The NetCDF C Interface Guide **************************** This document describes the C interface to the netCDF library; it applies to netCDF version 4.0.1-beta1 and was last updated on 6 October 2008. For a complete description of the netCDF format and utilities see *note The NetCDF Users Guide: (netcdf)Top. 1 Use of the NetCDF Library *************************** You can use the netCDF library without knowing about all of the netCDF interface. If you are creating a netCDF dataset, only a handful of routines are required to define the necessary dimensions, variables, and attributes, and to write the data to the netCDF dataset. (Even less are needed if you use the ncgen utility to create the dataset before running a program using netCDF library calls to write data.) Similarly, if you are writing software to access data stored in a particular netCDF object, only a small subset of the netCDF library is required to open the netCDF dataset and access the data. Authors of generic applications that access arbitrary netCDF datasets need to be familiar with more of the netCDF library. In this chapter we provide templates of common sequences of netCDF calls needed for common uses. For clarity we present only the names of routines; omit declarations and error checking; omit the type-specific suffixes of routine names for variables and attributes; indent statements that are typically invoked multiple times; and use ... to represent arbitrary sequences of other statements. Full parameter lists are described in later chapters. 1.1 Creating a NetCDF Dataset ============================= Here is a typical sequence of netCDF calls used to create a new netCDF dataset: nc_create /* create netCDF dataset: enter define mode */ ... nc_def_dim /* define dimensions: from name and length */ ... nc_def_var /* define variables: from name, type, ... */ ... nc_put_att /* put attribute: assign attribute values */ ... nc_enddef /* end definitions: leave define mode */ ... nc_put_var /* provide values for variables */ ... nc_close /* close: save new netCDF dataset */ Only one call is needed to create a netCDF dataset, at which point you will be in the first of two netCDF modes. When accessing an open netCDF dataset, it is either in define mode or data mode. In define mode, you can create dimensions, variables, and new attributes, but you cannot read or write variable data. In data mode, you can access data and change existing attributes, but you are not permitted to create new dimensions, variables, or attributes. One call to nc_def_dim is needed for each dimension created. Similarly, one call to nc_def_var is needed for each variable creation, and one call to a member of the nc_put_att family is needed for each attribute defined and assigned a value. To leave define mode and enter data mode, call nc_enddef. Once in data mode, you can add new data to variables, change old values, and change values of existing attributes (so long as the attribute changes do not require more storage space). Single values may be written to a netCDF variable with one of the members of the nc_put_var1 family, depending on what type of data you have to write. All the values of a variable may be written at once with one of the members of the nc_put_var family. Arrays or array cross-sections of a variable may be written using members of the nc_put_vara family. Subsampled array sections may be written using members of the nc_put_vars family. Mapped array sections may be written using members of the nc_put_varm family. (Subsampled and mapped access are general forms of data access that are explained later.) Finally, you should explicitly close all netCDF datasets that have been opened for writing by calling nc_close. By default, access to the file system is buffered by the netCDF library. If a program terminates abnormally with netCDF datasets open for writing, your most recent modifications may be lost. This default buffering of data is disabled by setting the NC_SHARE flag when opening the dataset. But even if this flag is set, changes to attribute values or changes made in define mode are not written out until nc_sync or nc_close is called. 1.2 Reading a NetCDF Dataset with Known Names ============================================= Here we consider the case where you know the names of not only the netCDF datasets, but also the names of their dimensions, variables, and attributes. (Otherwise you would have to do "inquire" calls.) The order of typical C calls to read data from those variables in a netCDF dataset is: nc_open /* open existing netCDF dataset */ ... nc_inq_dimid /* get dimension IDs */ ... nc_inq_varid /* get variable IDs */ ... nc_get_att /* get attribute values */ ... nc_get_var /* get values of variables */ ... nc_close /* close netCDF dataset */ First, a single call opens the netCDF dataset, given the dataset name, and returns a netCDF ID that is used to refer to the open netCDF dataset in all subsequent calls. Next, a call to nc_inq_dimid for each dimension of interest gets the dimension ID from the dimension name. Similarly, each required variable ID is determined from its name by a call to nc_inq_varid Once variable IDs are known, variable attribute values can be retrieved using the netCDF ID, the variable ID, and the desired attribute name as input to a member of the nc_get_att family (typically nc_get_att_text or nc_get_att_double) for each desired attribute. Variable data values can be directly accessed from the netCDF dataset with calls to members of the nc_get_var1 family for single values, the nc_get_var family for entire variables, or various other members of the nc_get_vara, nc_get_vars, or nc_get_varm families for array, subsampled or mapped access. Finally, the netCDF dataset is closed with nc_close. There is no need to close a dataset open only for reading. 1.3 Reading a netCDF Dataset with Unknown Names =============================================== It is possible to write programs (e.g., generic software) which do such things as processing every variable, without needing to know in advance the names of these variables. Similarly, the names of dimensions and attributes may be unknown. Names and other information about netCDF objects may be obtained from netCDF datasets by calling inquire functions. These return information about a whole netCDF dataset, a dimension, a variable, or an attribute. The following template illustrates how they are used: nc_open /* open existing netCDF dataset */ ... nc_inq /* find out what is in it */ ... nc_inq_dim /* get dimension names, lengths */ ... nc_inq_var /* get variable names, types, shapes */ ... nc_inq_attname /* get attribute names */ ... nc_inq_att /* get attribute types and lengths */ ... nc_get_att /* get attribute values */ ... nc_get_var /* get values of variables */ ... nc_close /* close netCDF dataset */ As in the previous example, a single call opens the existing netCDF dataset, returning a netCDF ID. This netCDF ID is given to the nc_inq routine, which returns the number of dimensions, the number of variables, the number of global attributes, and the ID of the unlimited dimension, if there is one. All the inquire functions are inexpensive to use and require no I/O, since the information they provide is stored in memory when a netCDF dataset is first opened. Dimension IDs use consecutive integers, beginning at 0. Also dimensions, once created, cannot be deleted. Therefore, knowing the number of dimension IDs in a netCDF dataset means knowing all the dimension IDs: they are the integers 0, 1, 2, ...up to the number of dimensions. For each dimension ID, a call to the inquire function nc_inq_dim returns the dimension name and length. Variable IDs are also assigned from consecutive integers 0, 1, 2, ... up to the number of variables. These can be used in nc_inq_var calls to find out the names, types, shapes, and the number of attributes assigned to each variable. Once the number of attributes for a variable is known, successive calls to nc_inq_attname return the name for each attribute given the netCDF ID, variable ID, and attribute number. Armed with the attribute name, a call to nc_inq_att returns its type and length. Given the type and length, you can allocate enough space to hold the attribute values. Then a call to a member of the nc_get_att family returns the attribute values. Once the IDs and shapes of netCDF variables are known, data values can be accessed by calling a member of the nc_get_var1 family for single values, or members of the nc_get_var, nc_get_vara, nc_get_vars, or nc_get_varm for various kinds of array access. 1.4 Adding New Dimensions, Variables, Attributes ================================================ An existing netCDF dataset can be extensively altered. New dimensions, variables, and attributes can be added or existing ones renamed, and existing attributes can be deleted. Existing dimensions, variables, and attributes can be renamed. The following code template lists a typical sequence of calls to add new netCDF components to an existing dataset: nc_open /* open existing netCDF dataset */ ... nc_redef /* put it into define mode */ ... nc_def_dim /* define additional dimensions (if any) */ ... nc_def_var /* define additional variables (if any) */ ... nc_put_att /* define additional attributes (if any) */ ... nc_enddef /* check definitions, leave define mode */ ... nc_put_var /* provide values for new variables */ ... nc_close /* close netCDF dataset */ A netCDF dataset is first opened by the nc_open call. This call puts the open dataset in data mode, which means existing data values can be accessed and changed, existing attributes can be changed (so long as they do not grow), but nothing can be added. To add new netCDF dimensions, variables, or attributes you must enter define mode, by calling nc_redef. In define mode, call nc_def_dim to define new dimensions, nc_def_var to define new variables, and a member of the nc_put_att family to assign new attributes to variables or enlarge old attributes. You can leave define mode and reenter data mode, checking all the new definitions for consistency and committing the changes to disk, by calling nc_enddef. If you do not wish to reenter data mode, just call nc_close, which will have the effect of first calling nc_enddef. Until the nc_enddef call, you may back out of all the redefinitions made in define mode and restore the previous state of the netCDF dataset by calling nc_abort. You may also use the nc_abort call to restore the netCDF dataset to a consistent state if the call to nc_enddef fails. If you have called nc_close from definition mode and the implied call to nc_enddef fails, nc_abort will automatically be called to close the netCDF dataset and leave it in its previous consistent state (before you entered define mode). For netCDF-4/HDF5 format files, define mode is still important, but the user does not have to called nc_enddef - it is called automatically when needed. It may also be called by the user. In netCDF-4/HDF5 files, there are some settings which can only be modified during the very first define mode of the file. For example the compression level of a variable may be set only after the nc_def_var call and before the next nc_enddef call, whether it is called by the user explicitly, or when the user tries to read of write some data. At most one process should have a netCDF dataset open for writing at one time. The library is designed to provide limited support for multiple concurrent readers with one writer, via disciplined use of the nc_sync function and the NC_SHARE flag. If a writer makes changes in define mode, such as the addition of new variables, dimensions, or attributes, some means external to the library is necessary to prevent readers from making concurrent accesses and to inform readers to call nc_sync before the next access. 1.5 Error Handling ================== The netCDF library provides the facilities needed to handle errors in a flexible way. Each netCDF function returns an integer status value. If the returned status value indicates an error, you may handle it in any way desired, from printing an associated error message and exiting to ignoring the error indication and proceeding (not recommended!). For simplicity, the examples in this guide check the error status and call a separate function, handle_err(), to handle any errors. One possible definition of handle_err() can be found within the documentation of nc_strerror (*note nc_strerror::). The nc_strerror function is available to convert a returned integer error status into an error message string. Occasionally, low-level I/O errors may occur in a layer below the netCDF library. For example, if a write operation causes you to exceed disk quotas or to attempt to write to a device that is no longer available, you may get an error from a layer below the netCDF library, but the resulting write error will still be reflected in the returned status value. 1.6 Compiling and Linking with the NetCDF Library ================================================= Details of how to compile and link a program that uses the netCDF C or FORTRAN interfaces differ, depending on the operating system, the available compilers, and where the netCDF library and include files are installed. Nevertheless, we provide here examples of how to compile and link a program that uses the netCDF library on a Unix platform, so that you can adjust these examples to fit your installation. Every C file that references netCDF functions or constants must contain an appropriate #include statement before the first such reference: #include Unless the netcdf.h file is installed in a standard directory where the C compiler always looks, you must use the -I option when invoking the compiler, to specify a directory where netcdf.h is installed, for example: cc -c -I/usr/local/netcdf/include myprogram.c Alternatively, you could specify an absolute path name in the #include statement, but then your program would not compile on another platform where netCDF is installed in a different location. Unless the netCDF library is installed in a standard directory where the linker always looks, you must use the -L and -l options to link an object file that uses the netCDF library. For example: cc -o myprogram myprogram.o -L/usr/local/netcdf/lib -lnetcdf Alternatively, you could specify an absolute path name for the library: cc -o myprogram myprogram.o -l/usr/local/netcdf/lib/libnetcdf.a 2 Datasets ********** This chapter presents the interfaces of the netCDF functions that deal with a netCDF dataset or the whole netCDF library. A netCDF dataset that has not yet been opened can only be referred to by its dataset name. Once a netCDF dataset is opened, it is referred to by a netCDF ID, which is a small non-negative integer returned when you create or open the dataset. A netCDF ID is much like a file descriptor in C or a logical unit number in FORTRAN. In any single program, the netCDF IDs of distinct open netCDF datasets are distinct. A single netCDF dataset may be opened multiple times and will then have multiple distinct netCDF IDs; however at most one of the open instances of a single netCDF dataset should permit writing. When an open netCDF dataset is closed, the ID is no longer associated with a netCDF dataset. Functions that deal with the netCDF library include: * Get version of library. * Get error message corresponding to a returned error code. The operations supported on a netCDF dataset as a single object are: * Create, given dataset name and whether to overwrite or not. * Open for access, given dataset name and read or write intent. * Put into define mode, to add dimensions, variables, or attributes. * Take out of define mode, checking consistency of additions. * Close, writing to disk if required. * Inquire about the number of dimensions, number of variables, number of global attributes, and ID of the unlimited dimension, if any. * Synchronize to disk to make sure it is current. * Set and unset nofill mode for optimized sequential writes. * After a summary of conventions used in describing the netCDF interfaces, the rest of this chapter presents a detailed description of the interfaces for these operations. 2.1 NetCDF Library Interface Descriptions ========================================= Each interface description for a particular netCDF function in this and later chapters contains: * a description of the purpose of the function; * a C function prototype that presents the type and order of the formal parameters to the function; * a description of each formal parameter in the C interface; * a list of possible error conditions; and * an example of a C program fragment calling the netCDF function (and perhaps other netCDF functions). The examples follow a simple convention for error handling, always checking the error status returned from each netCDF function call and calling a handle_error function in case an error was detected. For an example of such a function, see *note nc_strerror::. 2.2 Parallel Access for NetCDF Files ==================================== For netCDF-4 files only, parallel read/write access is possible on systems which support it, and only if parallel HDF5 was installed on the system before netCDF, and only if the HDF5 parallel compiler was used during the netCDF configure. (Parallel HDF5 requires the MPI library). To use parallel access, open or create the file with nc_open_par (see *note nc_open_par::) or nc_create_par (see *note nc_create_par::). Only netCDF-4 files can be opened or created for parallel access. The following example shows the creation of a file using parallel access, and how a program might write data to such a file. #include "netcdf.h" #include #include #include "hdf5.h" #include #include #define BAIL(e) do { \ printf("Bailing out in file %s, line %d, error:%s.\n", __FILE__, __LINE__, nc_strerror(e)); \ return e; \ } while (0) #define FILE "test_par.nc" #define NDIMS 2 #define DIMSIZE 24 #define QTR_DATA (DIMSIZE*DIMSIZE/4) #define NUM_PROC 4 int main(int argc, char **argv) { /* MPI stuff. */ int mpi_namelen; char mpi_name[MPI_MAX_PROCESSOR_NAME]; int mpi_size, mpi_rank; MPI_Comm comm = MPI_COMM_WORLD; MPI_Info info = MPI_INFO_NULL; /* Netcdf-4 stuff. */ int ncid, v1id, dimids[NDIMS]; size_t start[NDIMS], count[NDIMS]; int data[DIMSIZE*DIMSIZE], j, i, res; /* Initialize MPI. */ MPI_Init(&argc,&argv); MPI_Comm_size(MPI_COMM_WORLD, &mpi_size); MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank); MPI_Get_processor_name(mpi_name, &mpi_namelen); printf("mpi_name: %s size: %d rank: %d\n", mpi_name, mpi_size, mpi_rank); /* Create a parallel netcdf-4 file. */ if ((res = nc_create_par(FILE, NC_NETCDF4|NC_MPIIO, comm, info, &ncid))) BAIL(res); /* Create two dimensions. */ if ((res = nc_def_dim(ncid, "d1", DIMSIZE, dimids))) BAIL(res); if ((res = nc_def_dim(ncid, "d2", DIMSIZE, &dimids[1]))) BAIL(res); /* Create one var. */ if ((res = nc_def_var(ncid, "v1", NC_INT, NDIMS, dimids, &v1id))) BAIL(res); if ((res = nc_enddef(ncid))) BAIL(res); /* Set up slab for this process. */ start[0] = mpi_rank * DIMSIZE/mpi_size; start[1] = 0; count[0] = DIMSIZE/mpi_size; count[1] = DIMSIZE; printf("mpi_rank=%d start[0]=%d start[1]=%d count[0]=%d count[1]=%d\n", mpi_rank, start[0], start[1], count[0], count[1]); /* Create phony data. We're going to write a 24x24 array of ints, in 4 sets of 144. */ printf("mpi_rank*QTR_DATA=%d (mpi_rank+1)*QTR_DATA-1=%d\n", mpi_rank*QTR_DATA, (mpi_rank+1)*QTR_DATA); for (i=mpi_rank*QTR_DATA; i<(mpi_rank+1)*QTR_DATA; i++) data[i] = mpi_rank; /*if ((res = nc_var_par_access(ncid, v1id, NC_COLLECTIVE))) BAIL(res);*/ if ((res = nc_var_par_access(ncid, v1id, NC_INDEPENDENT))) BAIL(res); /* Write slabs of phony data. */ if ((res = nc_put_vara_int(ncid, v1id, start, count, &data[mpi_rank*QTR_DATA]))) BAIL(res); /* Close the netcdf file. */ if ((res = nc_close(ncid))) BAIL(res); /* Shut down MPI. */ MPI_Finalize(); return 0; } 2.3 Get error message corresponding to error status: nc_strerror ================================================================ The function nc_strerror returns a static reference to an error message string corresponding to an integer netCDF error status or to a system error number, presumably returned by a previous call to some other netCDF function. The list of netCDF error status codes is available in the appropriate include file for each language binding. Usage ===== const char * nc_strerror(int ncerr); `ncerr' An error status that might have been returned from a previous call to some netCDF function. Errors ====== If you provide an invalid integer error status that does not correspond to any netCDF error message or or to any system error message (as understood by the system strerror function), nc_strerror returns a string indicating that there is no such error status. Example ======= Here is an example of a simple error handling function that uses nc_strerror to print the error message corresponding to the netCDF error status returned from any netCDF function call and then exit: #include ... void handle_error(int status) { if (status != NC_NOERR) { fprintf(stderr, "%s\n", nc_strerror(status)); exit(-1); } } 2.4 Get netCDF library version: nc_inq_libvers ============================================== The function nc_inq_libvers returns a string identifying the version of the netCDF library, and when it was built. Usage ===== const char * nc_inq_libvers(void); Errors ====== This function takes no arguments, and thus no errors are possible in its invocation. Example ======= Here is an example using nc_inq_libvers to print the version of the netCDF library with which the program is linked: #include ... printf("%s\n", nc_inq_libvers()); 2.5 Create a NetCDF Dataset: nc_create ====================================== This function creates a new netCDF dataset, returning a netCDF ID that can subsequently be used to refer to the netCDF dataset in other netCDF function calls. The new netCDF dataset opened for write access and placed in define mode, ready for you to add dimensions, variables, and attributes. A creation mode flag specifies: * whether to overwrite any existing dataset with the same name, * whether access to the dataset is shared, * whether this file should be in netCDF classic format (the default), the new 64-bit offset format (use NC_64BIT_OFFSET), or NC_NETCDF4 for a netCDF-4/HDF5 file. Usage ===== NOTE: When creating a netCDF-4 file HDF5 error reporting is turned off, if it is on. This doesn't stop the HDF5 error stack from recording the errors, it simply stops their display to the user through stderr. int nc_create (const char* path, int cmode, int *ncidp); `path' The file name of the new netCDF dataset. `cmode' The creation mode flag. The following flags are available: NC_NOCLOBBER, NC_SHARE, NC_64BIT_OFFSET, NC_NETCDF4, NC_CLASSIC_MODEL. Setting NC_NOCLOBBER means you do not want to clobber (overwrite) an existing dataset; an error (NC_EEXIST) is returned if the specified dataset already exists. The NC_SHARE flag is appropriate when one process may be writing the dataset and one or more other processes reading the dataset concurrently; it means that dataset accesses are not buffered and caching is limited. Since the buffering scheme is optimized for sequential access, programs that do not access data sequentially may see some performance improvement by setting the NC_SHARE flag. This flag is ignored for netCDF-4 files. (See below.) Setting NC_64BIT_OFFSET causes netCDF to create a 64-bit offset format file, instead of a netCDF classic format file. The 64-bit offset format imposes far fewer restrictions on very large (i.e. over 2 GB) data files. *Note Large File Support: (netcdf)Large File Support. A zero value (defined for convenience as NC_CLOBBER) specifies the default behavior: overwrite any existing dataset with the same file name and buffer and cache accesses for efficiency. The dataset will be in netCDF classic format. *Note NetCDF Classic Format Limitations: (netcdf)NetCDF Classic Format Limitations. Setting NC_NETCDF4 causes netCDF to create a HDF5/NetCDF-4 file. Setting NC_CLASSIC_MODEL causes netCDF to enforce the classic data model in this file. (This only has effect for netCDF-4/HDF5 files, as classic and 64-bit offset files always use the classic model.) When used with NC_NETCDF4, this flag ensures that the resulting netCDF-4/HDF5 file may never contain any new constructs from the enhanced data model. That is, it cannot contain groups, user defined types, multiple unlimited dimensions, or new atomic types. The advantage of this restriction is that such files are guarenteed to work with existing netCDF software. `ncidp' Pointer to location where returned netCDF ID is to be stored. Errors ====== nc_create returns the value NC_NOERR if no errors occurred. Possible causes of errors include: * Passing a dataset name that includes a directory that does not exist. * Specifying a dataset name of a file that exists and also specifying NC_NOCLOBBER. * Specifying a meaningless value for the creation mode. * Attempting to create a netCDF dataset in a directory where you don't have permission to create files. Return Codes ============ `NC_NOERR' No error. `NC_ENOMEM' System out of memory. `NC_EHDFERR' HDF5 error (netCDF-4 files only). `NC_EFILEMETA' Error writing netCDF-4 file-level metadata in HDF5 file. (netCDF-4 files only). Examples ======== In this example we create a netCDF dataset named foo.nc; we want the dataset to be created in the current directory only if a dataset with that name does not already exist: #include ... int status; int ncid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); In this example we create a netCDF dataset named foo_large.nc. It will be in the 64-bit offset format. #include ... int status; int ncid; ... status = nc_create("foo_large.nc", NC_NOCLOBBER|NC_64BIT_OFFSET, &ncid); if (status != NC_NOERR) handle_error(status); In this example we create a netCDF dataset named foo_HDF5.nc. It will be in the HDF5 format. #include ... int status; int ncid; ... status = nc_create("foo_HDF5.nc", NC_NOCLOBBER|NC_NETCDF4, &ncid); if (status != NC_NOERR) handle_error(status); In this example we create a netCDF dataset named foo_HDF5_classic.nc. It will be in the HDF5 format, but will not allow the use of any netCDF-4 advanced features. That is, it will conform to the classic netCDF-3 data model. #include ... int status; int ncid; ... status = nc_create("foo_HDF5_classic.nc", NC_NOCLOBBER|NC_NETCDF4|NC_CLASSIC_MODEL, &ncid); if (status != NC_NOERR) handle_error(status); A variant of nc_create, nc__create (note the double underscore) allows users to specify two tuning parameters for the file that it is creating. These tuning parameters are not written to the data file, they are only used for so long as the file remains open after an nc__create. *Note nc__create::. 2.6 Create a NetCDF Dataset With Performance Options: nc__create ================================================================ This function is a variant of nc_create, nc__create (note the double underscore) allows users to specify two tuning parameters for the file that it is creating. These tuning parameters are not written to the data file, they are only used for so long as the file remains open after an nc__create. This function creates a new netCDF dataset, returning a netCDF ID that can subsequently be used to refer to the netCDF dataset in other netCDF function calls. The new netCDF dataset opened for write access and placed in define mode, ready for you to add dimensions, variables, and attributes. A creation mode flag specifies whether to overwrite any existing dataset with the same name and whether access to the dataset is shared, and whether this file should be in netCDF classic format (the default), or the new 64-bit offset format. Usage ===== int nc__create(const char *path, int cmode, size_t initialsz, size_t *chunksizehintp, int *ncidp); `path' The file name of the new netCDF dataset. `cmode' The creation mode flag. The following flags are available: NC_NOCLOBBER, NC_SHARE, and NC_64BIT_OFFSET. Setting NC_NOCLOBBER means you do not want to clobber (overwrite) an existing dataset; an error (NC_EEXIST) is returned if the specified dataset already exists. The NC_SHARE flag is appropriate when one process may be writing the dataset and one or more other processes reading the dataset concurrently; it means that dataset accesses are not buffered and caching is limited. Since the buffering scheme is optimized for sequential access, programs that do not access data sequentially may see some performance improvement by setting the NC_SHARE flag. Setting NC_64BIT_OFFSET causes netCDF to create a 64-bit offset format file, instead of a netCDF classic format file. The 64-bit offset format imposes far fewer restrictions on very large (i.e. over 2 GB) data files. *Note Large File Support: (netcdf)Large File Support. A zero value (defined for convenience as NC_CLOBBER) specifies the default behavior: overwrite any existing dataset with the same file name and buffer and cache accesses for efficiency. The dataset will be in netCDF classic format. *Note NetCDF Classic Format Limitations: (netcdf)NetCDF Classic Format Limitations. `initialsz' On some systems, and with custom I/O layers, it may be advantageous to set the size of the output file at creation time. This parameter sets the initial size of the file at creation time. `chunksizehintp' The argument referenced by chunksizehintp controls a space versus time tradeoff, memory allocated in the netcdf library versus number of system calls. Because of internal requirements, the value may not be set to exactly the value requested. The actual value chosen is returned by reference. Using the value NC_SIZEHINT_DEFAULT causes the library to choose a default. How the system chooses the default depends on the system. On many systems, the "preferred I/O block size" is available from the stat() system call, struct stat member st_blksize. If this is available it is used. Lacking that, twice the system pagesize is used. Lacking a call to discover the system pagesize, we just set default chunksize to 8192. The chunksize is a property of a given open netcdf descriptor ncid, it is not a persistent property of the netcdf dataset. `ncidp' Pointer to location where returned netCDF ID is to be stored. Errors ====== nc_create returns the value NC_NOERR if no errors occurred. Possible causes of errors include: * Passing a dataset name that includes a directory that does not exist. * Specifying a dataset name of a file that exists and also specifying NC_NOCLOBBER. * Specifying a meaningless value for the creation mode. * Attempting to create a netCDF dataset in a directory where you don't have permission to create files. Examples ======== In this example we create a netCDF dataset named foo.nc; we want the dataset to be created in the current directory only if a dataset with that name does not already exist: #include ... int status; int ncid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); In this example we create a netCDF dataset named foo_large.nc; we want the dataset to be created in the current directory only if a dataset with that name does not already exist. We also specify that chunksize and initial size for the file. #include ... int status; int ncid; int intialsz = 2048; int *chunksize; ... *chunksize = 1024; status = nc__create("foo.nc", NC_NOCLOBBER, initialsz, chunksize, &ncid); if (status != NC_NOERR) handle_error(status); 2.7 Create a NetCDF Dataset With Performance Options: nc_create_par =================================================================== This function is a variant of nc_create, nc_create_par allows users to open a file on a MPI/IO or MPI/Posix parallel file system. The parallel parameters are not written to the data file, they are only used for so long as the file remains open after an nc_create_par. This function creates a new netCDF dataset, returning a netCDF ID that can subsequently be used to refer to the netCDF dataset in other netCDF function calls. The new netCDF dataset opened for write access and placed in define mode, ready for you to add dimensions, variables, and attributes. This function is only available for netCDF-4 files. The creation mode flag must include NC_NETCDF4. When a netCDF-4 file is created for parallel access, collective operations are the default. To use independent access on a variable, *Note nc_var_par_access::. Usage ===== int nc_create_par(const char *path, int cmode, MPI_Comm comm, MPI_Info info, int ncidp); `path' The file name of the new netCDF dataset. `cmode' Either the NC_MPIIO or NC_MPIPOSIX flags must be present. The NC_NETCDF4 flag is also required. Setting NC_NOCLOBBER means you do not want to clobber (overwrite) an existing dataset; an error (NC_EEXIST) is returned if the specified dataset already exists. The NC_SHARE flag is ignored. `comm' The MPI_Comm object returned by the MPI layer. `info' The MPI_Info object returned by the MPI layer, if MPI/IO is being used, or 0 if MPI/Posix is being used. `ncidp' Pointer to location where returned netCDF ID is to be stored. Return Codes ============ `NC_NOERR' No error. `NC_ENOTNC4' Parallel access with netcdf-4 files only. `NC_EPARINIT' Could not initialize parallel access at the HDF5 layer. `NC_EFILEMETA' HDF5 layer cannot handle create or root group open. `NC_EEXIST' Specifying a dataset name of a file that exists and also specifying NC_NOCLOBBER. `NC_EINVAL' Bad value for the creation mode. `NC_EHDFERR' Unexpected error from the HDF5 layer. Examples ======== #include ... int status; int ncid; ... *chunksize = 1024; status = nc__create("foo.nc", NC_NOCLOBBER, initialsz, chunksize, &ncid); if (status != NC_NOERR) handle_error(status); 2.8 Open a NetCDF Dataset for Access: nc_open ============================================= The function nc_open opens an existing netCDF dataset for access. It determines the underlying file format automatically. Use the same call to open a netCDF classic, 64-bit offset, or netCDF-4 file. Usage ===== int nc_open (const char *path, int omode, int *ncidp); `path' File name for netCDF dataset to be opened. When DAP support is enabled, then the path may be an OPeNDAP URL rather than a file path. `omode' A zero value (or NC_NOWRITE) specifies the default behavior: open the dataset with read-only access, buffering and caching accesses for efficiency Otherwise, the open mode is NC_WRITE, NC_SHARE, or NC_WRITE|NC_SHARE. Setting the NC_WRITE flag opens the dataset with read-write access. ("Writing" means any kind of change to the dataset, including appending or changing data, adding or renaming dimensions, variables, and attributes, or deleting attributes.) The NC_SHARE flag is only used for netCDF classic and 64-bit offset files. It is appropriate when one process may be writing the dataset and one or more other processes reading the dataset concurrently; it means that dataset accesses are not buffered and caching is limited. Since the buffering scheme is optimized for sequential access, programs that do not access data sequentially may see some performance improvement by setting the NC_SHARE flag. It is not necessary to pass any information about the format of the file being opened. The file type will be detected automatically by the netCDF library. If a the path is a DAP URL, then the open mode is read-only. Setting NC_WRITE will be ignored. `ncidp' Pointer to location where returned netCDF ID is to be stored. Errors ====== When opening a netCDF-4 file HDF5 error reporting is turned off, if it is on. This doesn't stop the HDF5 error stack from recording the errors, it simply stops their display to the user through stderr. nc_open returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF dataset does not exist. * A meaningless mode was specified. Return Codes ============ `NC_NOERR' No error. `NC_NOMEM' Out of memory. `NC_EHDFERR' HDF5 error. (NetCDF-4 files only.) `NC_EDIMMETA' Error in netCDF-4 dimension metadata. (NetCDF-4 files only.) `NC_ENOCOMPOIND' (NetCDF-4 files only.) Example ======= Here is an example using nc_open to open an existing netCDF dataset named foo.nc for read-only, non-shared access: #include ... int status; int ncid; ... status = nc_open("foo.nc", 0, &ncid); if (status != NC_NOERR) handle_error(status); 2.9 Open a NetCDF Dataset for Access with Performance Tuning: nc__open ====================================================================== A function opens a netCDF dataset for access with an additional performance tuning parameter. When DAP support is enabled, it is possible to open a DAP data source through this interface, but it is deprecated because all of the performance tuning parameters are ignored. The standard nc_open interface should be used instead. Usage ===== int nc__open(const char *path, int mode, size_t *chunksizehintp, int *ncidp); `path' File name for netCDF dataset to be opened. `omode' A zero value (or NC_NOWRITE) specifies the default behavior: open the dataset with read-only access, buffering and caching accesses for efficiency Otherwise, the open mode is NC_WRITE, NC_SHARE, or NC_WRITE|NC_SHARE. Setting the NC_WRITE flag opens the dataset with read-write access. ("Writing" means any kind of change to the dataset, including appending or changing data, adding or renaming dimensions, variables, and attributes, or deleting attributes.) The NC_SHARE flag is appropriate when one process may be writing the dataset and one or more other processes reading the dataset concurrently; it means that dataset accesses are not buffered and caching is limited. Since the buffering scheme is optimized for sequential access, programs that do not access data sequentially may see some performance improvement by setting the NC_SHARE flag. `chunksizehintp' The argument referenced by chunksizehintp controls a space versus time tradeoff, memory allocated in the netcdf library versus number of system calls. Because of internal requirements, the value may not be set to exactly the value requested. The actual value chosen is returned by reference. Using the value NC_SIZEHINT_DEFAULT causes the library to choose a default. How the system chooses the default depends on the system. On many systems, the "preferred I/O block size" is available from the stat() system call, struct stat member st_blksize. If this is available it is used. Lacking that, twice the system pagesize is used. Lacking a call to discover the system pagesize, we just set default chunksize to 8192. The chunksize is a property of a given open netcdf descriptor ncid, it is not a persistent property of the netcdf dataset. `ncidp' Pointer to location where returned netCDF ID is to be stored. Errors ====== nc__open returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF dataset does not exist. * A meaningless mode was specified. Example ======= Here is an example using nc__open to open an existing netCDF dataset named foo.nc for read-only, non-shared access: #include ... int status; int ncid; int *chunksize; ... *chunksize = 1024; status = nc_open("foo.nc", 0, chunksize, &ncid); if (status != NC_NOERR) handle_error(status); 2.10 Open a NetCDF Dataset for Parallel Access ============================================== This function opens a netCDF-4 dataset for parallel access. This opens the file using either MPI-IO or MPI-POSIX. The file must be a netCDF-4 file. (That is, it must have been created using NC_NETCDF4 in the creation mode). DAP access is not allowed through this interface. This function is only available if netCDF-4 was configured with the -use-parallel option before being built. Also HDF5 parallel must be installed (before netCDF-4 is installed.) Before either HDF5 or netCDF-4 can be installed with support for parallel programming, and MPI layer must also be installed on the machine, and usually a parallel file system. NetCDF-4 exposes the parallel access functionality of HDF5. For more information about what is required to install and use the parallel access functions, see the HDF5 web site. When a netCDF-4 file is opened for parallel access, collective operations are the default. To use independent access on a variable, *Note nc_var_par_access::. Usage ===== int nc_open_par(const char *path, int mode, MPI_Comm comm, MPI_Info info, int *ncidp); `path' File name for netCDF dataset to be opened. `omode' Either the NC_MPIIO or NC_MPIPOSIX flags must be present. The flag NC_WRITE opens the dataset with read-write access. ("Writing" means any kind of change to the dataset, including appending or changing data, adding or renaming dimensions, variables, and attributes, or deleting attributes.) All other flags are ignored or not allowed. The NC_NETCDF4 flag is not required, as the file type is detected when the file is opened. `comm' MPI_Comm object returned by the MPI layer. `info' MPI_Info object returned by the MPI layer, or NULL if MPI-POSIX access is desired. `ncidp' Pointer to location where returned netCDF ID is to be stored. Return Codes ============ `NC_NOERR' No error. `NC_ENOTNC4' Not a netCDF-4 file. `' The specified netCDF dataset does not exist. `' A meaningless mode was specified. Example ======= Here is an example using nc_open_par to open an existing netCDF dataset named foo.nc for read-only, non-shared, MPI/IO access: #include ... int status; int ncid; int *chunksize; ... 2.11 Put Open NetCDF Dataset into Define Mode: nc_redef ======================================================= The function nc_redef puts an open netCDF dataset into define mode, so dimensions, variables, and attributes can be added or renamed and attributes can be deleted. Usage ===== For netCDF-4 files (i.e. files created with NC_NETCDF4 in the cmode, *note nc_create::), it is not necessary to call nc_redef unless the file was also created with NC_STRICT_NC3. For straight-up netCDF-4 files, nc_redef is called automatically, as needed. For all netCDF-4 files, the root ncid must be used. This is the ncid returned by nc_open and nc_create, and points to the root of the hierarchy tree for netCDF-4 files. int nc_redef(int ncid); `ncid' netCDF ID, from a previous call to nc_open or nc_create. Errors ====== nc_redef returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF dataset is already in define mode. This error code will only be returned for classic and 64-bit offset format files. * The specified netCDF dataset was opened for read-only. * The specified netCDF ID does not refer to an open netCDF dataset. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad ncid. `NC_EBADGRPID' The ncid must refer to the root group of the file, that is, the group returned by nc_open or nc_create. (*note nc_open:: *note nc_create::). `NC_EINDEFINE' Already in define mode. `NC_EPERM' File is read-only. Example ======= Here is an example using nc_redef to open an existing netCDF dataset named foo.nc and put it into define mode: #include ... int status; int ncid; ... status = nc_open("foo.nc", NC_WRITE, &ncid); /* open dataset */ if (status != NC_NOERR) handle_error(status); ... status = nc_redef(ncid); /* put in define mode */ if (status != NC_NOERR) handle_error(status); 2.12 Leave Define Mode: nc_enddef ================================= The function nc_enddef takes an open netCDF dataset out of define mode. The changes made to the netCDF dataset while it was in define mode are checked and committed to disk if no problems occurred. Non-record variables may be initialized to a "fill value" as well. *Note nc_set_fill::. The netCDF dataset is then placed in data mode, so variable data can be read or written. It's not necessary to call nc_enddef for netCDF-4 files. With netCDF-4 files, nc_enddef is called when needed by the netcdf-4 library. User calls to nc_enddef for netCDF-4 files still flush the metadata to disk. This call may involve copying data under some circumstances. For a more extensive discussion see *note File Structure and Performance: (netcdf)File Structure and Performance. For netCDF-4/HDF5 format files there are some variable settings (the compression, endianness, fletcher32 error correction, and fill value) which must be set (if they are going to be set at all) between the nc_def_var and the next nc_enddef. Once the nc_enddef is called, these settings can no longer be changed for a variable. Usage ===== int nc_enddef(int ncid); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. If you use a group id, the enddef will apply to the entire file. That all, the enddef will not just end define mode in one group, but in the entire file. Errors ====== nc_enddef returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF dataset is not in define mode. * The specified netCDF ID does not refer to an open netCDF dataset. * The size of one or more variables exceed the size constraints for whichever variant of the file format is in use). *Note Large File Support: (netcdf)Large File Support. Example ======= Here is an example using nc_enddef to finish the definitions of a new netCDF dataset named foo.nc and put it into data mode: #include ... int status; int ncid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); ... /* create dimensions, variables, attributes */ status = nc_enddef(ncid); /*leave define mode*/ if (status != NC_NOERR) handle_error(status); 2.13 Leave Define Mode with Performance Tuning: nc__enddef ========================================================== The function nc__enddef takes an open netCDF dataset out of define mode. The changes made to the netCDF dataset while it was in define mode are checked and committed to disk if no problems occurred. Non-record variables may be initialized to a "fill value" as well. *Note nc_set_fill::. The netCDF dataset is then placed in data mode, so variable data can be read or written. This call may involve copying data under some circumstances. For a more extensive discussion see *note File Structure and Performance: (netcdf)File Structure and Performance. Caution: this function exposes internals of the netcdf version 1 file format. Users should use nc_enddef in most circumstances. This function may not be available on future netcdf implementations. The current netcdf file format has three sections, the "header" section, the data section for fixed size variables, and the data section for variables which have an unlimited dimension (record variables). The header begins at the beginning of the file. The index (offset) of the beginning of the other two sections is contained in the header. Typically, there is no space between the sections. This causes copying overhead to accrue if one wishes to change the size of the sections, as may happen when changing names of things, text attribute values, adding attributes or adding variables. Also, for buffered i/o, there may be advantages to aligning sections in certain ways. The minfree parameters allow one to control costs of future calls to nc_redef, nc_enddef by requesting that minfree bytes be available at the end of the section. The align parameters allow one to set the alignment of the beginning of the corresponding sections. The beginning of the section is rounded up to an index which is a multiple of the align parameter. The flag value ALIGN_CHUNK tells the library to use the chunksize (see above) as the align parameter. The file format requires mod 4 alignment, so the align parameters are silently rounded up to multiples of 4. The usual call, nc_enddef(ncid); is equivalent to nc__enddef(ncid, 0, 4, 0, 4); The file format does not contain a "record size" value, this is calculated from the sizes of the record variables. This unfortunate fact prevents us from providing minfree and alignment control of the "records" in a netcdf file. If you add a variable which has an unlimited dimension, the third section will always be copied with the new variable added. Usage ===== int nc__enddef(int ncid, size_t h_minfree, size_t v_align, size_t v_minfree, size_t r_align); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `h_minfree' Sets the pad at the end of the "header" section. `v_align' Controls the alignment of the beginning of the data section for fixed size variables. `v_minfree' Sets the pad at the end of the data section for fixed size variables. `r_align' Controls the alignment of the beginning of the data section for variables which have an unlimited dimension (record variables). Errors ====== nc__enddef returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF dataset is not in define mode. * The specified netCDF ID does not refer to an open netCDF dataset. * The size of one or more variables exceed the size constraints for whichever variant of the file format is in use). *Note Large File Support: (netcdf)Large File Support. Example ======= Here is an example using nc_enddef to finish the definitions of a new netCDF dataset named foo.nc and put it into data mode: #include ... int status; int ncid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); ... /* create dimensions, variables, attributes */ status = nc_enddef(ncid); /*leave define mode*/ if (status != NC_NOERR) handle_error(status); 2.14 Close an Open NetCDF Dataset: nc_close =========================================== The function nc_close closes an open netCDF dataset. If the dataset in define mode, nc_enddef will be called before closing. (In this case, if nc_enddef returns an error, nc_abort will automatically be called to restore the dataset to the consistent state before define mode was last entered.) After an open netCDF dataset is closed, its netCDF ID may be reassigned to the next netCDF dataset that is opened or created. Usage ===== For netCDF-4 files, the ncid of the root group must be passed into nc_close. int nc_close(int ncid); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. Errors ====== nc_close returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * Define mode was entered and the automatic call made to nc_enddef failed. * The specified netCDF ID does not refer to an open netCDF dataset. `NC_NOERR' No error. `NC_EBADID' Invalid id passed. `NC_EBADGRPID' ncid did not contain the root group id of this file. (NetCDF-4 only). Example ======= Here is an example using nc_close to finish the definitions of a new netCDF dataset named foo.nc and release its netCDF ID: #include ... int status; int ncid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); ... /* create dimensions, variables, attributes */ status = nc_close(ncid); /* close netCDF dataset */ if (status != NC_NOERR) handle_error(status); 2.15 Inquire about an Open NetCDF Dataset: nc_inq Family ======================================================== Members of the nc_inq family of functions return information about an open netCDF dataset, given its netCDF ID. Dataset inquire functions may be called from either define mode or data mode. The first function, nc_inq, returns values for the number of dimensions, the number of variables, the number of global attributes, and the dimension ID of the dimension defined with unlimited length, if any. The other functions in the family each return just one of these items of information. For C, these functions include nc_inq, nc_inq_ndims, nc_inq_nvars, nc_inq_natts, and nc_inq_unlimdim. An additional function, nc_inq_format, returns the (rarely needed) format version. No I/O is performed when these functions are called, since the required information is available in memory for each open netCDF dataset. Usage ===== int nc_inq (int ncid, int *ndimsp, int *nvarsp, int *ngattsp, int *unlimdimidp); int nc_inq_ndims (int ncid, int *ndimsp); int nc_inq_nvars (int ncid, int *nvarsp); int nc_inq_natts (int ncid, int *ngattsp); int nc_inq_unlimdim (int ncid, int *unlimdimidp); int nc_inq_format (int ncid, int *formatp); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `ndimsp' Pointer to location for returned number of dimensions defined for this netCDF dataset. `nvarsp' Pointer to location for returned number of variables defined for this netCDF dataset. `ngattsp' Pointer to location for returned number of global attributes defined for this netCDF dataset. `unlimdimidp' Pointer to location for returned ID of the unlimited dimension, if there is one for this netCDF dataset. If no unlimited length dimension has been defined, -1 is returned. `formatp' Pointer to location for returned format version, one of NC_FORMAT_CLASSIC, NC_FORMAT_64BIT, NC_FORMAT_NETCDF4, NC_FORMAT_NETCDF4_CLASSIC. Errors ====== All members of the nc_inq family return the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_inq to find out about a netCDF dataset named foo.nc: #include ... int status, ncid, ndims, nvars, ngatts, unlimdimid; ... status = nc_open("foo.nc", NC_NOWRITE, &ncid); if (status != NC_NOERR) handle_error(status); ... status = nc_inq(ncid, &ndims, &nvars, &ngatts, &unlimdimid); if (status != NC_NOERR) handle_error(status); 2.16 Synchronize an Open NetCDF Dataset to Disk: nc_sync ======================================================== The function nc_sync offers a way to synchronize the disk copy of a netCDF dataset with in-memory buffers. There are two reasons you might want to synchronize after writes: * To minimize data loss in case of abnormal termination, or * To make data available to other processes for reading immediately after it is written. But note that a process that already had the dataset open for reading would not see the number of records increase when the writing process calls nc_sync; to accomplish this, the reading process must call nc_sync. This function is backward-compatible with previous versions of the netCDF library. The intent was to allow sharing of a netCDF dataset among multiple readers and one writer, by having the writer call nc_sync after writing and the readers call nc_sync before each read. For a writer, this flushes buffers to disk. For a reader, it makes sure that the next read will be from disk rather than from previously cached buffers, so that the reader will see changes made by the writing process (e.g., the number of records written) without having to close and reopen the dataset. If you are only accessing a small amount of data, it can be expensive in computer resources to always synchronize to disk after every write, since you are giving up the benefits of buffering. An easier way to accomplish sharing (and what is now recommended) is to have the writer and readers open the dataset with the NC_SHARE flag, and then it will not be necessary to call nc_sync at all. However, the nc_sync function still provides finer granularity than the NC_SHARE flag, if only a few netCDF accesses need to be synchronized among processes. It is important to note that changes to the ancillary data, such as attribute values, are not propagated automatically by use of the NC_SHARE flag. Use of the nc_sync function is still required for this purpose. Sharing datasets when the writer enters define mode to change the data schema requires extra care. In previous releases, after the writer left define mode, the readers were left looking at an old copy of the dataset, since the changes were made to a new copy. The only way readers could see the changes was by closing and reopening the dataset. Now the changes are made in place, but readers have no knowledge that their internal tables are now inconsistent with the new dataset schema. If netCDF datasets are shared across redefinition, some mechanism external to the netCDF library must be provided that prevents access by readers during redefinition and causes the readers to call nc_sync before any subsequent access. When calling nc_sync, the netCDF dataset must be in data mode. A netCDF dataset in define mode is synchronized to disk only when nc_enddef is called. A process that is reading a netCDF dataset that another process is writing may call nc_sync to get updated with the changes made to the data by the writing process (e.g., the number of records written), without having to close and reopen the dataset. Data is automatically synchronized to disk when a netCDF dataset is closed, or whenever you leave define mode. Usage ===== int nc_sync(int ncid); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. Errors ====== nc_sync returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The netCDF dataset is in define mode. * The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_sync to synchronize the disk writes of a netCDF dataset named foo.nc: #include ... int status; int ncid; ... status = nc_open("foo.nc", NC_WRITE, &ncid); /* open for writing */ if (status != NC_NOERR) handle_error(status); ... /* write data or change attributes */ status = nc_sync(ncid); /* synchronize to disk */ if (status != NC_NOERR) handle_error(status); 2.17 Back Out of Recent Definitions: nc_abort ============================================= You no longer need to call this function, since it is called automatically by nc_close in case the dataset is in define mode and something goes wrong with committing the changes. The function nc_abort just closes the netCDF dataset, if not in define mode. If the dataset is being created and is still in define mode, the dataset is deleted. If define mode was entered by a call to nc_redef, the netCDF dataset is restored to its state before definition mode was entered and the dataset is closed. Usage ===== int nc_abort(int ncid); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. Errors ====== nc_abort returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * When called from define mode while creating a netCDF dataset, deletion of the dataset failed. * The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_abort to back out of redefinitions of a dataset named foo.nc: #include ... int ncid, status, latid; ... status = nc_open("foo.nc", NC_WRITE, &ncid);/* open for writing */ if (status != NC_NOERR) handle_error(status); ... status = nc_redef(ncid); /* enter define mode */ if (status != NC_NOERR) handle_error(status); ... status = nc_def_dim(ncid, "lat", 18L, &latid); if (status != NC_NOERR) { handle_error(status); status = nc_abort(ncid); /* define failed, abort */ if (status != NC_NOERR) handle_error(status); } 2.18 Set Fill Mode for Writes: nc_set_fill ========================================== This function is intended for advanced usage, to optimize writes under some circumstances described below. The function nc_set_fill sets the fill mode for a netCDF dataset open for writing and returns the current fill mode in a return parameter. The fill mode can be specified as either NC_FILL or NC_NOFILL. The default behavior corresponding to NC_FILL is that data is pre-filled with fill values, that is fill values are written when you create non-record variables or when you write a value beyond data that has not yet been written. This makes it possible to detect attempts to read data before it was written. For more information on the use of fill values see *note Fill Values::. For information about how to define your own fill values see *note Attribute Conventions: (netcdf)Attribute Conventions. The behavior corresponding to NC_NOFILL overrides the default behavior of prefilling data with fill values. This can be used to enhance performance, because it avoids the duplicate writes that occur when the netCDF library writes fill values that are later overwritten with data. A value indicating which mode the netCDF dataset was already in is returned. You can use this value to temporarily change the fill mode of an open netCDF dataset and then restore it to the previous mode. After you turn on NC_NOFILL mode for an open netCDF dataset, you must be certain to write valid data in all the positions that will later be read. Note that nofill mode is only a transient property of a netCDF dataset open for writing: if you close and reopen the dataset, it will revert to the default behavior. You can also revert to the default behavior by calling nc_set_fill again to explicitly set the fill mode to NC_FILL. There are three situations where it is advantageous to set nofill mode: 1. Creating and initializing a netCDF dataset. In this case, you should set nofill mode before calling nc_enddef and then write completely all non-record variables and the initial records of all the record variables you want to initialize. 2. Extending an existing record-oriented netCDF dataset. Set nofill mode after opening the dataset for writing, then append the additional records to the dataset completely, leaving no intervening unwritten records. 3. Adding new variables that you are going to initialize to an existing netCDF dataset. Set nofill mode before calling nc_enddef then write all the new variables completely. If the netCDF dataset has an unlimited dimension and the last record was written while in nofill mode, then the dataset may be shorter than if nofill mode was not set, but this will be completely transparent if you access the data only through the netCDF interfaces. The use of this feature may not be available (or even needed) in future releases. Programmers are cautioned against heavy reliance upon this feature. Usage ===== int nc_set_fill (int ncid, int fillmode, int *old_modep); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `fillmode' Desired fill mode for the dataset, either NC_NOFILL or NC_FILL. `old_modep' Pointer to location for returned current fill mode of the dataset before this call, either NC_NOFILL or NC_FILL. Return Codes ============ `NC_NOERR' No error. `NC_EBADID' The specified netCDF ID does not refer to an open netCDF dataset. `NC_EPERM' The specified netCDF ID refers to a dataset open for read-only access. `NC_EINVAL' The fill mode argument is neither NC_NOFILL nor NC_FILL. Example ======= Here is an example using nc_set_fill to set nofill mode for subsequent writes of a netCDF dataset named foo.nc: #include ... int ncid, status, old_fill_mode; ... status = nc_open("foo.nc", NC_WRITE, &ncid); /* open for writing */ if (status != NC_NOERR) handle_error(status); ... /* write data with default prefilling behavior */ status = nc_set_fill(ncid, NC_NOFILL, &old_fill_mode); /* set nofill */ if (status != NC_NOERR) handle_error(status); ... /* write data with no prefilling */ 2.19 Set Default Creation Format: nc_set_default_format ======================================================= This function is intended for advanced users. Starting in version 3.6, netCDF introduced a new data format, the first change in the underlying binary data format since the netCDF interface was released. The new format, 64-bit offset format, was introduced to greatly relax the limitations on creating very large files. Users are warned that creating files in the 64-bit offset format makes them unreadable by the netCDF library prior to version 3.6.0. For reasons of compatibility, users should continue to create files in netCDF classic format. Users who do want to use 64-bit offset format files can create them directory from nc_create, using the proper cmode flag. (*note nc_create::). The function nc_set_default_format allows the user to change the format of the netCDF file to be created by future calls to nc_create (or nc__create) without changing the cmode flag. This allows the user to convert a program to use 64-bit offset formation without changing all calls the nc_create. *Note Large File Support: (netcdf)Large File Support. Once the default format is set, all future created files will be in the desired format. Two constants are provided in the netcdf.h file to be used with this function, NC_FORMAT_64BIT and NC_FORMAT_CLASSIC. If a non-NULL pointer is provided, it is assumed to point to an int, where the existing default format will be written. Using nc_create with a cmode including NC_64BIT_OFFSET overrides the default format, and creates a 64-bit offset file. Usage ===== int nc_set_default_format(int format, int *old_formatp); `format' Valid formats include NC_FORMAT_CLASSIC (the default), NC_FORMAT_64BIT, and, if -enable-netcdf-4 was used during configure, NC_FORMAT_NETCDF4 and NC_FORMAT_NETCDF4_CLASSIC `old_formatp' Either NULL (in which case it will be ignored), or a pointer to an int where the existing default format (i.e. before being changed to the new format) will be written. This allows you to get the existing default format while setting a new default format. Return Codes ============ `NC_NOERR' No error. `NC_EINVAL' Invalid format. Valid formats include NC_FORMAT_CLASSIC, NC_FORMAT_64BIT, and, if -enable-netcdf-4 was used during configure, NC_FORMAT_NETCDF4 and NC_FORMAT_NETCDF4_CLASSIC. Trying to set the default format to something else will result in an invalid argument error. Example ======= Here is an example using nc_set_default_format to create the same file in four formats with the same nc_create call (from libsrc4/tst_utf8.c): #include ... int ncid, varid, dimids[NDIMS]; int f; for (f = NC_FORMAT_CLASSIC; f < NC_FORMAT_NETCDF4_CLASSIC; f++) { if (nc_set_default_format(f, NULL)) ERR; if (nc_create(FILE_NAME, NC_CLOBBER, &ncid)) ERR; ... 3 Groups ******** NetCDF-4 added support for hierarchical groups within netCDF datasets. Groups are identified with a ncid, which identifies both the open file, and the group within that file. When a file is opened with nc_open or nc_create, the ncid for the root group of that file is provided. Using that as a starting point, users can add new groups, or list and navigate existing groups. All netCDF calls take a ncid which determines where the call will take its action. For example, the nc_def_var function takes a ncid as its first parameter. It will create a variable in whichever group its ncid refers to. Use the root ncid provided by nc_create or nc_open to create a variable in the root group. Or use nc_def_grp to create a group and use its ncid to define a variable in the new group. Variable are only visible in the group in which they are defined. The same applies to attributes. "Global" attributes are associated with the group whose ncid is used. Dimensions are visible in their groups, and all child groups. Group operations are only permitted on netCDF-4 files - that is, files created with the HDF5 flag in nc_create. (*note nc_create::). Groups are not compatible with the netCDF classic data model, so files created with the NC_CLASSIC_MODEL file cannot contain groups (except the root group). 3.1 Find a Group ID: nc_inq_ncid ================================ Given an ncid and group name (NULL or "" gets root group), return ncid of the named group. Usage ===== int nc_inq_ncid(int ncid, char *name, int *grp_ncid); `ncid' The group id for this operation. `name' A char array that holds the name of the desired group. `grp_ncid' An int pointer that will receive the group id, if the group is found. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= int root_ncid, child_ncid; char file[] = "nc4_test.nc"; /* Open the file. */ if ((res = nc_open(file, NC_NOWRITE, &root_ncid))) return res; /* Get the ncid of an existing group. */ if ((res = nc_inq_ncid(root_ncid, "group1", &child_ncid))) return res; 3.2 Get a List of Groups in a Group: nc_inq_grps ================================================ Given a location id, return the number of groups it contains, and an array of their ncids. Usage ===== int nc_inq_grps(int ncid, int *numgrps, int *ncids); `ncid' The group id for this operation. `numgrps' Pointer to an int which will get number of groups in this group. If NULL, it's ignored. `ncids' Pointer to a already allocated array of ints which will receive the ids of all the groups in this group. If NULL, it's ignored. Call this function with NULL for ncids parameter to find out how many groups there are. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= int root_ncid, numgrps; int *ncids; char file[] = "nc4_test.nc"; /* Open the file. */ if ((res = nc_open(file, NC_NOWRITE, &root_ncid))) return res; /* Get a list of ncids for the root group. (That is, find out of there are any groups already defined. */ if ((res = nc_inq_grps(root_ncid, &numgrps, NULL))) return res; ncids = malloc(sizeof(int) * numgrps); if ((res = nc_inq_grps(root_ncid, NULL, ncids))) return res; 3.3 Find all the Variables in a Group: nc_inq_varids ==================================================== Find all varids for a location. Usage ===== nc_inq_varids(int ncid, int *nvars, int *varids); `ncid' The group id for this operation. `nvars' The integer pointed to by this parameter will get the number of variable IDs found. `varids' An already allocated array to store the list of varids. Ignored if NULL. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from libsrc4/tst_vars.c. int nvars_in, varids_in[2]; ... /* Open the file and make sure nc_inq_varids yeilds correct * result. */ if (nc_open(FILE_NAME, NC_NOWRITE, &ncid)) ERR; if (nc_inq_varids(ncid, &nvars_in, varids_in)) ERR; 3.4 Find all Dimensions Visible in a Group: nc_inq_dimids ========================================================= Find all dimids for a location. This finds all dimensions in a group, or any of its parents. Usage ===== int nc_inq_dimids(int ncid, int *ndims, int *dimids, int include_parents); `ncid' The group id for this operation. `dimids' An already allocated array of ints when the dimids of the visible dimensions will be stashed. Use nc_inq_ndims to find out how many dims are visible from this group. (*note nc_inq Family::). `include_parents' If non-zero then all the dimensions in all parent groups will also be retrieved. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from libsrc4/tst_dims.c. int ncid, dimid; int ndims_in, dimids_in[MAX_DIMS]; ... /* Open the file and make sure nc_inq_dimids yeilds correct * result. */ if (nc_open(FILE_NAME, NC_NOWRITE, &ncid)) ERR; if (nc_inq_dimids(ncid, &ndims_in, dimids_in, 0)) ERR; if (ndims_in != 1 || dimids_in[0] != 0) ERR; 3.5 Find a Group's Name: nc_inq_grpname ======================================= Given a group ID find its name. (Root group is named "/"). *Note nc_inq_grpname_full::. Usage ===== int nc_inq_grpname(int ncid, char *name); `ncid' The group id for this operation. `name' Pointer to allocated space of correct length. The name of the group will be copied there. The name will be less than NC_MAX_NAME, not including a terminating NULL byte. `' Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from libsrc4/tst_grps.c. int grpid_in[MAX_SIBLING_GROUPS]; char name_in[NC_MAX_NAME + 1]; ... if (nc_inq_grpname(grpid_in[0], name_in)) ERR; 3.6 Find a Group's Full Name: nc_inq_grpname_full ================================================= Given ncid, find complete name of group. (Root group is named "/", a full "path" for each group is provided in the name, with groups separated with a forward slash / as in Unix directory names. For example "/group1/subgrp1/subsubgrp1") Usage ===== int nc_inq_grpname_full(int ncid, size_t *lenp, char *full_name); `ncid' The group id for this operation. `lenp' Pointer to a size_t which will get the length. (Note that this does not include the NULL terminator byte.) `full_name' Pointer to allocated space of correct length. The name of the group will be copied there. To find the required length, call nc_inq_grpname_len call this function with NULL for the full_name parameter, or call (*note nc_inq_grpname_len::). `' Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from the test program libsrc4/tst_grps.c. int grpid_in[MAX_SIBLING_GROUPS]; char full_name_in[NC_MAX_NAME * 10]; size_t len; ... if (nc_inq_grpname_full(grpid_in[0], &len, full_name_in)) ERR; 3.7 Find the Length of a Group's Full Name: nc_inq_grpname_len ============================================================== Given ncid, find len of the full name, as returned by nc_inq_grpname_full *Note nc_inq_grpname_full::. (Root group is named "/", with length 1.) Usage ===== int nc_inq_grpname_len(int ncid, size_t *lenp); `ncid' The group id for this operation. `lenp' Pointer to an int where the length will be placed. Length does not include NULL terminator. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from libsrc4/tst_grps.c. int ncid; size_t len; ... if (nc_inq_grpname_len(ncid, &len)) ERR; 3.8 Find a Group's Parent: nc_inq_grp_parent ============================================ Given ncid, find the ncid of the parent group. When used with the root group, this function returns the NC_ENOGRP error (since the root group has no parent.) Usage ===== int nc_inq_grp_parent(int ncid, int *parent_ncid); `ncid' The group id. `parent_ncid' Pointer to an int. The ncid of the group will be copied there. `' Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOGRP' No parent group found (i.e. this is the root group). `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= if (nc_create(FILE_NAME, NC_NETCDF4, &ncid)) ERR; if (nc_def_grp(ncid, HENRY_VII, &henry_vii_id)) ERR; if (nc_inq_grp_parent(henry_vii_id, &parent_ncid)) ERR; if (parent_ncid != ncid) ERR; if (nc_close(ncid)) ERR; 3.9 Create a New Group: nc_def_grp ================================== Create a group. Its location id is returned in the new_ncid pointer. Usage ===== int nc_def_grp(int parent_ncid, char *name, int *new_ncid); `parent_ncid' The group id of the parent group. `name' The name of the new group. `new_ncid' A pointer to an int. The ncid of the new group will be placed there. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENAMEINUSE' That name is in use. Group names must be unique within a group, and must not be the same as any variable or type in the group. `NC_EMAXNAME' Name exceed max length NC_MAX_NAME. `NC_EBADNAME' Name contains illegal characters. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. `NC_EPERM' Attempt to write to a read-only file. `NC_ENOTINDEFINE' Not in define mode. Example ======= int root_ncid, a1_ncid; char grpname[] = "assimilation1"; /* Get the ncid of the root group. */ if ((res = nc_inq_ncid(root_ncid, NULL, &root_ncid))) return res; /* Create a group. */ if ((res = nc_def_grp(root_ncid, grpname, &a1_ncid))) return res; 4 Dimensions ************ 4.1 Dimensions Introduction =========================== Dimensions for a netCDF dataset are defined when it is created, while the netCDF dataset is in define mode. Additional dimensions may be added later by reentering define mode. A netCDF dimension has a name and a length. In a netCDF classic or 64-bit offset file, at most one dimension can have the unlimited length, which means variables using this dimension can grow along this dimension. In a netCDF-4 file multiple unlimited dimensions are supported. There is a suggested limit (100) to the number of dimensions that can be defined in a single netCDF dataset. The limit is the value of the predefined macro NC_MAX_DIMS. The purpose of the limit is to make writing generic applications simpler. They need only provide an array of NC_MAX_DIMS dimensions to handle any netCDF dataset. The implementation of the netCDF library does not enforce this advisory maximum, so it is possible to use more dimensions, if necessary, but netCDF utilities that assume the advisory maximums may not be able to handle the resulting netCDF datasets. Ordinarily, the name and length of a dimension are fixed when the dimension is first defined. The name may be changed later, but the length of a dimension (other than the unlimited dimension) cannot be changed without copying all the data to a new netCDF dataset with a redefined dimension length. Dimension lengths in the C interface are type size_t rather than type int to make it possible to access all the data in a netCDF dataset on a platform that only supports a 16-bit int data type, for example MSDOS. If dimension lengths were type int instead, it would not be possible to access data from variables with a dimension length greater than a 16-bit int can accommodate. A netCDF dimension in an open netCDF dataset is referred to by a small integer called a dimension ID. In the C interface, dimension IDs are 0, 1, 2, ..., in the order in which the dimensions were defined. Operations supported on dimensions are: * Create a dimension, given its name and length. * Get a dimension ID from its name. * Get a dimension's name and length from its ID. * Rename a dimension. 4.2 Create a Dimension: nc_def_dim ================================== The function nc_def_dim adds a new dimension to an open netCDF dataset in define mode. It returns (as an argument) a dimension ID, given the netCDF ID, the dimension name, and the dimension length. At most one unlimited length dimension, called the record dimension, may be defined for each classic or 64-bit offset netCDF dataset. NetCDF-4 datasets may have multiple unlimited dimensions. Usage ===== int nc_def_dim (int ncid, const char *name, size_t len, int *dimidp); `ncid' NetCDF group ID, from a previous call to nc_open, nc_create, nc_def_grp, etc. `name' Dimension name. Must begin with an alphabetic character, followed by zero or more alphanumeric characters including the underscore ('_'). Case is significant. `len' Length of dimension; that is, number of values for this dimension as an index to variables that use it. This should be either a positive integer (of type size_t) or the predefined constant NC_UNLIMITED. `dimidp' Pointer to location for returned dimension ID. Errors ====== nc_def_dim returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The netCDF dataset is not in definition mode. * The specified dimension name is the name of another existing dimension. * The specified length is not greater than zero. * The specified length is unlimited, but there is already an unlimited length dimension defined for this netCDF dataset. * The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_def_dim to create a dimension named lat of length 18 and a unlimited dimension named rec in a new netCDF dataset named foo.nc: #include ... int status, ncid, latid, recid; ... status = nc_create("foo.nc", NC_NOCLOBBER, &ncid); if (status != NC_NOERR) handle_error(status); ... status = nc_def_dim(ncid, "lat", 18L, &latid); if (status != NC_NOERR) handle_error(status); status = nc_def_dim(ncid, "rec", NC_UNLIMITED, &recid); if (status != NC_NOERR) handle_error(status); 4.3 Get a Dimension ID from Its Name: nc_inq_dimid ================================================== The function nc_inq_dimid returns (as an argument) the ID of a netCDF dimension, given the name of the dimension. If ndims is the number of dimensions defined for a netCDF dataset, each dimension has an ID between 0 and ndims-1. Usage ===== When searching for a dimension, the specified group is searched, and then its parent group, and then its grandparent group, etc., up to the root group. int nc_inq_dimid (int ncid, const char *name, int *dimidp); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `name' Dimension name, a character string beginning with a letter and followed by any sequence of letters, digits, or underscore ('_') characters. Case is significant in dimension names. `dimidp' Pointer to location for the returned dimension ID. Errors ====== nc_inq_dimid returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: The name that was specified is not the name of a dimension in the netCDF dataset. The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_inq_dimid to determine the dimension ID of a dimension named lat, assumed to have been defined previously in an existing netCDF dataset named foo.nc: #include ... int status, ncid, latid; ... status = nc_open("foo.nc", NC_NOWRITE, &ncid); /* open for reading */ if (status != NC_NOERR) handle_error(status); ... status = nc_inq_dimid(ncid, "lat", &latid); if (status != NC_NOERR) handle_error(status); 4.4 Inquire about a Dimension: nc_inq_dim Family ================================================ This family of functions returns information about a netCDF dimension. Information about a dimension includes its name and its length. The length for the unlimited dimension, if any, is the number of records written so far. The functions in this family include nc_inq_dim, nc_inq_dimname, and nc_inq_dimlen. The function nc_inq_dim returns all the information about a dimension; the other functions each return just one item of information. Usage ===== int nc_inq_dim (int ncid, int dimid, char* name, size_t* lengthp); int nc_inq_dimname (int ncid, int dimid, char *name); int nc_inq_dimlen (int ncid, int dimid, size_t *lengthp); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `dimid' Dimension ID, from a previous call to nc_inq_dimid or nc_def_dim. `name' Returned dimension name. The caller must allocate space for the returned name. The maximum possible length, in characters, of a dimension name is given by the predefined constant NC_MAX_NAME. (This doesn't include the null terminator, so declare your array to be size NC_MAX_NAME+1). The returned character array will be null-terminated. `lengthp' Pointer to location for returned length of dimension. For the unlimited dimension, this is the number of records written so far. Errors ====== These functions return the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The dimension ID is invalid for the specified netCDF dataset. * The specified netCDF ID does not refer to an open netCDF dataset. Example ======= Here is an example using nc_inq_dim to determine the length of a dimension named lat, and the name and current maximum length of the unlimited dimension for an existing netCDF dataset named foo.nc: #include ... int status, ncid, latid, recid; size_t latlength, recs; char recname[NC_MAX_NAME+1]; ... status = nc_open("foo.nc", NC_NOWRITE, &ncid); /* open for reading */ if (status != NC_NOERR) handle_error(status); status = nc_inq_unlimdim(ncid, &recid); /* get ID of unlimited dimension */ if (status != NC_NOERR) handle_error(status); ... status = nc_inq_dimid(ncid, "lat", &latid); /* get ID for lat dimension */ if (status != NC_NOERR) handle_error(status); status = nc_inq_dimlen(ncid, latid, &latlength); /* get lat length */ if (status != NC_NOERR) handle_error(status); /* get unlimited dimension name and current length */ status = nc_inq_dim(ncid, recid, recname, &recs); if (status != NC_NOERR) handle_error(status); 4.5 Rename a Dimension: nc_rename_dim ===================================== The function nc_rename_dim renames an existing dimension in a netCDF dataset open for writing. You cannot rename a dimension to have the same name as another dimension. For netCDF classic and 64-bit offset files, if the new name is longer than the old name, the netCDF dataset must be in define mode. For netCDF-4 files the dataset is switched to define more for the rename, regardless of the name length. Usage ===== int nc_rename_dim(int ncid, int dimid, const char* name); `ncid' NetCDF ID, from a previous call to nc_open or nc_create. `dimid' Dimension ID, from a previous call to nc_inq_dimid or nc_def_dim. `name' New dimension name. Errors ====== nc_rename_dim returns the value NC_NOERR if no errors occurred. Otherwise, the returned status indicates an error. Possible causes of errors include: * The new name is the name of another dimension. * The dimension ID is invalid for the specified netCDF dataset. * The specified netCDF ID does not refer to an open netCDF dataset. * The new name is longer than the old name and the netCDF dataset is not in define mode. Example ======= Here is an example using nc_rename_dim to rename the dimension lat to latitude in an existing netCDF dataset named foo.nc: #include ... int status, ncid, latid; ... status = nc_open("foo.nc", NC_WRITE, &ncid); /* open for writing */ if (status != NC_NOERR) handle_error(status); ... status = nc_redef(ncid); /* put in define mode to rename dimension */ if (status != NC_NOERR) handle_error(status); status = nc_inq_dimid(ncid, "lat", &latid); if (status != NC_NOERR) handle_error(status); status = nc_rename_dim(ncid, latid, "latitude"); if (status != NC_NOERR) handle_error(status); status = nc_enddef(ncid); /* leave define mode */ if (status != NC_NOERR) handle_error(status); 4.6 Find All Unlimited Dimension IDs: nc_inq_unlimdims ====================================================== In netCDF-4 files, it's possible to have multiple unlimited dimensions. This function returns a list of the unlimited dimension ids visible in a group. Dimensions are visible in a group if they have been defined in that group, or any ancestor group. Usage ===== int nc_inq_unlimdims(int ncid, int *nunlimdimsp, int *unlimdimidsp); `ncid' NetCDF group ID, from a previous call to nc_open, nc_create, nc_def_grp, etc. `nunlimdimsp' A pointer to an int which will get the number of visible unlimited dimensions. Ignored if NULL. `unlimdimidsp' A pointer to an already allocated array of int which will get the ids of all visible unlimited dimensions. Ignored if NULL. To allocate the correct length for this array, call nc_inq_unlimdims with a NULL for this parameter and use the nunlimdimsp parameter to get the number of visible unlimited dimensions. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag HDF5. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= int root_ncid, num_unlimdims, unlimdims[NC_MAX_DIMS]; char file[] = "nc4_test.nc"; int res; /* Open the file. */ if ((res = nc_open(file, NC_NOWRITE, &root_ncid))) return res; /* Find out if there are any unlimited dimensions in the root group. */ if ((res = nc_inq_unlimdims(root_ncid, &num_unlimdims, unlimdims))) return res; printf("nc_inq_unlimdims reports %d unlimited dimensions\n", num_unlimdims); 5 User Defined Data Types ************************* 5.1 User Defined Types Introduction =================================== NetCDF-4 has added support for four different user defined data types. `compound type' Like a C struct, a compound type is a collection of types, including other user defined types, in one package. `variable length array type' The variable length array may be used to store ragged arrays. `opaque type' This type has only a size per element, and no other type information. `enum type' Like an enumeration in C, this type lets you assign text values to integer values, and store the integer values. Users may construct user defined type with the various nc_def_* functions described in this section. They may learn about user defined types by using the nc_inq_ functions defined in this section. 5.2 Learn the IDs of All Types in Group: nc_inq_typeids ======================================================= Learn the number of types defined in a group, and their IDs. Usage ===== int nc_inq_typeids(int ncid, int *ntypes, int *typeids); `ncid' The group id. `ntypes' A pointer to int which will get the number of types defined in the group. If NULL, ignored. `typeids' A pointer to an int array which will get the typeids. If NULL, ignored. Errors ====== `NC_NOERR' No error. `NC_BADID' Bad ncid. Example ======= The following example is from the test program libsrc4/tst_enums.c. if (nc_open(FILE_NAME, NC_NOWRITE, &ncid)) ERR; /* Get type info. */ if (nc_inq_typeids(ncid, &ntypes, typeids)) ERR; if (ntypes != 1 || !typeids[0]) ERR; 5.3 Learn About an User Defined Type: nc_inq_type ================================================= Given an ncid and a typeid, get the information about a type. This function will work on any type, including atomic and any user defined type, whether compound, opaque, enumeration, or variable length array. For even more information about a user defined type *note nc_inq_user_type::. Usage ===== nc_inq_type(int ncid, nc_type xtype, char *name, size_t *sizep); `ncid' The ncid for the group containing the type (ignored for atomic types). `xtype' The typeid for this type, as returned by nc_def_compound, nc_def_opaque, nc_def_enum, nc_def_vlen, or nc_inq_var, or as found in netcdf.h in the list of atomic types (NC_CHAR, NC_INT, etc.). `name' If non-NULL, the name of the user defined type will be copied here. It will be NC_MAX_NAME bytes or less. For atomic types, the type name from CDL will be given. `sizep' If non-NULL, the size of the type (in bytes) will be copied here. VLEN type size is the size of one element of the VLEN. String size is returned as zero, since it varies from string to string. Return Codes ============ `NC_NOERR' No error. `NC_EBADTYPEID' Bad typeid. `NC_ENOTNC4' Seeking a user-defined type in a netCDF-3 file. `NC_ESTRICTNC3' Seeking a user-defined type in a netCDF-4 file for which classic model has been turned on. `NC_EBADGRPID' Bad group ID in ncid. `NC_EBADID' Type ID not found. `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= This example is from the test program tst_enums.c, and it uses all the possible inquiry functions on an enum type. /* Check it out. */ if (nc_inq_user_type(ncid, typeids[0], name_in, &base_size_in, &base_nc_type_in, &nfields_in, &class_in)) ERR; if (strcmp(name_in, TYPE_NAME) || base_size_in != sizeof(int) || base_nc_type_in != NC_INT || nfields_in != NUM_MEMBERS || class_in != NC_ENUM) ERR; if (nc_inq_type(ncid, typeids[0], name_in, &base_size_in)) ERR; if (strcmp(name_in, TYPE_NAME) || base_size_in != sizeof(int)) ERR; if (nc_inq_enum(ncid, typeids[0], name_in, &base_nc_type, &base_size_in, &num_members)) ERR; if (strcmp(name_in, TYPE_NAME) || base_nc_type != NC_INT || num_members != NUM_MEMBERS) ERR; for (i = 0; i < NUM_MEMBERS; i++) { if (nc_inq_enum_member(ncid, typeid, i, name_in, &value_in)) ERR; if (strcmp(name_in, member_name[i]) || value_in != member_value[i]) ERR; if (nc_inq_enum_ident(ncid, typeid, member_value[i], name_in)) ERR; if (strcmp(name_in, member_name[i])) ERR; } if (nc_close(ncid)) ERR; 5.4 Learn About an User Defined Type: nc_inq_user_type ====================================================== Given an ncid and a typeid, get the information about a user defined type. This function will work on any user defined type, whether compound, opaque, enumeration, or variable length array. Usage ===== nc_inq_user_type(int ncid, nc_type xtype, char *name, size_t *sizep, nc_type *base_nc_typep, size_t *nfieldsp, int *classp); `ncid' The ncid for the group containing the user defined type. `xtype' The typeid for this type, as returned by nc_def_compound, nc_def_opaque, nc_def_enum, nc_def_vlen, or nc_inq_var. `name' If non-NULL, the name of the user defined type will be copied here. It will be NC_MAX_NAME bytes or less. `sizep' If non-NULL, the size of the user defined type will be copied here. `base_nc_typep' If non-NULL, the base typeid will be copied here for vlen and enum types. `nfieldsp' If non-NULL, the number of fields will be copied here for enum and compound types. `classp' Return the class of the user defined type, NC_VLEN, NC_OPAQUE, NC_ENUM, or NC_COMPOUND. Errors ====== `NC_NOERR' No error. `NC_EBADTYPEID' Bad typeid. `NC_EBADFIELDID' Bad fieldid. `NC_EHDFERR' An error was reported by the HDF5 layer. Example ======= /* Create a file. */ if (nc_create(FILE_NAME, NC_NETCDF4, &ncid)) ERR; /* Create an enum type. */ if (nc_def_enum(ncid, NC_INT, TYPE_NAME, &typeid)) ERR; for (i = 0; i < NUM_MEMBERS; i++) if (nc_insert_enum(ncid, typeid, member_name[i], &member_value[i])) ERR; /* Check it out. */ if (nc_inq_user_type(ncid, typeid, name_in, &base_size_in, &base_nc_type_in, &nfields_in, &class_in)) ERR; if (strcmp(name_in, TYPE_NAME) || base_size_in != sizeof(int) || base_nc_type_in != NC_INT || nfields_in != NUM_MEMBERS || class_in != NC_ENUM) ERR; 5.5 Compound Types Introduction =============================== NetCDF-4 added support for compound types, which allow users to construct a new type - a combination of other types, like a C struct. Compound types are not supported in classic or 64-bit offset format files. To write data in a compound type, first use nc_def_compound to create the type, multiple calls to nc_insert_compound to add to the compound type, and then write data with the appropriate nc_put_var1, nc_put_vara, nc_put_vars, or nc_put_varm call. To read data written in a compound type, you must know its structure. Use the nc_inq_compound functions to learn about the compound type. 5.6 Creating a Compound Type: nc_def_compound ============================================= Create a compound type. Provide an ncid, a name, and a total size (in bytes) of one element of the completed compound type. After calling this function, fill out the type with repeated calls to nc_insert_compound (*note nc_insert_compound::). Call nc_insert_compound once for each field you wish to insert into the compound type. Note that there does not seem to be a way to read such types into structures in Fortran 90 (and there are no structures in Fortran 77). Usage ===== int nc_def_compound(int ncid, size_t size, char *name, nc_type *typeidp); `ncid' The groupid where this compound type will be created. `size' The size, in bytes, of the compound type. `name' The name of the new compound type. `typeidp' A pointer to an nc_type. The typeid of the new type will be placed there. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENAMEINUSE' That name is in use. `NC_EMAXNAME' Name exceeds max length NC_MAX_NAME. `NC_EBADNAME' Name contains illegal characters. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag NC_NETCDF4. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. `NC_EPERM' Attempt to write to a read-only file. `NC_ENOTINDEFINE' Not in define mode. Example ======= struct s1 { int i1; int i2; }; struct s1 data[DIM_LEN], data_in[DIM_LEN]; /* Create a file with a compound type. Write a little data. */ if (nc_create(FILE_NAME, NC_NETCDF4, &ncid)) ERR; if (nc_def_compound(ncid, sizeof(struct s1), SVC_REC, &typeid)) ERR; if (nc_insert_compound(ncid, typeid, BATTLES_WITH_KLINGONS, HOFFSET(struct s1, i1), NC_INT)) ERR; if (nc_insert_compound(ncid, typeid, DATES_WITH_ALIENS, HOFFSET(struct s1, i2), NC_INT)) ERR; if (nc_def_dim(ncid, STARDATE, DIM_LEN, &dimid)) ERR; if (nc_def_var(ncid, SERVICE_RECORD, typeid, 1, dimids, &varid)) ERR; if (nc_put_var(ncid, varid, data)) ERR; if (nc_close(ncid)) ERR; 5.7 Inserting a Field into a Compound Type: nc_insert_compound ============================================================== Insert a named field into a compound type. Usage ===== int nc_insert_compound(nc_type typeid, char *name, size_t offset, nc_type field_typeid); `typeid' The typeid for this compound type, as returned by nc_def_compound, or nc_inq_var. `name' The name of the new field. `offset' Offset in byte from the beginning of the compound type for this field. `field_typeid' The type of the field to be inserted. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENAMEINUSE' That name is in use. Field names must be unique within a compound type. `NC_EMAXNAME' Name exceed max length NC_MAX_NAME. `NC_EBADNAME' Name contains illegal characters. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag NC_NETCDF4. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. `NC_ENOTINDEFINE' Not in define mode. Example ======= 5.8 Inserting an Array Field into a Compound Type: nc_insert_array_compound =========================================================================== Insert a named field into a compound type. Usage ===== int nc_insert_array_compound(int ncid, nc_type xtype, char *name, size_t offset, nc_type field_typeid, int ndims, int *dim_sizes); `ncid' The ID of the file that contains the array type and the compound type. `xtype' The typeid for this compound type, as returned by nc_def_compound, or nc_inq_var. `name' The name of the new field. `offset' Offset in byte from the beginning of the compound type for this field. `field_typeid' The base type of the array to be inserted. Errors ====== `NC_NOERR' No error. `NC_EBADID' Bad group id. `NC_ENAMEINUSE' That name is in use. Field names must be unique within a compound type. `NC_EMAXNAME' Name exceed max length NC_MAX_NAME. `NC_EBADNAME' Name contains illegal characters. `NC_ENOTNC4' Attempting a netCDF-4 operation on a netCDF-3 file. NetCDF-4 operations can only be performed on files defined with a create mode which includes flag NC_NETCDF4. (*note nc_open::). `NC_ESTRICTNC3' This file was created with the strict netcdf-3 flag, therefore netcdf-4 operations are not allowed. (*note nc_open::). `NC_EHDFERR' An error was reported by the HDF5 layer. `NC_ENOTINDEFINE' Not in define mode. `NC_ETYPEDEFINED' Attempt to change type that has already been committed. The first time the file leaves define mode, all defined types are committed, and can't be changed. If you wish to add an array to a compound type, you must do so before the compound type is committed. Example ======= This example comes from the test file libsrc4/tst_compounds.c, which writes data about some Star Fleet officers who are known to use netCDF data. /* Since some aliens exists in different, or more than one, * dimensions, StarFleet keeps track of the dimensional abilities * of everyone on 7 dimensions. */ #define NUM_DIMENSIONS 7 struct dim_rec { int starfleet_id; int abilities[NUM_DIMENSIONS]; }; struct dim_rec dim_data_out[DIM_LEN], dim_data_in[DIM_LEN]; /* Create some phoney data. */ for (i=0; i