCodes library version : 2.7.0
NetCDF library version : 4.7.4 of May 11 2020 19:29:28 $
hdf5 library version : 1.12.0 threadsafe
exse library version : 1.4.1
FILE library version : 1.8.3
Lorenzo