\chapter{Application Programming Interface (API)}


\exo{} files can be written and read by application codes
written in C, C++, or Fortran via calls to functions in the application
programming interface (API). Functions within the API are categorized
as data file utilities, model description functions, or results
data functions.

In general, the following pattern is followed for writing
data objects to a file:

\begin{enumerate}
\item create the file with \cfuncref{ex_create};
\item write out global parameters to the file using \cfuncref{ex_put_init};
\item write out specific data object parameters; for example,
put out element block parameters with \cfuncref{ex_put_elem_block};
\item write out the data object; for example, put out the connectivity
for an element block with \cfuncref{ex_put_elem_conn};
\item close the file with \cfuncref{ex_close}.
\end{enumerate}

Steps 3 and 4 are repeated within this pattern for each data object
(i.e., nodal coordinates, element blocks, node sets, side sets,
results variables, etc.). For some data object types, steps 3 and 4
are combined in a single call. For instance, \cfuncref{ex_put_qa}
writes out the parameters (number of QA records) as well as the data
object itself (the QA records). During the database writing process,
there are a few order dependencies (e.g., an element block must be
written before element variables for that element block are written)
which are documented in the description of each library function.


The invocation of the \exo{} API functions for reading
data is order independent, providing random read access. The
following steps are typically used for reading data:

\begin{enumerate}

\item open the file with \cfuncref{ex_open};

\item read the global parameters for dimensioning purposes with
\cfuncref{ex_get_init};

\item read specific data object parameters; for example, read
node set parameters with \cfuncref{ex_get_node_set_param};

\item read the data object; for example, read the node set node
list with \cfuncref{ex_get_node_set};

\item close the file with \cfuncref{ex_close}.
\end{enumerate}

Again, steps 3 and 4 are repeated for each object. For some object
parameters, step 3 may be accomplished with a call to \cfuncref{ex_inquire}
 to inquire the size of certain objects.


In developing applications using the \exo{} API, the following
points may prove beneficial:

\begin{itemize}

 \item All functions that write objects to the database begin with
 \cfuncref{ex_put_}; functions that read objects from
 the database begin with \cfuncref{ex_get_}.

 \item Function arguments are classified as readable \R{}, writable
 \W{}, or both \RW{}. Readable arguments are not modified by the API
 routines; writable arguments are modified; read-write arguments may
 be either depending on the value of the argument.

 \item All application codes which use the \exo{} API must include
 the file `exodusII.h' for C. This file defines constants that are
 used (1) as arguments to the API routines, (2) to set global
 parameters such as maximum string length and database version, and
 (3) as error condition or function return values.

 \item Throughout this section, sample code segments have been
 included to aid the application developer in using the API
 routines. These segments are not complete and there has been no
 attempt to include all calling sequence dependencies within
 them.

 \item Because 2-dimensional arrays cannot be statically dimensioned,
 either dynamic dimensioning or user indexing is required. Most of the
 sample code segments utilize user indexing within 1-dimensional
 arrays even though the variables are logically 2-dimensional.

 \item There are many \code{NetCDF} utilities that prove useful. ncdump,
 which converts a binary \code{NetCDF} file to a readable ASCII version of
 the file, is the most notable.

 \item Because \code{NetCDF} buffers I/O, it is important to flush
 all buffers with \cfuncref{ex_update} when debugging an
 application that produces an \exo{} file.
\end{itemize}

\section{Data File Utilities}\label{sec:utilities}

This section describes data file utility functions for creating /
opening a file, initializing a file with global parameters, reading /
writing information text, inquiring on parameters stored in the data
file, and error reporting.

\subsection{Create \exo{} File}

The function \cfuncref{ex_create} creates a new
\exo{} file and returns an ID that can subsequently be used to
refer to the file.

All floating point values in an \exo{} file are stored as either
4-byte (``float'') or 8-byte (``double'') numbers; no mixing
of 4- and 8-byte numbers in a single file is allowed. An application
code can compute either 4- or 8-byte values and can designate that the
values be stored in the \exo{} file as either 4- or 8-byte numbers;
conversion between the 4- and 8-byte values is performed automatically
by the API routines. Thus, there are four possible combinations of
compute word size and storage (or I/O) word size.

In case of an error, \cfuncref{ex_create} returns a negative
number. Possible causes of errors include:

\begin{itemize}

 \item Passing a file name that includes a directory that does not
 exist.

 \item Specifying a file name of a file that exists and also
 specifying a no clobber option.

 \item Attempting to create a file in a directory without permission
 to create files there.

 \item Passing an invalid file clobber mode.
\end{itemize}

\funcdef{ex_create}{char~*path, int~mode, int~*comp_ws, int~*io_ws}

\begin{parameters}
\item[{char* path \R{}}]
The file name of the new \exo{} file. This can be given as either an
absolute path name (from the root of the file system) or a relative
path name (from the current directory).


\item[{int mode \R{}}]
Mode. Use one of the following predefined constants:

\begin{description}
\item[\param{EX_NOCLOBBER}]
To create the new file only if the given file name does not refer to a
file that already exists.

\item[\param{EX_CLOBBER}]
To create the new file, regardless of whether a file with the same
name already exists. If a file with the same name does exist, its
contents will be erased.

\item[\param{EX_LARGE_MODEL}]
To create a model that can store individual datasets larger than
2~gigabytes. This modifies the internal storage used by exodusII and
also puts the underlying \code{NetCDF} file into the ``64-bit offset''
mode. See Appendix~\ref{app:largemodel} for more details on this
mode.\footnote{A ``large model'' file will also be created if the
environment variable
\code{EXODUS_LARGE_MODEL}\index{Environment
Variable!EXODUS_LARGE_MODEL}\index{EXODUS_LARGE_MODEL} is defined
in the users environment. A message will be printed to standard output
if this environment variable is found.}

\item[\param{EX_NORMAL_MODEL}]
Create a standard model.

\item[\param{EX_NETCDF4}]
To create a model using the \code{HDF5}-based \code{NetCDF-4}
output. (Future capability)\footnote{\code{NetCDF-4} is currently in
beta mode; however, it will be used for ExodusII when available, so
this mode is being defined here for future completeness. An
\code{HDF5}-based \code{NetCDF-4} file will also be created if the
environment variable \code{EXODUS_NETCDF4}\index{Environment
Variable!EXODUS_NETCDF4}\index{EXODUS_NETCDF4} is defined in the
users environment. A message will be printed to standard output if
this environment variable is found.}

\item[\param{EX_NOSHARE}]
Do not open the underlying \code{NetCDF} file in ``share'' mode. See the
\code{NetCDF} documentation for more details.

\item[\param{EX_SHARE}]
Do open the underlying \code{NetCDF} file in ``share'' mode. See the \code{NetCDF}
documentation for more details.
\end{description}

\item[{int* comp_ws \RW{}}]
{The word size in bytes (0, 4 or 8) of the floating point variables
used in the application program. If 0 (zero) is passed, the default
sizeof(float) will be used and returned in this variable. WARNING: all
\exo{} functions requiring floats must be passed floats declared with
this passed in or returned compute word size (4 or 8).}

\item[{int* io_ws \R{}}] {The word size in bytes (4 or 8) of the
floating point data as they are to be stored in the \exo{} file.}
\end{parameters}

The following code segment creates an \exo{} file called \file{test.exo}:

\begin{lstlisting}
#include "exodusII.h"
int CPU_word_size, IO_word_size, exoid;
CPU_word_size = sizeof(float); /* use float or double */
IO_word_size = 8; /* store variables as doubles */

/* create \exo{} file */
exoid = ex_create ("test.exo"      /* filename path */
		    EX_CLOBBER,     /* create mode */
		    &CPU_word_size, /* CPU float word size in bytes */
	            &IO_word_size); /* I/O float word size in bytes */
\end{lstlisting}


\subsection{Open \exo{} File}

The function \cfuncref{ex_open} opens an existing \exo{} file and
returns an ID that can subsequently be used to refer to the file, the
word size of the floating point values stored in the file, and the
version of the \exo{} database (returned as a ``float'', regardless of
the compute or I/O word size). Multiple files may be ``open''
simultaneously.


In case of an error, \cfuncref{ex_open} returns a negative
number. Possible causes of errors include:

\begin{itemize}
\item The specified file does not exist.

 \item The mode specified is something other than the predefined
constant \param{EX_READ} or \param{EX_WRITE}.


 \item Database version is earlier than 2.0.
\end{itemize}


\funcdef{ex_open}{char~*path, int~mode, int~*comp_ws, int~*io_ws, float~*version}

\begin{parameters}
\item[{char* path \R{}}]
The file name of the \exo{} file. This can be given as either an
absolute path name (from the root of the file system) or a relative
path name (from the current directory).

\item[{int mode \R{}}]
Access mode. Use one of the following predefined constants:

\begin{description}
 \item [EX_READ] To open the file just for reading.

 \item [EX_WRITE] To open the file for writing and reading.
\end{description}

\item[{int* comp_ws \RW{}}]
The word size in bytes (0, 4 or 8) of the floating point variables
used in the application program. If 0 (zero) is passed, the default
size of floating point values for the machine will be used and
returned in this variable. WARNING: all \exo{} functions requiring
reals must be passed reals declared with this passed in or returned
compute word size (4 or 8).


\item[{int* io_ws \RW{}}]
The word size in bytes (0, 4 or 8) of the floating
point data as they are stored in the \exo{} file. If the word
size does not match the word size of data stored in the file,
a fatal error is returned. If this argument is 0, the word size
of the floating point data already stored in the file is returned.

\item[{float* version \W{}}]
Returned \exo{} database version number. The current version is
\version{}
\end{parameters}

The following opens an \exo{} file named \file{test.exo} for read
only, using default settings for compute and I/O word sizes:

\begin{lstlisting}
#include "exodusII.h"
int CPU_word_size,IO_word_size, exoid;
float version;

CPU_word_size = sizeof(float);   /* float or double */
IO_word_size = 0;                /* use what is stored in file */

/* open \exo{} files */
exoid = ex_open ("test.exo",     /* filename path */
	         EX_READ,        /* access mode = READ */
		 &CPU_word_size, /* CPU word size */
		 &IO_word_size,  /* IO word size */
	         &version);      /* ExodusII library version */
\end{lstlisting}


\subsection{Close \exo{} File}

The function \cfuncref{ex_close} updates and
then closes an open \exo{} file.

In case of an error, \cfuncref{ex_close} returns a negative number; a
warning will return a positive number. Possible causes of errors
include:
\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}
\end{itemize}


\funcdef{ex_close}{int~exoid}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.
\end{parameters}

The following code segment closes an open \exo{} file:

\begin{lstlisting}
int error,exoid;
error = ex_close (exoid);
\end{lstlisting}



\subsection{Write Initialization Parameters}

The function \cfuncref{ex_put_init} writes the
initialization parameters to the \exo{}
file. This function must be called once (and only once) before
writing any data to the file.


In case of an error, \cfuncref{ex_put_init} returns a negative number;
a warning will return a positive number.  Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}

 \item data file opened for read only.

 \item this routine has been called previously.
\end{itemize}


\funcdef{ex_put_init}{int~exoid, char~*title, int~num_dim, int~num_nodes,
int~num_elem, int~num_elem_blk, int~num_node_sets, int~num_side_sets}

\begin{parameters}
\item[int exoid \R{}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{char* titletitle \R{}}]
Database title. Maximum length is \param{MAX_LINE_LENGTH}.

\item[{int num_dim \R{}}]
The dimensionality of the database. This is the number of coordinates
per node.

\item[{int num_nodes \R{}}]
The number of nodal points.

\item[{int num_elem \R{}}]
The number of elements.

\item[{int num_elem_blk \R{}}]
The number of element blocks.

\item[{int num_node_sets \R{}}]
The number of node sets.

\item[{int num_side_sets \R{}}]
The number of side sets.
\end{parameters}

The following code segment will initialize an open \exo{} file with
the specified parameters:

\begin{lstlisting}
int num_dim, num_nods, num_el, num_el_blk, num_ns, num_ss, error, exoid;

/* initialize file with parameters */
num_dim = 3; num_nods = 46; num_el = 5; num_el_blk = 5;
num_ns = 2; num_ss = 5;

error = ex_put_init (exoid, "This is the title", num_dim,
                     num_nods, num_el,num_el_blk, num_ns, num_ss);
\end{lstlisting}


\subsection{Read Initialization Parameters}

The function \cfuncref{ex_get_init} reads the
initialization parameters from an opened
\exo{} file.

In case of an error, \cfuncref{ex_get_init} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}.
\end{itemize}

\funcdef{ex_get_init}{int~exoid, char~*title, int~num_dim, int~num_nodes,
int~num_elem, int~num_elem_blk, int~num_node_sets, int~num_side_sets}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{char* titletitle \W{}}]
Returned database title. String length may be up to
\param{MAX_LINE_LENGTH} bytes.

\item[{int* num_dim \W{}}]
Returned dimensionality of the database. This is the number of
coordinates per node.

\item[{int* num_nodes \W{}}]
Returned number of nodal points.

\item[{int* num_elem \W{}}]
Returned number of elements.

\item[{int* num_elem_blk \W{}}]
Returned number of element blocks.

\item[{int* num_node_sets \W{}}]
Returned number of node sets.

\item[{int* num_side_sets \W{}}]
Returned number of side sets.
\end{parameters}

The following code segment will read the initialization parameters
from the open \exo{} file:

\begin{lstlisting}
#include "exodusII.h"
int num_dim, num_nodes, num_elem, num_elem_blk,
    num_node_sets, num_side_sets, error, exoid;

char title[MAX_LINE_LENGTH+1];

/* read database parameters */
error = ex_get_init (exoid, title, &num_dim, &num_nodes,
                     &num_elem, &num_elem_blk, &num_node_sets, &num_side_sets);
\end{lstlisting}


\subsection{Write Quality Assurance (QA) Records}


The function \cfuncref{ex_put_qa} writes the QA records to the
database. Each QA record contains four \param{MAX_STR_LENGTH}-byte
character strings. The character strings are:

\begin{itemize}
\item the analysis code name
\item the analysis code QA descriptor
\item the analysis date
\item the analysis time
\end{itemize}

In case of an error, \cfuncref{ex_put_qa} returns a negative number; a
warning will return a positive number. Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}

 \item data file opened for read only.

 \item QA records already exist in file.
\end{itemize}


\funcdef{ex_put_qa}{int~exoid, int~num_qa_records, char~*qa_record[][4]}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int num_qa_records \R{}}]
The number of QA records.

\item[{char* qa_record \R{}}]
Array containing the QA records.
\end{parameters}

The following code segment will write out two QA records:

\begin{lstlisting}
#include "exodusII.h"
int num_qa_rec, error, exoid;
char *qa_record[2][4];

/* write QA records */
num_qa_rec = 2;

qa_record[0][0] = "TESTWT1";
qa_record[0][1] = "testwt1";
qa_record[0][2] = "07/07/93";
qa_record[0][3] = "15:41:33";
qa_record[1][0] = "FASTQ";
qa_record[1][1] = "fastq";
qa_record[1][2] = "07/07/93";
qa_record[1][3] = "16:41:33";

error = ex_put_qa (exoid, num_qa_rec, qa_record);
\end{lstlisting}




\subsection{Read Quality Assurance (QA) Records}

The function \cfuncref{ex_get_qa} reads the QA records from the
database. Each QA record contains four \param{MAX_STR_LENGTH}-byte
character strings. The character strings are:

\begin{itemize}
\item the analysis code name
\item the analysis code QA descriptor
\item the analysis date
\item the analysis time
\end{itemize}

Memory must be allocated for the QA records before this call is
made. The number of QA records can be determined by invoking
\cfuncref{ex_inquire}.

In case of an error, \cfuncref{ex_get_qa} returns a negative number; a
warning will return a positive number.  Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}

 \item a warning value is returned if no QA records were stored.
\end{itemize}

\funcdef{ex_get_qa}{int~exoid, char~*qa_record[][4]}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{char* qa_record \W{}}]
Returned array containing the QA records.
\end{parameters}


The following will determine the number of QA records and
read them from the open \exo{} file:

\begin{lstlisting}
#include "exodusII.h"
int num_qa_rec, error, exoid
char *qa_record[MAX_QA_REC][4];

/* read QA records */
num_qa_rec = ex_inquire_int(exoid, EX_INQ_QA);

for (i=0; i\texttt{<}num_qa_rec; i++) {
    for (j=0; j\texttt{<}4; j++)
    qa_record[i][j] = (char *) calloc ((MAX_STR_LENGTH+1), sizeof(char));
}
error = ex_get_qa (exoid, qa_record);
\end{lstlisting}


\subsection{Write Information Records}

The function \cfuncref{ex_put_info} writes information records to the
database. The records are MAX_LINE_LENGTH-character strings.

In case of an error, \cfuncref{ex_put_info} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item information records already exist in file.
\end{itemize}

\funcdef{ex_put_info}{int~exoid, int~num_info, char~**info}


\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.


\item[{int num_info \R{}}]
The number of information records.


\item[{char** info \R{}}]
Array containing the information records.
\end{parameters}

The following code will write out three information records
to an open \exo{} file:

\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_info;
char *info[3];

/* write information records */
num_info = 3;

info[0] = "This is the first information record.";
info[1] = "This is the second information record.";
info[2] = "This is the third information record.";

error = ex_put_info(exoid, num_info, info);
\end{lstlisting}

\subsection{Read Information Records}

The function \cfuncref{ex_get_info} reads information records from the
database. The records are \param{MAX_LINE_LENGTH}-character
strings. Memory must be allocated for the information records before
this call is made. The number of records can be determined by invoking
\cfuncref{ex_inquire} or \cfuncref{ex_inquire_int}.


In case of an error, \cfuncref{ex_get_info} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:


\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}

 \item a warning value is returned if no information records were
 stored.
\end{itemize}



\funcdef{ex_get_info}{int~exoid, char`**info}


\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[char** info \W{}]
Returned array containing the information records.
\end{parameters}

The following code segment will determine the number of information
records and read them from an open \exo{} file:


\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_info;
char *info[MAXINFO];

/* read information records */
num_info = ex_inquire_int (exoid,EX_INQ_INFO);
for (i=0; i < num_info; i++) {
   info[i] = (char *) calloc ((MAX_LINE_LENGTH+1), sizeof(char));
}
error = ex_get_info (exoid, info);
\end{lstlisting}



\subsection{Inquire \exo{} Parameters}\label{s:inquire}


The function \cfuncref{ex_inquire} is used to inquire values of certain
data entities in an \exo{} file. Memory must be allocated for the
returned values before this function is invoked.query database


In case of an error, \cfuncref{ex_inquire} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}.
 \item requested information not stored in the file.
 \item invalid request flag.
\end{itemize}


\funcdef{ex_inquire}{int~exoid, ex_inquiry~req_info, int~*ret_int, float~*ret_float, char~*ret_char}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_inquiry req_info \R{}}]
A flag which designates what information is requested. It must be one
of the following constants (predefined in the file
\file{exodusII.h}):

\begin{longtable}{@{}lp{4.4in}}
 \param{EX_INQ_API_VERS}& The \exo{} API version number is returned
 in \cmd{ret_float} and an ``undotted'' version number is returned in
 \cmd{ret_int}. The API version number reflects the release of the
 function library (i.e., function names, argument list, etc.). The
 current API version is {\version} or {\versionud}\footnote{The API
 and DB version numbers are synchronized and will always
 match. Initially, it was thought that maintaining the two versions
 separately would be a benefit, but that was more confusing than
 helpful, so the numbers were made the same awhile ago}.\\

 \param{EX_INQ_DB_VERS}& The \exo{} database version number is
 returned in \cmd{ret_float} and an ``undotted'' version number is
 returned in \cmd{ret_int}. The database version number reflects the
 version of the library that was used to write the file pointed to by
 \cmd{exoid}. The current database version is {\version} or
 {\versionud}.\\

 \param{EX_INQ_LIB_VERS}& The \exo{} library version number is
 returned in \cmd{ret_float} and an ``undotted'' version number is
 returned in \cmd{ret_int}. The API library version number reflects
 the version number of the \exo{} library linked with this
 application. The current library version is {\version} or
 {\versionud}\\

 \param{EX_INQ_TITLE}& The title stored in the database is returned
 in \cmd{ret_char}.\\

 \param{EX_INQ_DIM}& The dimensionality, or number of coordinates
 per node (1, 2 or 3), of the database is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_NODES}& The number of nodes is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_ELEM}& The number of elements is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_ELEM_BLK}& The number of element blocks is returned
 in \cmd{ret_int}.\\

 \param{EX_INQ_NODE_SETS}& The number of node sets is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_NS_NODE_LEN}& The length of the concatenated node
 sets node list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_NS_DF_LEN}& The length of the concatenated node
 sets distribution list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_SIDE_SETS}& The number of side sets is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_SS_ELEM_LEN}& The length of the concatenated side
 sets element list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_SS_DF_LEN}& The length of the concatenated side
 sets distribution factor list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_SS_NODE_LEN}& The aggregate length of all of the
 side sets node lists is returned in \cmd{ret_int}.\\

 \param{EX_INQ_EB_PROP}& The number of integer properties stored
 for each element block is returned in \cmd{ret_int}; this number
 includes the property named ``ID''.\\

 \param{EX_INQ_NS_PROP}& The number of integer properties stored
 for each node set is returned in \cmd{ret_int}; this number includes
 the property named ``ID''.\\

 \param{EX_INQ_SS_PROP}& The number of integer properties stored
 for each side set is returned in \cmd{ret_int}; this number includes
 the property named ``ID''.\\

 \param{EX_INQ_QA}& The number of QA records is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_INFO}& The number of information records is returned
 in \cmd{ret_int}.\\

 \param{EX_INQ_TIME}& The number of time steps stored in the
 database is returned in \cmd{ret_int}.\\

 \param{EX_INQ_EDGE_BLK} & The number of edge blocks is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_EDGE_MAP} & The number of edge maps is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_EDGE_PROP} & The number of properties stored per
 edge blockis returned in \cmd{ret_int}. \\

 \param{EX_INQ_EDGE_SETS} & The number of edge sets is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_EDGE} & The number of edges is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_FACE} & The number of faces is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_EB_PROP} & The number of element block properties is
 returned in \cmd{ret_int}.\\

 \param{EX_INQ_ELEM_MAP} & The number of element maps is returned
 in \cmd{ret_int}.\\

 \param{EX_INQ_ELEM_SETS} & The number of element sets is returned
 in \cmd{ret_int}.\\

 \param{EX_INQ_ELS_DF_LEN} & The length of the concatenated
 element set distribution factor list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_ELS_LEN} & The length of the concatenated element
 set element list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_ELS_PROP} & The number of properties stored per elem
 set is returned in \cmd{ret_int}.\\

 \param{EX_INQ_EM_PROP} & The number of element map properties is
 returned in \cmd{ret_int}.\\

 \param{EX_INQ_ES_DF_LEN} & The length of the concatenated edge
 set distribution factor list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_ES_LEN} & The length of the concatenated edge set
 edge list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_ES_PROP} & The number of properties stored per edge
 set is returned in \cmd{ret_int}.\\

 \param{EX_INQ_FACE_BLK} & The number of face blocks is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_FACE_MAP} & The number of face maps is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_FACE_PROP} & The number of properties stored per
 face block is returned in \cmd{ret_int}.\\

 \param{EX_INQ_FACE_SETS} & The number of face sets is returned in
 \cmd{ret_int}.\\

 \param{EX_INQ_FS_DF_LEN} & The length of the concatenated face
 set distribution factor list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_FS_LEN} & The length of the concatenated face set
 face list is returned in \cmd{ret_int}.\\

 \param{EX_INQ_FS_PROP} & The number of properties stored per face
 set is returned in \cmd{ret_int}.\\

 \param{EX_INQ_NM_PROP} & The number of node map properties is
 returned in \cmd{ret_int}.\\

 \param{EX_INQ_NODE_MAP} & The number of node maps is returned in
 \cmd{ret_int}.\\
\end{longtable}

\item[{int* ret_int \W{}}]
Returned integer, if an integer value is requested according
to \cmd{req_info}); otherwise, supply a dummy argument.

\item[float* ret_float \W{}]
Returned float, if a float value is requested (according
to \cmd{req_info}); otherwise, supply a dummy argument\footnote{NOTE:
This argument is always a float even if the database IO and/or CPU word
size is a double.}.

\item[{char* ret_char \W{}}]
Returned character string, if a character value is requested according
to \cmd{req_info}); otherwise, supply a dummy argument.
\end{parameters}

As an example, the following will return the number of element
block properties stored in the \exo{} file:

\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_props;
float fdum;
char *cdum;

/* determine the number of element block properties */
error = ex_inquire (exoid, EX_INQ_EB_PROP, &num_props,
                    &fdum, cdum);
\end{lstlisting}


\subsection{Inquire \exo{} Integer Parameters}


The function \cfuncref{ex_inquire_int} is used to query or inquire
values of certain integer data entities in an \exo{} file. It is a
short-cut to the \cfuncref{ex_inquire} function defined in the previous
section.  If there is no error, the queried value will be returned as
a positive number. In case of an error, \cfuncref{ex_inquire} returns a
negative number.

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create} or
 \cfuncref{ex_open}.
 \item requested information not stored in the file.
 \item invalid request flag.
\end{itemize}


\funcdef{ex_inquire_int}{int~exoid, ex_inquiry~req_info}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_inquiry req_info \R{}}]
A flag which designates what information is requested. It must be one
of the following constants (predefined in the file
\file{exodusII.h}):

\begin{longtable}{@{}lp{4.4in}}
 \param{EX_INQ_API_VERS}& The ``undotted'' \exo{} API version
 number is returned. The API version number reflects the release of
 the function library (i.e., function names, argument list, etc.). The
 current ``undotted'' API version is {\versionud}.\\

 \param{EX_INQ_LIB_VERS}& The ``undotted'' \exo{} API library
 version number is returned. The API library version number reflects
 the format of the data as it is stored in the \code{NetCDF}
 database. The current API version is {\versionud}\\

 \param{EX_INQ_DB_VERS}& The ``undotted'' \exo{} database version
 number is returned. The database version number reflects the version
 of the library that was used to write the file pointed to by
 \cmd{exoid}. The current database version is {\versionud}.\\

 \param{EX_INQ_DIM}& The dimensionality, or number of coordinates
 per node (1, 2 or 3), of the database is returned.\\

 \param{EX_INQ_NODES}& The number of nodes is returned.\\

 \param{EX_INQ_ELEM}& The number of elements is returned.\\

 \param{EX_INQ_ELEM_BLK}& The number of element blocks is
 returned.\\

 \param{EX_INQ_NODE_SETS}& The number of node sets is returned.\\

 \param{EX_INQ_NS_NODE_LEN}& The length of the concatenated node
 sets node list is returned.\\

 \param{EX_INQ_NS_DF_LEN}& The length of the concatenated node
 sets distribution list is returned.\\

 \param{EX_INQ_SIDE_SETS}& The number of side sets is returned.\\

 \param{EX_INQ_SS_ELEM_LEN}& The length of the concatenated side
 sets element list is returned.\\

 \param{EX_INQ_SS_DF_LEN}& The length of the concatenated side
 sets distribution factor list is returned.\\

 \param{EX_INQ_SS_NODE_LEN}& The aggregate length of all of the
 side sets node lists is returned.\\

 \param{EX_INQ_EB_PROP}& The number of integer properties stored
 for each element block is returned; this number includes the property
 named ``ID''.\\

 \param{EX_INQ_NS_PROP}& The number of integer properties stored
 for each node set is returned; this number includes the property
 named ``ID''.\\

 \param{EX_INQ_SS_PROP}& The number of integer properties stored
 for each side set is returned; this number includes the property
 named ``ID''.\\

 \param{EX_INQ_QA}& The number of QA records is returned.\\

 \param{EX_INQ_INFO}& The number of information records is returned.\\

 \param{EX_INQ_TIME}& The number of time steps stored in the
 database is returned.\\

 \param{EX_INQ_EDGE_BLK} & The number of edge blocks is returned.\\

 \param{EX_INQ_EDGE_MAP} & The number of edge maps is returned.\\

 \param{EX_INQ_EDGE_PROP} & The number of properties stored per
 edge block is returned. \\

 \param{EX_INQ_EDGE_SETS} & The number of edge sets is returned.\\

 \param{EX_INQ_EDGE} & The number of edges is returned.\\

 \param{EX_INQ_FACE} & The number of faces is returned.\\

 \param{EX_INQ_EB_PROP} & The number of element block properties is
 returned.\\

 \param{EX_INQ_ELEM_MAP} & The number of element maps is returned.\\

 \param{EX_INQ_ELEM_SETS} & The number of element sets is returned.\\

 \param{EX_INQ_ELS_DF_LEN} & The length of the concatenated
 element set distribution factor list is returned.\\

 \param{EX_INQ_ELS_LEN} & The length of the concatenated element
 set element list is returned.\\

 \param{EX_INQ_ELS_PROP} & The number of properties stored per elem
 set is returned.\\

 \param{EX_INQ_EM_PROP} & The number of element map properties is
 returned.\\

 \param{EX_INQ_ES_DF_LEN} & The length of the concatenated edge
 set distribution factor list is returned.\\

 \param{EX_INQ_ES_LEN} & The length of the concatenated edge set
 edge list is returned.\\

 \param{EX_INQ_ES_PROP} & The number of properties stored per edge
 set is returned.\\

 \param{EX_INQ_FACE_BLK} & The number of face blocks is returned.\\

 \param{EX_INQ_FACE_MAP} & The number of face maps is returned.\\

 \param{EX_INQ_FACE_PROP} & The number of properties stored per
 face block is returned.\\

 \param{EX_INQ_FACE_SETS} & The number of face sets is returned.\\

 \param{EX_INQ_FS_DF_LEN} & The length of the concatenated face
 set distribution factor list is returned.\\

 \param{EX_INQ_FS_LEN} & The length of the concatenated face set
 face list is returned.\\

 \param{EX_INQ_FS_PROP} & The number of properties stored per face
 set is returned.\\

 \param{EX_INQ_NM_PROP} & The number of node map properties is
 returned.\\

 \param{EX_INQ_NODE_MAP} & The number of node maps is returned.\\
\end{longtable}
\end{parameters}

As an example, the following will return the number of nodes,
elements, and element blocks stored in the \exo{} file:

\begin{lstlisting}
#include "exodusII.h"
int exoid;
int num_nodes = ex_inquire_int(exoid, EX_INQ_NODES);
int num_elems = ex_inquire_int(exoid, EX_INQ_ELEM);
int num_block = ex_inquire_int(exoid, EX_INQ_ELEM_BLK);
\end{lstlisting}


\subsection{Error Reporting}


The function \cfuncref{ex_err} logs an error to \cmd{stderr}. It is intended
to provide explanatory messages for error codes returned from other
\exo{} routines.This function

The passed in error codes and corresponding messages are listed in
Appendix C. The programmer may supplement the error message printed
for standard errors by providing an error message. If the error code
is provided with no error message, the predefined message will be
used. The error code \param{EX_MSG} is available to log application
specific messages.



\funcdefv{ex_err}{char~*module_name, char~*message, int~err_num}

\begin{parameters}
\item[{char* module_name \R{}}]
{This is a string containing the name of the calling function.}

\item[{char* message \R{}}]
This is a string containing a message explaining the error
or problem. If \param{EX_VERBOSE} (see \cfuncref{ex_opts}) is true,
this message will be printed to \cmd{stderr}. Otherwise,
nothing will be printed.

\item[{int err_num \R{}}]
This is an integer code identifying the error. \exo{} C functions
place an error code value in \cmd{exerrval}, an external int. Negative
values are considered fatal errors while positive values are
warnings. There is a set of predefined values defined in
\file{exodusII.h}. The predefined constant \param{EX_PRTLASTMSG} will
cause the last error message to be output, regardless of the setting
of the error reporting level (see \cfuncref{ex_opts}).
\end{parameters}

The following is an example of the use of this function:
\begin{lstlisting}
#include "exodusII.h"
int exoid, CPU_word_size, IO_word_size, errval;
float version;
char errmsg[MAX_ERR_LENGTH];

CPU_word_size = sizeof(float);
IO_word_size = 0;

/* open \exo{} file */
if (exoid = ex_open ("test.exo", EX_READ, &CPU_word_size,
                     &IO_word_size, &version)) {
   errval = 999;
   sprintf(errmsg,"Error: cannot open file test.exo");
   ex_err("prog_name", errmsg, errval);
}
\end{lstlisting}


\subsection{Set Error Reporting Level}

The function \cfuncref{ex_opts} is used to set message reporting
options.

In case of an error, \cfuncref{ex_opts} returns a negative number; a
warning will return a positive number.

\funcdef{ex_opts}{ex_options~option_val}

\begin{parameters}
\item[{int option_val \R{}}]
Integer option value. Current options are:

\begin{description}
\item[\param{EX_ABORT}] Causes fatal errors to force program
exit. (Default is false.)

\item[\param{EX_DEBUG}] Causes certain messages to print
for debug use. (Default is false.)

\item[\param{EX_VERBOSE}] Causes all error messages to
print when true, otherwise no error messages will print. (Default
is false.)
\end{description}
\end{parameters}


NOTE: Values may be OR'ed together to provide any combination
of these capabilities.

For example, the following will cause all messages to print
and will cause the program to exit upon receipt of fatal error:

\begin{lstlisting}
#include "exodusII.h"
ex_opts(EX_ABORT|EX_VERBOSE);
\end{lstlisting}

\section{Model Description}

The routines in this section read and write information which
describe an \exo{} finite element model. This includes nodal
coordinates, element order map, element connectivity arrays,
element attributes, node sets, side sets, and object properties.

\subsection{Write Nodal Coordinates}

The function \cfuncref{ex_put_coord} writes the nodal coordinates of
the nodes in the model. The function \cfuncref{ex_put_init} must be
invoked before this call is made.

Because the coordinates are floating point values, the application
code must declare the arrays passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_coord} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.
\end{itemize}



\funcdef{ex_put_coord}{int~exoid, void~*x_coor, void~*y_coor, void~*z_coor}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{void* x_coor \R{}}]
The X-coordinates of the nodes. If this is \code{NULL}, the
X-coordinates will not be written.

\item[{void* y_coor \R{}}]
The Y-coordinates of the nodes. These are stored only if \cmd{num_dim}
\texttt{>} 1; otherwise, pass in dummy address. If this is \code{NULL}, the
Y-coordinates will not be written.

\item[{void* z_coor \R{}}]
The Z-coordinates of the nodes. These are stored only if \cmd{num_dim}
\texttt{>} 2; otherwise, pass in dummy address. If this is \code{NULL}, the
Z-coordinates will not be written.
\end{parameters}

The following will write the nodal coordinates to an open
\exo{} file:

\begin{lstlisting}
int error, exoid;

/* if file opened with compute word size of sizeof(float) */
float x[8], y[8], z[8];

/* write nodal coordinates values to database */
x[0] = 0.0; y[0] = 0.0; z[0] = 0.0;
x[1] = 0.0; y[1] = 0.0; z[1] = 1.0;
x[2] = 1.0; y[2] = 0.0; z[2] = 1.0;
x[3] = 1.0; y[3] = 0.0; z[3] = 0.0;
x[4] = 0.0; y[4] = 1.0; z[4] = 0.0;
x[5] = 0.0; y[5] = 1.0; z[5] = 1.0;
x[6] = 1.0; y[6] = 1.0; z[6] = 1.0;
x[7] = 1.0; y[7] = 1.0; z[7] = 0.0;

error = ex_put_coord(exoid, x, y, z);

/* Do the same as the previous call in three separate calls */
error = ex_put_coord(exoid, x,    NULL, NULL);
error = ex_put_coord(exoid, NULL, y,    NULL);
error = ex_put_coord(exoid, NULL, NULL, z);
\end{lstlisting}


\subsection{Read Nodal Coordinates}




The function \cfuncref{ex_get_coord} reads the nodal coordinates of the
nodes. Memory must be allocated for the coordinate arrays (\cmd{x_coor},
\cmd{y_coor}, and \cmd{z_coor}) before this call is made. The length of each
of these arrays is the number of nodes in the mesh.

Because the coordinates are floating point values, the application
code must declare the arrays passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_get_coord} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if nodal coordinates were not
 stored.
\end{itemize}

\funcdef{ex_get_coord}{int~exoid, void~*x_coor, void~*y_coor, void~*z_coor}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{void* x_coor \W{}}]
Returned X coordinates of the nodes. If this is \code{NULL}, the
X-coordinates will not be read.

\item[{void* y_coor \W{}}]
Returned Y coordinates of the nodes. These are returned only if
{num_dim} \texttt{>} 1; otherwise, pass in a dummy address. If this
is \code{NULL}, the Y-coordinates will not be read.

\item[{void* z_coor \W{}}]
Returned Z coordinates of the nodes. These are returned only if
{num_dim} \texttt{>} 2; otherwise, pass in a dummy address. If this
is \code{NULL}, the Z-coordinates will not be read.
\end{parameters}

The following code segment will read the nodal coordinates
from an open \exo{} file:

\begin{lstlisting}
int error, exoid;

float *x, *y, *z;

/* read nodal coordinates values from database */
x = (float *)calloc(num_nodes, sizeof(float));
y = (float *)calloc(num_nodes, sizeof(float));
if (num_dim >= 3)
   z = (float *)calloc(num_nodes, sizeof(float));
else
   z = 0;

error = ex_get_coord(exoid, x, y, z);

/* Do the same as the previous call in three separate calls */
error = ex_get_coord(exoid, x,    NULL, NULL);
error = ex_get_coord(exoid, NULL, y,    NULL);
if (num_dim >= 3)
   error = ex_get_coord(exoid, NULL, NULL, z);
\end{lstlisting}


\subsection{Write Coordinate Names}

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.
\end{itemize}

\funcdef{ex_put_coord_names}{int~exoid, char~**coord_names}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{char** coord_names \R{}}]
Array containing \cmd{num_dim} names of length \param{MAX_STR_LENGTH}
of the nodal coordinate arrays.
\end{parameters}

The following coding will write the coordinate names to an
open \exo{} file:

\begin{lstlisting}
int error, exoid;

char *coord_names[3];
coord_names[0] = "xcoor";
coord_names[1] = "ycoor";
coord_names[2] = "zcoor";

error = ex_put_coord_names (exoid, coord_names);
\end{lstlisting}

\subsection{Read Coordinate Names}

The function \cfuncref{ex_get_coord_names} reads the names
(\param{MAX_STR_LENGTH}-characters in length) of the coordinate arrays
from the database. Memory must be allocated for the character strings
before this function is invoked.


In case of an error, \cfuncref{ex_get_coord_names} returns
a negative number; a warning will return a positive number.
Possible causes of errors include:


\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if coordinate names were not
 stored.
\end{itemize}

\funcdef{ex_get_coord_names}{int~exoid, char~**coord_names}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{char** coord_names \W{}}]
Returned pointer to a vector containing \cmd{num_dim} names of the nodal
coordinate arrays.
\end{parameters}

The following code segment will read the coordinate names from an open
\exo{} file:

\begin{lstlisting}
int error, exoid;
char *coord_names[3];

for (i=0; i < num_dim; i++) {
   coord_names[i] = (char *)calloc((MAX_STR_LENGTH+1), sizeof(char));
}

error = ex_get_coord_names (exoid, coord_names);
\end{lstlisting}


\subsection{Write Node Number Map}

The function \cfuncref{ex_put_node_num_map} writes out the optional
node number map to the database. See Section~\ref{s:nnm} for a
description of the node number map. The function \cfuncref{ex_put_init}
must be invoked before this call is made.

In case of an error, \cfuncref{ex_put_node_num_map} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item a node number map already exists in the file.
\end{itemize}


\funcdef{ex_put_node_num_map}{int~exoid, int~*node_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* node_map \R{}}]
The node number map.
\end{parameters}

The following code generates a default node number map and outputs it
to an open \exo{} file. This is a trivial case and included just for
illustration. Since this map is optional, it should be written out
only if it contains something other than the default map.

\begin{lstlisting}
int error, exoid;
int *node_map = (int *)calloc(num_nodes, sizeof(int));

for (i=1; i <= num_nodes; i++)
   node_map[i-1] = i;

error = ex_put_node_num_map(exoid, node_map);
\end{lstlisting}

\subsection{Read Node Number Map}

The function \cfuncref{ex_get_node_num_map} reads the optional node
number map from the database. See Section~\ref{s:nnm} for a
description of the node number map. If a node number map is
not stored in the data file, a default array (1,2,3,. .. \cmd{num_nodes})
is returned. Memory must be allocated for the node number map array
({num_nodes} in length) before this call is made.

In case of an error, \cfuncref{ex_get_node_num_map} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item if a node number map is not stored, a default map
and a warning value are returned.
\end{itemize}


\funcdef{ex_get_node_num_map}{int~exoid, int~*node_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* node_map \W{}}]
Returned node number map.
\end{parameters}

The following code will read a node number map from an open
\exo{} file:

\begin{lstlisting}
int *node_map, error, exoid;

/* read node number map */
node_map = (int *)calloc(num_nodes, sizeof(int));
error = ex_get_node_num_map(exoid, node_map);
\end{lstlisting}

\subsection{Write Element Number Map}

The function \cfuncref{ex_put_elem_num_map} writes out the optional
element number map to the database. See Section~\ref{s:enm} for a
description of the element number map. The function
\cfuncref{ex_put_init} must be invoked before this call is made.

In case of an error, \cfuncref{ex_put_elem_num_map} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item an element number map already exists in the file.
\end{itemize}



\funcdef{ex_put_elem_num_map}{int~exoid, int~*elem_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* elem_map \R{}}]
The element number map.
\end{parameters}


The following code generates a default element number map and outputs
it to an open \exo{} file. This is a trivial case and included just
for illustration. Since this map is optional, it should be written out
only if it contains something other than the default map.

\begin{lstlisting}
int error, exoid;
int *elem_map = (int *)calloc(num_elem, sizeof(int));

for (i=1; i <= num_elem; i++)
   elem_map[i-1] = i;

error = ex_put_elem_num_map(exoid, elem_map);
\end{lstlisting}

\subsection{Read Element Number Map}

The function \cfuncref{ex_get_elem_num_map} reads the optional
element number map from the database. See Section~\ref{s:enm} for a
description of the element number map. If an element number map is not
stored in the data file, a default array (1,2,3,. .. \cmd{num_elem}) is
returned. Memory must be allocated for the element number map array
({num_elem} in length) before this call is made.

In case of an error, \cfuncref{ex_get_elem_num_map} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item if an element number map is not stored, a default map and a
 warning value are returned.
\end{itemize}



\funcdef{ex_get_elem_num_map}{int~exoid, int~*elem_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* elem_map \W{}}]
Returned element number map.
\end{parameters}

The following code will read an element number map from an
open \exo{} file:
\begin{lstlisting}
int *elem_map, error, exoid;

/* read element number map */
elem_map = (int *) calloc(num_elem, sizeof(int));
error = ex_get_elem_num_map (exoid, elem_map);
\end{lstlisting}

\subsection{Write Element Order Map}

The function \cfuncref{ex_put_map} writes out the optional element
order map to the database. See Section~\ref{s:eom} for a description
of the element order map. The function \cfuncref{ex_put_init} must be
invoked before this call is made.

In case of an error, \cfuncref{ex_put_map} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item an element map already exists in the file.
\end{itemize}


\funcdef{ex_put_map}{int~exoid, int~*elem_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* elem_map \R{}}]
The element order map.
\end{parameters}

The following code generates a default element order map and outputs
it to an open \exo{} file. This is a trivial case and included just
for illustration. Since this map is optional, it should be written out
only if it contains something other than the default map.

\begin{lstlisting}
int error, exoid;
int *elem_map = (int *)calloc(num_elem, sizeof(int));
for (i=0; i < num_elem; i++) {
   elem_map[i] = i+1;
}
error = ex_put_map(exoid, elem_map);
\end{lstlisting}

\subsection{Read Element Order Map}

The function \cfuncref{ex_get_map} reads the element order map from
the database. See Section~\ref{s:eom} for a description of the element
order map. If an element order map is not stored in the data file, a
default array (1,2,3,. .. \cmd{num_elem}) is returned. Memory must be
allocated for the element map array ({num_elem} in length) before this
call is made.

In case of an error, \cfuncref{ex_get_map} returns a negative number; a
warning will return a positive number. Possible causes of errors
include:
\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item if an element order map is not stored, a default map and a
 warning value are returned.
\end{itemize}

\funcdef{ex_get_map}{int~exoid, int~*elem_map}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* elem_map \W{}}]
Returned element order map.
\end{parameters}

The following code will read an element order map from an
open \exo{} file:

\begin{lstlisting}
int *elem_map, error, exoid;

/* read element order map */
elem_map = (int *)calloc(num_elem, sizeof(int));

error = ex_get_map(exoid, elem_map);
\end{lstlisting}

\subsection{Write Element Block Parameters}\label{s:pebparam}

The function \cfuncref{ex_put_elem_block} writes the parameters used
to describe an element block.

In case of an error, \cfuncref{ex_put_elem_block} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item an element block with the same ID has already been specified.

 \item the number of element blocks specified in the call to
 \cfuncref{ex_put_init} has been exceeded.
\end{itemize}

\funcdef{ex_put_elem_block}{int~exoid, int~elem_blk_id,
char~*elem_type, int~num_elem_this_blk, int~num_nodes_per_elem, int~num_attr}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{char* elem_type \R{}}]
The type of elements in the element block. The maximum length of this
string is \param{MAX_STR_LENGTH}.

\item[{int num_elem_this_blk \R{}}]
The number of elements in the element block.

\item[{int num_nodes_per_elem \R{}}]
The number of nodes per element in the element block.

\item[{int num_attr \R{}}]
The number of attributes per element in the element block.
\end{parameters}

For example, the following code segment will initialize an
element block with an ID of 10, write out the connectivity array,
and write out the element attributes array:
\begin{lstlisting}
int id, error, exoid, num_elem_in_blk, num_nodes_per_elem,
    *connect, num_attr;

float *attrib;

/* write element block parameters */
id = 10;
num_elem_in_blk = 2;
num_nodes_per_elem = 4;   /* elements are 4-node shells */
num_attr = 1;             /* one attribute per element */

error = ex_put_elem_block(exoid, id, "SHEL", num_elem_in_blk,
                          num_nodes_per_elem, num_attr);

/* write element connectivity */
connect = (int *)calloc(num_elem_in_blk*num_nodes_per_elem, sizeof(int));

/* fill connect with node numbers; nodes for first element*/
connect[0] = 1; connect[1] = 2; connect[2] = 3; connect[3] = 4;

/* nodes for second element */
connect[4] = 5; connect[5] = 6; connect[6] = 7; connect[7] = 8;

error = ex_put_elem_conn (exoid, id, connect);

/* write element block attributes */
attrib = (float *) calloc (num_attr*num_elem_in_blk, sizeof(float));

for (i=0, cnt=0; i < num_elem_in_blk; i++) {
   for (j=0; j < num_attr; j++, cnt++) {
      attrib[cnt] = 1.0;
   }
}

error = ex_put_elem_attr (exoid, id, attrib);
\end{lstlisting}

\subsection{Read Element Block Parameters}\label{s:gebparam}

The function \cfuncref{ex_get_elem_block} reads the parameters used to
describe an element block. IDs of all element blocks stored can be
determined by calling \cfuncref{ex_get_elem_blk_ids}.

In case of an error, \cfuncref{ex_get_elem_block} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:


\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item element block with specified ID is not stored in
the data file.
\end{itemize}

\funcdef{ex_get_elem_block}{int~exoid, int~elem_blk_id,
char~*elem_type, int~*num_elem_this_blk,
int~*num_nodes_per_elem, int~*num_attr}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{char* elem_type \W{}}]
Returned element type of elements in the element block.
The maximum length of this string is \param{MAX_STR_LENGTH}.

\item[{int* num_elem_this_blk \W{}}]
Returned number of elements in the element block.

\item[{int* num_nodes_per_elem \W{}}]
Returned number of nodes per element in the element block.

\item[{int* num_attr \W{}}]
Returned number of attributes per element in the element block.
\end{parameters}


As an example, the following code segment will read the parameters for
the element block with an ID of 10 and read the connectivity and
element attributes arrays from an open \exo{} file:
\begin{lstlisting}
#include "exodusII.h"
int id, error, exoid, num_el_in_blk, num_nod_per_el, num_attr,
    *connect;
float *attrib;
char elem_type[MAX_STR_LENGTH+1];

/* read element block parameters */
id = 10;

error = ex_get_elem_block(exoid, id, elem_type, &num_el_in_blk,
                          &num_nod_per_elem, &num_attr);

/* read element connectivity */
connect = (int *) calloc(num_nod_per_el*num_el_in_blk,
                         sizeof(int));

error = ex_get_elem_conn(exoid, id, connect);

/* read element block attributes */
attrib = (float *) calloc (num_attr * num_el_in_blk, sizeof(float));
error = ex_get_elem_attr (exoid, id, attrib);
\end{lstlisting}

\subsection{Read Element Blocks IDs}

The function \cfuncref{ex_get_elem_blk_ids} reads the IDs of all of
the element blocks. Memory must be allocated for the returned array of
({num_elem_blk}) IDs before this function is invoked. The required
size(\cmd{num_elem_blk}) can be determined via a call to
\cfuncref{ex_inquire} or \cfuncref{ex_inquire_int}.


In case of an error, \cfuncref{ex_get_elem_blk_ids} returns
a negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}
\end{itemize}

\funcdef{ex_get_elem_blk_ids}{int~exoid, int~*elem_blk_ids}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* elem_blk_ids \W{}}]
Returned array of the element blocks IDs. The order of the IDs in this
array reflects the sequence that the element blocks were introduced
into the file.
\end{parameters}

The following code segment reads all the element block IDs:

\begin{lstlisting}
int error, exoid, *idelbs, num_elem_blk;
idelbs = (int *) calloc(num_elem_blk, sizeof(int));

error = ex_get_elem_blk_ids (exoid, idelbs);
\end{lstlisting}

\subsection{Write Element Block Connectivity}

The function \cfuncref{ex_put_elem_conn} writes the connectivity array
for an element block. The function \cfuncref{ex_put_elem_block} must
be invoked before this call is made.


In case of an error, \cfuncref{ex_put_elem_conn} returns a
negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_elem_block} was not called previously.
\end{itemize}

\funcdef{ex_put_elem_conn}{int~exoid, int~elem_blk_id, int~*connect}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{int connect[num_elem_this_blk,num_nodes_per_elem] \R{}}]
The connectivity array; a list of nodes (internal node IDs;
See Section~\ref{s:nnm}) that define each element in the element
block. The node index cycles faster than the element index.
\end{parameters}

Refer to the code example in Section~\ref{s:ebparam} for an example of
writing the connectivity array for an element block.


\subsection{Read Element Block Connectivity}

The function \cfuncref{ex_get_elem_conn} reads the connectivity array
for an element block. Memory must be allocated for the connectivity
array(\cmd{num_elem_this_blk} $\times$ \cmd{num_nodes_per_elem} in length)
before this routine is called.

In case of an error, \cfuncref{ex_get_elem_conn} returns a
negative number; a warning will return a positive number.
Possible causes of errors include:
\begin{itemize}
 \item an element block with the specified ID is not stored in the
 file.
\end{itemize}

\funcdef{ex_get_elem_conn}{int~exoid, int~elem_blk_id, int~*connect}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{int connect[num_elem_this_blk,num_nodes_per_elem] \W{}}]
Returned connectivity array; a list of nodes (internal node
IDs; See Section~\ref{s:nnm}) that define each element. The
node index cycles faster than the element index.
\end{parameters}

Refer to the code example in Section~\ref{s:gebparam} for an example
of reading the connectivity for an element block.


\subsection{Write Element Block Attributes}

The function \cfuncref{ex_put_elem_attr} writes the attributes for an
element block. Each element in the element block must have the same
number of attributes, so there are(\cmd{num_attr} $\times$
{num_elem_this_blk}) attributes for each element block. The
function \cfuncref{ex_put_elem_block} must be invoked before this call
is made.

Because the attributes are floating point values, the application code
must declare the array passed to be the appropriate type (``float'' or
``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_elem_attr} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_elem_block} was not called previously for
 specified element block ID.

 \item \cfuncref{ex_put_elem_block} was called with 0 attributes
 specified.
\end{itemize}

\funcdef{ex_put_elem_attr}{int~exoid, int~elem_blk_id, void~*attrib}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{void attrib}{[} {num_elem_this_blk,num_attr}{]} {\R{}}]
The list of attributes for the element block. The \cmd{num_attr}
index cycles faster.
\end{parameters}

Refer to the code example in Section~\ref{s:pebparam} for an example
of writing the attributes array for an element block.


\subsection{Read Element Block Attributes}

The function \cfuncref{ex_get_elem_attr} reads the attributes for an
element block. Memory must be allocated for(\cmd{num_attr} $\times$
\cmd{num_elem_this_blk}) attributes before this routine is called.

Because the attributes are floating point values, the application code
must declare the array passed to be the appropriate type (``float'' or
``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_elem_attr} returns a
negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid element block ID.

 \item a warning value is returned if no attributes are stored in the
 file.
\end{itemize}


\funcdef{ex_get_elem_attr}{int~exoid, int~elem_blk_id, void~*attrib}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{void attrib}{[} {num_elem_this_blk,num_attr}{]} {\W{}}]
Returned list of(\cmd{num_attr} $\times$ {num_elem_this_blk}) attributes for
the element block, with the \cmd{num_attr} index cycling faster.
\end{parameters}

Refer to the code example in Section~\ref{s:gebparam} for an example
of reading the element attributes for an element block.

\subsection{Write Node Set Parameters}

The function \cfuncref{ex_put_node_set_param} writes the node set ID,
the number of nodes which describe a single node set, and the number
of node set distribution factors for the node set.

In case of an error, \cfuncref{ex_put_node_set_param} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item the number of node sets specified in the call to
 \cfuncref{ex_put_init} was zero or has been exceeded.

 \item a node set with the same ID has already been stored.

 \item the specified number of distribution factors is not zero and is
 not equal to the number of nodes.
\end{itemize}


\funcdef{ex_put_node_set_param}{int~exoid, int~node_set_id, int~num_nodes_in_set, int~num_dist_in_set}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int node_set_id \R{}}]
The node set ID.

\item[{int num_nodes_in_set \R{}}]
The number of nodes in the node set.

\item[{int num_dist_in_set \R{}}]
The number of distribution factors in the node set. This should be
either 0 (zero) for no factors, or should equal \cmd{num_nodes_in_set}.
\end{parameters}

The following code segment will write out a node set to an open \exo{}
file:

\begin{lstlisting}
int id, num_nodes_in_set, num_dist_in_set, error, exoid,
    *node_list;

float *dist_fact;

/* write node set parameters */
id = 20; num_nodes_in_set = 5; num_dist_in_set = 5;
error = ex_put_node_set_param(exoid, id, num_nodes_in_set,
                              num_dist_in_set);

/* write node set node list */
node_list = (int *) calloc (num_nodes_in_set, sizeof(int));
node_list[0] = 100; node_list[1] = 101; node_list[2] = 102;
node_list[3] = 103; node_list[4] = 104;

error = ex_put_node_set(exoid, id, node_list);

/* write node set distribution factors */
dist_fact = (float *) calloc (num_dist_in_set, sizeof(float));
dist_fact[0] = 1.0; dist_fact[1] = 2.0; dist_fact[2] = 3.0;
dist_fact[3] = 4.0; dist_fact[4] = 5.0;

error = ex_put_node_set_dist_fact(exoid, id, dist_fact);
\end{lstlisting}

\subsection{Read Node Set Parameters}

The function \cfuncref{ex_get_node_set_param} reads the number of
nodes which describe a single node set and the number of distribution
factors for the node set.

In case of an error, \cfuncref{ex_get_node_set_param} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no node sets are stored
in the file.

 \item incorrect node set ID.
\end{itemize}

\funcdef{ex_get_node_set_param}{int~exoid, int~node_set_id, int~*num_nodes_in_set, int~*num_dist_in_set}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int node_set_id \R{}}]
The node set ID.

\item[{int* num_nodes_in_set \W{}}]
Returned number of nodes in the node set.

\item[{int* num_dist_in_set \W{}}]
Returned number of distribution factors in the node set.
\end{parameters}

The following code segment will read a node set from an open
\exo{} file:
\begin{lstlisting}
int error, exoid, id, num_nodes_in_set, num_df_in_set, *node_list;

float *dist_fact;

/* read node set parameters */
id = 100;

error = ex_get_node_set_param(exoid, id, &num_nodes_in_set,
                              &num_df_in_set);

/* read node set node list */
node_list = (int *) calloc(num_nodes_in_set, sizeof(int));
error = ex_get_node_set(exoid, id, node_list);

/* read node set distribution factors */
if (num_df_in_set > 0) {
   dist_fact = (float *) calloc(num_nodes_in_set, sizeof(float));
   error = ex_get_node_set_dist_fact(exoid, id, dist_fact);
}
\end{lstlisting}

\subsection{Write Node Set}

The function \cfuncref{ex_put_node_set} writes the node list for a
single node set. The function \cfuncref{ex_put_node_set_param} must
be called before this routine is invoked.

In case of an error, \cfuncref{ex_put_node_set} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_node_set_param} not called previously.
\end{itemize}

\funcdef{ex_put_node_set}{int~exoid, int~node_set_id, int~*node_set_node_list}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int node_set_id \R{}}]
The node set ID.

\item[{int* node_set_node_list \R{}}]
Array containing the node list for the node set. Internal node IDs are
used in this list (See Section~\ref{s:nnm}).
\end{parameters}

Refer to the description of \cfuncref{ex_put_node_set_param} for a
sample code segment to write out a node set.



\subsection{Write Node Set Distribution Factors}

The function \cfuncref{ex_put_node_set_dist_fact} writes node set
distribution factors for a single node set. The function
\cfuncref{ex_put_node_set_param} must be called before this routine
is invoked.


Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_put_node_set_dist_fact} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_node_set_param} not called previously.

 \item a call to \cfuncref{ex_put_node_set_param} specified zero
 distribution factors.
\end{itemize}

\funcdef{ex_put_node_set_dist_fact}{int~exoid, int~node_set_id, void~8node_set_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int node_set_id \R{}}]
The node set ID.

\item[{void* node_set_dist_fact \R{}}]
Array containing the distribution factors in the node set.
\end{parameters}

Refer to the description of \cfuncref{ex_put_node_set_param} for a
sample code segment to write out the distribution factors for a node
set.


\subsection{Read Node Set Distribution Factors}

The function \cfuncref{ex_get_node_set_dist_fact} returns the node
set distribution factors for a single node set. Memory must be
allocated for the list of distribution factors(\cmd{num_dist_in_set}
in length) before this function is invoked.


Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_get_node_set_dist_fact} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item a warning value is returned if no distribution factors
were stored.
\end{itemize}

\funcdef{ex_get_node_set_dist_fact}{int~exoid, int~node_set_id, void~8node_set_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int node_set_id \R{}}]
The node set ID.

\item[{void* node_set_dist_fact \W{}}]
Returned array containing the distribution factors in the node set.
\end{parameters}

Refer to the description of \cfuncref{ex_get_node_set_param} for a
sample code segment to read a node set's distribution factors.


\subsection{Read Node Sets IDs }

The function \cfuncref{ex_get_node_set_ids} reads the IDs of all of
the node sets. Memory must be allocated for the returned array of
({num_node_sets}) IDs before this function is invoked.

In case of an error, \cfuncref{ex_get_node_set_ids} returns
a negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item a warning value is returned if no node sets are stored
in the file.
\end{itemize}

\funcdef{ex_get_node_set_ids}{int~exoid, int~*node_set_ids}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int}{* node_set_ids \W{}}]
Returned array of the node sets IDs. The order of the IDs in this array
reflects the sequence the node sets were introduced into the file.
\end{parameters}

As an example, the following code will read all of the node set IDs
from an open data file:

\begin{lstlisting}
int *ids, num_node_sets, error, exoid;

/* read node sets IDs */
ids = (int *) calloc(num_node_sets, sizeof(int));

error = ex_get_node_set_ids (exoid, ids);
\end{lstlisting}

\subsection{Write Concatenated Node Sets}

The function \cfuncref{ex_put_concat_node_sets} writes the node set
ID's, node sets node count array, node sets distribution factor count
array, node sets node list pointers array, node sets distribution
factor pointer, node set node list, and node set distribution factors
for all of the node sets. ``Concatenated node sets'' refers to the
arrays required to define all of the node sets (ID array, counts
arrays, pointers arrays, node list array, and distribution factors
array) as described in Section 3.10 on page 11. Writing concatenated
node sets is more efficient than writing individual node sets. See
Appendix~\ref{app:efficiency} for a discussion of efficiency issues.


Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_put_concat_node_sets} returns
a negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item the number of node sets specified in a call to
 \cfuncref{ex_put_init} was zero or has been exceeded.

 \item a node set with the same ID has already been stored.

 \item the number of distribution factors specified for one of the
 node sets is not zero and is not equal to the number of nodes in the
 same node set.
\end{itemize}


\funcdef{ex_put_concat_node_sets}{int~exoid, int~*node_set_ids,
int~*num_nodes_per_set,
int~*num_dist_per_set,
int~*node_sets_node_index,
int~*node_sets_dist_index,
int~*node_sets_node_list,
void~*node_sets_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* node_set_ids \R{}}]
Array containing the node set ID for each set.

\item[{int* num_nodes_per_set \R{}}]
Array containing the number of nodes for each set.

\item[{int* num_dist_per_set \R{}}]
Array containing the number of distribution factors for each set.

\item[{int* node_sets_node_index \R{}}]
Array containing the indices into the \cmd{node_set_node_list} which
are the locations of the first node for each set. These indices are
0-based.

\item[{int* node_sets_dist_index \R{}}]
Array containing the indices into the \cmd{node_set_dist_list} which
are the locations of the first distribution factor for each set. These
indices are 0-based.

\item[{int* node_sets_node_list \R{}}]
Array containing the nodes for all sets. Internal node IDs are used in
this list (See Section~\ref{s:nnm}).

\item[{void* node_sets_dist_fact \R{}}]
Array containing the distribution factors for all sets.
\end{parameters}

For example, the following code will write out two node sets
in a concatenated format:

\begin{lstlisting}
int ids[2], num_nodes_per_set[2], node_ind[2], node_list[8],
    num_df_per_set[2], df_ind[2], error, exoid;

float dist_fact[8];

ids[0] = 20; ids[1] = 21;
num_nodes_per_set[0] = 5; num_nodes_per_set[1] = 3;

node_ind[0] = 0; node_ind[1] = 5;

node_list[0] = 100; node_list[1] = 101; node_list[2] = 102;
node_list[3] = 103; node_list[4] = 104;
node_list[5] = 200; node_list[6] = 201; node_list[7] = 202;

num_df_per_set[0] = 5; num_df_per_set[1] = 3;

df_ind[0] = 0; df_ind[1] = 5;

dist_fact[0] = 1.0; dist_fact[1] = 2.0; dist_fact[2] = 3.0;
dist_fact[3] = 4.0; dist_fact[4] = 5.0;
dist_fact[5] = 1.1; dist_fact[6] = 2.1;
dist_fact[7] = 3.1;

error = ex_put_concat_node_sets (exoid, ids, num_nodes_per_set,
                                 num_df_per_set, node_ind, df_ind,
                                 node_list, dist_fact);
\end{lstlisting}

\subsection{Read Concatenated Node Sets}

The function \cfuncref{ex_get_concat_node_sets} reads the node set
ID's, node set node count array, node set distribution factors count
array, node set node pointers array, node set distribution factors
pointer array, node set node list, and node set distribution factors
for all of the node sets. ``Concatenated node sets'' refers to the
arrays required to define all of the node sets (ID array, counts
arrays, pointers arrays, node list array, and distribution factors
array) as described in Section 3.10 on page 11.

Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

The length of each of the returned arrays can be determined by
invoking \cfuncref{ex_inquire} or \cfuncref{ex_inquire_int}.

In case of an error, \cfuncref{ex_get_concat_node_sets} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:
\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no node sets are stored in the
 file.
\end{itemize}

\funcdef{ex_get_concat_node_sets}
{int~exoid,
int~*node_set_ids,
int~*num_nodes_per_set,
int~*num_dist_per_set,
int~*node_sets_node_index,
int~*node_sets_dist_index,
int~*node_sets_node_list,
void~*node_sets_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* node_set_ids \W{}}]
Returned array containing the node set ID for each set.

\item[{int* num_nodes_per_set \W{}}]
Returned array containing the number of nodes for each set.

\item[{int* num_dist_per_set \W{}}]
Returned array containing the number of distribution factors for each
set.

\item[{int* node_sets_node index \W{}}]
Returned array containing the indices into the \cmd{node_set_node_list}
which are the locations of the first node for each set. These indices
are 0-based.

\item[{int* node_sets_dist_index \W{}}]
Returned array containing the indices into the \cmd{node_set_dist_fact}
which are the locations of the first distribution factor for each
set. These indices are 0-based.

\item[{int* node_sets_node_list \W{}}]
Returned array containing the nodes for all sets. Internal node IDs
are used in this list (see Section~\ref{s:nnm}).

\item[{void* node_sets_dist_fact \W{}}]
Returned array containing the distribution factors for all sets.
\end{parameters}

As an example, the following code segment will read concatenated
node sets:

\begin{lstlisting}
#include "exodusII.h"

int error, exoid, num_node_sets, list_len, *ids,
    *num_nodes_per_set, *num_df_per_set, *node_ind,
    *df_ind, *node_list;

float *dist_fact

/* read concatenated node sets */
num_node_sets = ex_inquire_int(exoid, EX_INQ_NODE_SETS);

ids               = (int *) calloc(num_node_sets, sizeof(int));
num_nodes_per_set = (int *) calloc(num_node_sets, sizeof(int));
num_df_per_set    = (int *) calloc(num_node_sets, sizeof(int));
node_ind          = (int *) calloc(num_node_sets, sizeof(int));
df_ind            = (int *) calloc(num_node_sets, sizeof(int));

list_len = ex_inquire_int(exoid, EX_INQ_NS_NODE_LEN);
node_list = (int *) calloc(list_len, sizeof(int));

list_len = ex_inquire_int(exoid, EX_INQ_NS_DF_LEN);
dist_fact = (float *) calloc(list_len, sizeof(float));

error = ex_get_concat_node_sets (exoid, ids, num_nodes_per_set,
                                 num_df_per_set, node_ind, df_ind,
                                 node_list, dist_fact);
\end{lstlisting}


\subsection{Write Side Set Parameters}

The function \cfuncref{ex_put_side_set_param} writes the side set set
ID and the number of sides (faces on 3D element types; edges on 2D
element types) which describe a single side set, and the number of
side set distribution factors on the side set. Because each side of a
side set is completely defined by an element and a local side number,
the number of sides is equal to the number of elements in a side set.

In case of an error, \cfuncref{ex_put_side_set_param} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item the number of side sets specified in the call to
 \cfuncref{ex_put_init} was zero or has been exceeded.

 \item a side set with the same ID has already been stored.
\end{itemize}


\funcdef{ex_put_side_set_param}
{int~exoid,
int~side_set_id,
int~num_side_in_set,
int~num_dist_fact_in_set}

\begin{parameters}
\item[{int} {exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id} {\R{}}]
The side set ID.

\item[{int num_side_in_set} {\R{}}]
The number of sides (faces or edges) in the side set.

\item[{int num_dist_fact_in_set} {\R{}}]
The number of distribution factors on the side set.
\end{parameters}

The following code segment will write a side set to an open
\exo{} file:

\begin{lstlisting}
int error, exoid, id, num_sides, num_df,
    elem_list[2], side_list[2];

float dist_fact[4];

/* write side set parameters */
id = 30;

num_sides = 2;
num_df    = 4;

error = ex_put_side_set_param (exoid, id, num_sides, num_df);

/* write side set element and side lists */
elem_list[0] = 1; elem_list[1] = 2;
side_list[0] = 1; side_list[1] = 1;

error = ex_put_side_set (exoid, id, elem_list, side_list);

/* write side set distribution factors */
dist_fact[0] = 30.0; dist_fact[1] = 30.1;
dist_fact[2] = 30.2; dist_fact[3] = 30.3;

error = ex_put_side_set_dist_fact (exoid, id, dist_fact);
\end{lstlisting}



\subsection{Read Side Set Parameters}

The function \cfuncref{ex_get_side_set_param} reads the number of
sides (faces on 3D element types; edges on 2D element types) which
describe a single side set, and the number of side set distribution
factors on the side set.

In case of an error, \cfuncref{ex_get_side_set_param} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no side sets are stored
in the file.

 \item incorrect side set ID.
\end{itemize}

\funcdef{ex_get_side_set_param}
{int~exoid,
int~side_set_id,
int~*num_side_in_set,
int~*num_dist_fact_in_set}

\begin{parameters}
\item[{int exoid} {\R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[{int* num_side_in_set \W{}}]
Returned number of sides (faces or edges) in the side set.

\item[{int* num_dist_fact_in_set \W{}}]
Returned number of distribution factors on the side set.
\end{parameters}

The following coding will read all of the side sets from
an open \exo{} file:

\begin{lstlisting}
int num_side_sets, error, exoid, num_sides_in_set, num_df_in_set,
    num_elem_in_set, *ids, *elem_list, *side_list, *ctr_list,
    *node_list;

float *dist_fact;

num_side_sets = ex_inquire_int(exoid, EX_INQ_SIDE_SETS);
ids = (int *) calloc(num_side_sets, sizeof(int));
error = ex_get_side_set_ids (exoid, ids);

for (i=0; i < num_side_sets; i++) {
   error = ex_get_side_set_param (exoid, ids[i], tab &num_sides_in_set,
                                  tab &num_df_in_set);

   num_elem_in_set = num_sides_in_set;
   elem_list = (int *) calloc(num_elem_in_set, sizeof(int));
   side_list = (int *) calloc(num_sides_in_set, sizeof(int));
   error = ex_get_side_set (exoid, ids[i], elem_list, side_list);

   if (num_df_in_set > 0) {
      /* get side set node list to correlate to dist factors */
      ctr_list  = (int *) calloc(num_elem_in_set, sizeof(int));
      node_list = (int *) calloc(num_df_in_set, sizeof(int));
      dist_fact = (float *) calloc(num_df_in_set, sizeof(float));
      error = ex_get_side_set_node_list (exoid, ids[i], ctr_list,
                                         node_list);

      error = ex_get_side_set_dist_fact (exoid, ids[i], tab dist_fact);
   }
}
\end{lstlisting}


\subsection{Write Side Set}

The function \cfuncref{ex_put_side_set} writes the side set element
list and side set side (face on 3D element types; edge on 2D element
types) list for a single side set. The routine
\cfuncref{ex_put_side_set_param} must be called before this function
is invoked.

In case of an error, \cfuncref{ex_put_side_set} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_side_set_param} not called previously.
\end{itemize}

\funcdef{ex_put_side_set}
{int~exoid,
int~side_set_id,
int~*side_set_elem_list,
int~*side_set_side_list}

\begin{parameters}
\item[int exoid \R{}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[int* side_set_elem_list \R{}]
Array containing the elements in the side set. Internal element
IDs are used in this list (see  Section~\ref{s:nnm}).

\item[int* side_set_side_list \R{}]
Array containing the sides (faces or edges) in the side set.
\end{parameters}

For an example of a code segment to write a side set, refer
to the description for \cfuncref{ex_put_side_set_param}.

\subsection{Read Side Set}

The function \cfuncref{ex_get_side_set} reads the side set element
list and side set side (face for 3D element types; edge for 2D
element types) list for a single side set. Memory must be allocated
for the element list and side list (both are {num_side_in_set} in
length) before this function is invoked.


In case of an error, \cfuncref{ex_get_side_set} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no side sets are stored
in the file.

 \item incorrect side set ID.
\end{itemize}

\funcdef{ex_get_side_set}
{int~exoid,
int~side_set_id,
int~*side_set_elem_list,
int~*side_set_side_list}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[{int* side_set_elem_list \W{}}]
Returned array containing the elements in the side set. Internal
element IDs are used in this list (see  Section~\ref{s:nnm}).

\item[{int* side_set_side_list \W{}}]
Returned array containing the sides (faces or edges) in the
side set.
\end{parameters}

For an example of code to read a side set from an \exo{}
II file, refer to the description for \cfuncref{ex_get_side_set_param}.




\subsection{Write Side Set Distribution Factors}

The function \cfuncref{ex_put_side_set_dist_fact} writes side set
distribution factors for a single side set. The routine
\cfuncref{ex_put_side_set_param} must be called before this function
is invoked.


Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_side_set_dist_fact} returns
a negative number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_side_set_param} not called previously.

 \item a call to \cfuncref{ex_put_side_set_param} specified zero
 distribution factors.
\end{itemize}


\funcdef{ex_put_side_set_dist_fact}
{int~exoid,
int~side_set_id,
void~*side_set_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[{void* side_set_dist_fact \R{}}]
Array containing the distribution factors in the side set.
\end{parameters}

For an example of a code segment to write side set distribution
factors, refer to the description for
\cfuncref{ex_put_side_set_param}.




\subsection{Read Side Set Distribution Factors}

The function \cfuncref{ex_get_side_set_dist_fact} returns the side
set distribution factors for a single side set. Memory must be
allocated for the list of distribution factors
({num_dist_fact_in_set} in length) before this function is
invoked.

Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_side_set_dist_fact} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item a warning value is returned if no distribution factors
were stored.
\end{itemize}

\funcdef{ex_get_side_set_dist_fact}
{int~exoid,
int~side_set_id,
void~*side_set_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[{void* side_set_dist_fact \W{}}]
Returned array containing the distribution factors in the
side set.
\end{parameters}

For an example of code to read side set distribution factors from an
\exo{} file, refer to the description for
\cfuncref{ex_get_side_set_param}.




\subsection{Read Side Sets IDs }

The function \cfuncref{ex_get_side_set_ids} reads the IDs of all of
the side sets. Memory must be allocated for the returned array of
({num_side_sets}) IDs before this function is invoked.

In case of an error, \cfuncref{ex_get_side_set_ids} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no side sets are stored
in the file.
\end{itemize}



\funcdef{ex_get_side_set_ids}
{int~exoid,
int~*side_set_ids}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* side_set_ids \W{}}]
Returned array of the side set IDs. The order of the IDs in this array
reflects the sequence the side sets were introduced into the file.
\end{parameters}

For an example of code to read side set IDs from an \exo{} II file,
refer to the description for \cfuncref{ex_get_side_set_param}.

\subsection{Read Side Set Node List}

The function \cfuncref{ex_get_side_set_node_list} returns a node
count array and a list of nodes on a single side set. With the 2.0 and
later versions of the database, this node list isn't stored directly
but can be derived from the element number in the side set element
list, local side number in the side set side list, and the element
connectivity array. The application program must allocate memory for
the node count array and node list.

There is a one-to-one mapping (i.e., same order -- as shown in
Table{\nobreakspace}2, ``Side Set Node Ordering,'' on
page{\nobreakspace}16 -- and same number) between the nodes in the
side set node list and the side set distribution factors. Thus, if
distribution factors are stored for the side set of interest, the
required size for the node list is the number of distribution factors
returned by \cfuncref{ex_get_side_set_param}. If distribution factors
are not stored for the side set, the application program must allocate
a maximum size anticipated for the node list. This would be the
product of the number of elements in the side set and the maximum
number of nodes per side for all types of elements in the model, since
side sets can span across different element types.

The length of the node count array is the length of the side set
element list. For each entry in the side set element list, there is an
entry in the side set side list, designating a local side number. The
corresponding entry in the node count array is the number of nodes
which define the particular side. In conjunction with the side set
node list, this node count array gives an unambiguous nodal
description of the side set.

In case of an error, \cfuncref{ex_get_side_set_node_list} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no side sets are stored in the
 file.

 \item incorrect side set ID.
\end{itemize}


\funcdef{ex_get_side_set_node_list}
{int~exoid,
int~side_set_id,
int~*side_set_node_cnt_list,
int~*side_set_node_list}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int side_set_id \R{}}]
The side set ID.

\item[{int* side_set_node_cnt_list \W{}}]
Returned array containing the number of nodes for each side (face in
3D, edge in 2D) in the side set.

\item[{int* side_set_node_list \W{}}]
Returned array containing a list of nodes on the side set. Internal
node IDs are used in this list (see Section 3.5~\ref{s:nnm}).
\end{parameters}

For an example of code to read a side set node list from an \exo{}
file, refer to the description for \cfuncref{ex_get_side_set_param}.




\subsection{Write Concatenated Side Sets}

The function \cfuncref{ex_put_concat_side_sets} writes the side set
IDs, side set element count array, side set distribution factor count
array, side set element pointers array, side set distribution factors
pointers array, side set element list, side set side list, and side
set distribution factors. ``Concatenated side sets'' refers to the
arrays needed to define all of the side sets (ID array, side counts
array, node counts array, element pointer array, node pointer array,
element list, node list, and distribution factors array) as described
in Section 3.12 on page 15. Writing concatenated side sets is more
efficient than writing individual side sets. See
Appendix~\ref{app:efficiency} for a discussion of efficiency issues.

Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_concat_side_sets} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item the number of side sets specified in a call to
 \cfuncref{ex_put_init} was zero or has been exceeded.

 \item a side set with the same ID has already been stored.
\end{itemize}

\funcdef{ex_put_concat_side_sets}
{int~exoid,
int~*side_sets_ids,
int~*num_side_per_set,
int~*num_dist_per_set,
int~*side_sets_elem_index,
int~*side_sets_dist_index,
int~*side_sets_elem_list,
int~*side_sets_side_list,
void~*side_sets_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int* side_sets_ids \R{}}]
Array containing the side set ID for each set.

\item[{int* num_side_per_set \R{}}]
Array containing the number of sides for each set.

\item[{int* num_dist_per_set \R{}}]
Array containing the number of distribution factors for each set.

\item[{int* side_sets_elem_index \R{}}]
Array containing the indices into the \cmd{side_sets_elem_list} which
are the locations of the first element for each set. These indices are
0-based.

\item[{int* side_sets_dist_index \R{}}]
Array containing the indices into the \cmd{side_sets_dist_fact} which
are the locations of the first distribution factor for each set. These
indices are 0-based.

\item[{int* side_sets_elem_list \R{}}]
Array containing the elements for all side sets. Internal element IDs
are used in this list (see Section~\ref{s:enm}).

\item[{int* side_sets_side_list \R{}}]
Array containing the sides for all side sets.

\item[{void* side_sets_dist_fact \R{}}]
Array containing the distribution factors for all side sets.
\end{parameters}

The following coding will write out two side sets in a concatenated
format:
\begin{lstlisting}
int error, exoid, ids[2], num_side_per_set[2], elem_ind[2],
    num_df_per_set[2], df_ind[2], elem_list[4], side_list[4];

float dist_fact[8];

/* write concatenated side sets */
ids[0] = 30;
ids[1] = 31;

num_side_per_set[0] = 2;
num_side_per_set[1] = 2;

elem_ind[0] = 0;
elem_ind[1] = 2;

num_df_per_set[0] = 4;
num_df_per_set[1] = 4;

df_ind[0] = 0;
df_ind[1] = 4;

/* side set #1 */
elem_list[0] = 2; elem_list[1] = 2;
side_list[0] = 2; side_list[1] = 1;

dist_fact[0] = 30.0; dist_fact[1] = 30.1;
dist_fact[2] = 30.2; dist_fact[3] = 30.3;

/* side set #2 */
elem_list[2] = 1; elem_list[3] = 2;
side_list[2] = 4; side_list[3] = 3;

dist_fact[4] = 31.0; dist_fact[5] = 31.1;
dist_fact[6] = 31.2; dist_fact[7] = 31.3;

error = ex_put_concat_side_sets (exoid, ids, num_side_per_set,
                                 num_df_per_set, elem_ind, df_ind,
                                 elem_list, side_list, dist_fact);
\end{lstlisting}


\subsection{Read Concatenated Side Sets}

The function \cfuncref{ex_get_concat_side_sets} reads the side set
IDs, side set element count array, side set distribution factors count
array, side set element pointers array, side set distribution factors
pointers array, side set element list, side set side list, and side
set distribution factors. ``Concatenated side sets'' refers to the
arrays needed to define all of the side sets (ID array, side counts
array, node counts array, element pointer array, node pointer array,
element list, node list, and distribution factors array) as described
in Section 3.12 on page 15.


Because the distribution factors are floating point values, the
application code must declare the array passed to be the appropriate
type (``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.


The length of each of the returned arrays can be determined by
invoking \cfuncref{ex_inquire} or \cfuncref{ex_inquire_int}.


In case of an error, \cfuncref{ex_get_concat_side_sets} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:
\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item a warning value is returned if no side sets are stored in the
 file.
\end{itemize}

\funcdef{ex_get_concat_side_sets}
{int~exoid,
int~*side_set_ids,
int~*num_side_per_set,
int~*num_dist_per_set,
int~*side_sets_elem_index,
int~*side_sets_dist_index,
int~*side_sets_elem_list,
int~*side_sets_side_list,
void~*side_sets_dist_fact}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* side_set_ids \W{}}]
Returned array containing the side set ID for each set.

\item[{int* num_side_per_set \W{}}]
Returned array containing the number of sides for each set.

\item[{int* num_dist_per_set \W{}}]
Returned array containing the number of distribution factors
for each set.

\item[{int* side_sets_elem_index \W{}}]
Returned array containing the indices into the
{side_sets_elem_list} which are the locations of the first element
for each set. These indices are 0-based.

\item[{int* side_sets_dist_index \W{}}]
Returned array containing the indices into the \cmd{side_sets_dist_fact}
array which are the locations of the first distribution factor
for each set. These indices are 0-based.

\item[{int* side_sets_elem_list \W{}}]
Returned array containing the elements for all side sets.
Internal element IDs are used in this list (see  Section~\ref{s:enm}).

\item[{int* side_sets_side_list \W{}}]
Returned array containing the sides for all side sets.

\item[{void* side_sets_dist_fact \W{}}]
Returned array containing the distribution factors for all
side sets.
\end{parameters}

The following code segment will return in concatenated format
all the side sets stored in an \exo{} file:

\begin{lstlisting}
#include "exodusII.h"

int error, exoid, num_ss, elem_list_len, df_list_len,
    *ids, *side_list, *num_side_per_set, *num_df_per_set,
    *elem_ind, *df_ind, *elem_list;

float *dist_fact;

num_ss = ex_inquire_int(exoid, EX_INQ_SIDE_SETS);

if (num_ss > 0) {
   elem_list_len = ex_inquire_int(exoid, EX_INQ_SS_ELEM_LEN);
   df_list_len   = ex_inquire_int(exoid, EX_INQ_SS_DF_LEN);

   /* read concatenated side sets */
   ids = (int *) calloc(num_ss, sizeof(int));
   num_side_per_set = (int *) calloc(num_ss, sizeof(int));
   num_df_per_set   = (int *) calloc(num_ss, sizeof(int));
   elem_ind         = (int *) calloc(num_ss, sizeof(int));
   df_ind           = (int *) calloc(num_ss, sizeof(int));
   elem_list        = (int *) calloc(elem_list_len, sizeof(int));
   side_list        = (int *) calloc(elem_list_len, sizeof(int));
   dist_fact        = (float *) calloc(df_list_len, sizeof(float));

   error = ex_get_concat_side_sets (exoid, ids, num_side_per_set,
                                    num_df_per_set, elem_ind, df_ind,
                                    elem_list, side_list,dist_fact);
}
\end{lstlisting}

\subsection{Convert Side Set Nodes to Sides}

The function \cfuncref{ex_cvt_nodes_to_sides} is used to convert a
side set node list to a side set side list. This routine is provided
for application programs that utilize side sets defined by nodes (as
was done previous to release 2.0) rather than local faces or
edges. The application program must allocate memory for the returned
array of sides. The length of this array is the same as the length of
the concatenated side sets element list, which can be determined with
a call to \cfuncref{ex_inquire} or
\cfuncref{ex_inquire_int}.

In case of an error, \cfuncref{ex_cvt_nodes_to_sides} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item a warning value is returned if no side sets are stored in the
 file.

 \item because the faces of a wedge require a different number of
 nodes to describe them (quadrilateral vs. triangular faces), the
 function will abort with a fatal return code if a wedge is
 encountered in the side set element list.
\end{itemize}

\funcdef{ex_cvt_nodes_to_sides}
{int~exoid,
int~*num_side_per_set,
int~*num_nodes_per_set,
int~*side_sets_elem_index,
int~*side_sets_node_index,
int~*side_sets_elem_list,
int~*side_sets_node_list,
int~*side_sets_side_list}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int* num_side_per_set \R{}}]
Array containing the number of sides for each set. The number
of sides is equal to the number of elements for each set.

\item[{int* num_nodes_per_set \R{}}]
Array containing the number of nodes for each set.

\item[{int* side_sets_elem_index \R{}}]
Array containing indices into the \cmd{side_sets_elem_list} which are
the locations of the first element for each set. These indices are
0-based.

\item[{int* side_sets_node_index \R{}}]
Array containing indices into the \cmd{side_sets_node_list}
which are the locations of the first node for each set. These
indices are 0-based.

\item[{int* side_sets_elem_list \R{}}]
Array containing the elements for all side sets. Internal element IDs
are used in this list (see Section~\ref{s:enm}).

\item[{int* side_sets_node_list \R{}}]
Array containing the nodes for all side sets. Internal node
IDs are used in this list (see  Section~\ref{s:nnm}).

\item[{int* side_sets_side_list \W{}}]
Returned array containing the sides for all side sets.
\end{parameters}

The following code segment will convert side sets described
by nodes to side sets described by local side numbers:

\begin{lstlisting}
int error, exoid, ids[2], num_side_per_set[2],
    num_nodes_per_set[2], elem_ind[2], node_ind[2],
    elem_list[4], node_list[8], el_lst_len, *side_list;

ids[0] = 30             ; ids[1]  = 31;
num_side_per_set[0]  = 2; num_side_per_set[1] = 2;
num_nodes_per_set[0] = 4; num_nodes_per_set[1] = 4;

elem_ind[0] = 0; elem_ind[1] = 2;
node_ind[0] = 0; node_ind[1] = 4;

/* side set #1 */
elem_list[0] = 2; elem_list[1] = 2;
node_list[0] = 8; node_list[1] = 5;
node_list[2] = 6; node_list[3] = 7;

/* side set #2 */
elem_list[2] = 1; elem_list[3] = 2;
node_list[4] = 2; node_list[5] = 3;
node_list[6] = 7; node_list[7] = 8;

el_lst_len = ex_inquire_int(exoid, EX_INQ_SS_ELEM_LEN);

/* side set element list is same length as side list */
side_list = (int *) calloc (el_lst_len, sizeof(int));

ex_cvt_nodes_to_sides(exoid, num_side_per_set, num_nodes_per_set,
                      elem_ind, node_ind, elem_list,
                      node_list, side_list);
\end{lstlisting}



\subsection{Write Property Arrays Names}

The function \cfuncref{ex_put_prop_names} writes object property names
and allocates space for object property arrays used to assign integer
properties to element blocks, node sets, or side sets. The property
arrays are initialized to zero (0). Although this function is
optional, since \cfuncref{ex_put_prop} will allocate space within the
data file if it hasn't been previously allocated, it is more efficient
to use \cfuncref{ex_put_prop_names} if there is more than one property
to store. See Appendix~\ref{app:efficiency} for a discussion of
efficiency issues.

In case of an error, \cfuncref{ex_put_prop_names} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item invalid object type specified.

 \item no object of the specified type is stored in the file.
\end{itemize}

\funcdef{ex_put_prop_names}
{int~exoid,
ex_entity_type~obj_type,
int~num_props,
char~**prop_names}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_entity_type obj_typ \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{int num_props \R{}}]
The number of integer properties to be assigned to all of the objects
of the type specified (element blocks, node sets, or side sets).

\item[{char** prop_names \R{}}]
Array containing \cmd{num_props} names (of maximum length
of \param{MAX_STR_LENGTH}) of properties to be stored.
\end{parameters}


For instance, suppose a user wanted to assign the 1st, 3rd, and 5th
element blocks (those element blocks stored 1st, 3rd, and 5th,
regardless of their ID) to a group (property) called ``TOP'', and the
2nd, 3rd, and 4th element blocks to a group called ``LSIDE''. This
could be accomplished with the following code:

\begin{lstlisting}
#include "exodusII.h";

char* prop_names[2];
int top_part[]   = {1,0,1,0,1};
int lside_part[] = {0,1,1,1,0};

int id[] = {10, 20, 30, 40, 50};

prop_names[0] = ``TOP'';
prop_names[1] = ``LSIDE'';

/* This call to ex_put_prop_names is optional, but more efficient */
ex_put_prop_names (exoid, EX_ELEM_BLOCK, 2, prop_names);

/* The property values can be output individually thus */
for (i=0; i < 5; i++) {
   ex_put_prop (exoid, EX_ELEM_BLOCK, id[i], prop_names[0],
                top_part[i]);

   ex_put_prop (exoid, EX_ELEM_BLOCK, id[i], prop_names[1],
                lside_part[i]);
}

/* Alternatively, the values can be output as an array thus*/
ex_put_prop_array (exoid, EX_ELEM_BLOCK, prop_names[0],
                   top_part);

ex_put_prop_array (exoid, EX_ELEM_BLOCK, prop_names[1],
                   lside_part);

\end{lstlisting}


\subsection{Read Property Arrays Names}

The function \cfuncref{ex_get_prop_names} returns names of integer
properties stored for an element block, node set, or side set. The
number of properties (needed to allocate space for the property names)
can be obtained via a call to \cfuncref{ex_inquire} or
\cfuncref{ex_inquire_int}.

In case of an error, \cfuncref{ex_get_prop_names} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid object type specified.
\end{itemize}

\funcdef{ex_get_prop_names}
{int~exoid,
ex_entity_type~obj_type,
char~**prop_names}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type obj_type \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{char** prop_names \W{}}]
eturned array containing \cmd{num_props} (obtained from call to
\cfuncref{ex_inquire} or \cfuncref{ex_inquire_int}) names (of maximum
length \param{MAX_STR_LENGTH}) of properties to be stored. ``ID'', a
reserved property name, will be the first name in the array.
\end{parameters}

As an example, the following code segment reads in properties
assigned to node sets:

\begin{lstlisting}
#include "exodusII.h";
int error, exoid, num_props, *prop_values;
char *prop_names[MAX_PROPS];

/* read node set properties */
num_props = ex_inquire_int(exoid, EX_INQ_NS_PROP);

for (i=0; i < num_props; i++) {
   prop_names[i] = (char *) malloc ((MAX_STR_LENGTH+1), sizeof(char));
   prop_values = (int *) malloc (num_node_sets, sizeof(int));
}

error = ex_get_prop_names(exoid,EX_NODE_SET,prop_names);

for (i=0; i < num_props; i++) {
   error = ex_get_prop_array(exoid, EX_NODE_SET, prop_names[i],
                             prop_values);
}
\end{lstlisting}


\subsection{Write Object Property}

The function \cfuncref{ex_put_prop} stores an integer object property
value to a single element block, node set, or side set. Although it is
not necessary to invoke \cfuncref{ex_put_prop_names}, since
\cfuncref{ex_put_prop} will allocate space within the data file if it
hasn't been previously allocated, it is more efficient to use
\cfuncref{ex_put_prop_names} if there is more than one property to
store. See Appendix~\ref{app:efficiency} for a discussion of
efficiency issues.


It should be noted that the interpretation of the values
of the integers stored as properties is left to the application
code. In general, a zero (0) means the object does not have the
specified property (or is not in the specified group); a nonzero
value means the object does have the specified property. When
space is allocated for the properties using \cfuncref{ex_put_prop_names}
or \cfuncref{ex_put_prop}, the properties are initialized to
zero (0).


Because the ID of an element block, node set, or side set
is just another property (named ``ID''), this routine can be used
to change the value of an ID. This feature must be used with
caution, though, because changing the ID of an object to the
ID of another object of the same type (element block, node set,
or side set) would cause two objects to have the same ID, and
thus only the first would be accessible. Therefore, \cfuncref{ex_put_prop}
issues a warning if a user attempts to give two objects the
same ID.


In case of an error, \cfuncref{ex_put_prop} returns a negative
number; a warning will return a positive number.
Possible causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item invalid object type specified.

 \item a warning is issued if a user attempts to change the ID of an
 object to the ID of an existing object of the same type.
\end{itemize}

\funcdef{ex_put_prop}
{int~exoid,
ex_entity_type~obj_type,
int~obj_id,
char~*prop_name,
int~value}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type obj_type \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{int obj_id \R{}}]
The element block, node set, or side set ID.

\item[{char* prop_name \R{}}]
The name of the property for which the value will be stored.
Maximum length of this string is \param{MAX_STR_LENGTH}.

\item[{int value \R{}}]
he value of the property.
\end{parameters}
For an example of code to write out an object property, refer
to the description for \cfuncref{ex_put_prop_names}.

\subsection{Read Object Property}

The function \cfuncref{ex_get_prop} reads an integer object property
value stored for a single element block, node set, or side set.

In case of an error, \cfuncref{ex_get_prop} returns a negative number;
a warning will return a positive number.  Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid object type specified.

 \item a warning value is returned if a property with the specified
 name is not found.
\end{itemize}

\funcdef{ex_get_prop}
{int~exoid,
ex_entity_type~obj_type,
int~obj_id,
char~*prop_name,
int~value}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type obj_type \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{int obj_id \R{}}]
The element block, node set, or side set ID.

\item[{char* prop_name \R{}}]
The name of the property (maximum length is \param{MAX_STR_LENGTH}) for
which the value is desired.

\item[{int* value \W{}}]
Returned value of the property.
\end{parameters}

For an example of code to read an object property, refer to the
description for \cfuncref{ex_get_prop_names}.


\subsection{Write Object Property Array}

The function \cfuncref{ex_put_prop_array} stores an array of
({num_elem_blk}, \cmd{num_node_sets}, or \cmd{num_side_sets}) integer
property values for all element blocks, node sets, or side sets. The
order of the values in the array must correspond to the order in which
the element blocks, node sets, or side sets were introduced into the
file. For instance, if the parameters for element block with ID 20
were written to a file (via \cfuncref{ex_put_elem_block}), and then
parameters for element block with ID 10, followed by the parameters
for element block with ID 30, the first, second, and third elements in
the property array would correspond to element block 20, element block
10, and element block 30, respectively.

One should note that this same functionality (writing properties to
multiple objects) can be accomplished with multiple calls to
\cfuncref{ex_put_prop}.

Although it is not necessary to invoke \cfuncref{ex_put_prop_names},
since \cfuncref{ex_put_prop_array} will allocate space within the data
file if it hasn't been previously allocated, it is more efficient to
use \cfuncref{ex_put_prop_names} if there is more than one property to
store. See Appendix~\ref{app:efficiency} for a discussion of
efficiency issues.

In case of an error, \cfuncref{ex_put_prop_array} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item invalid object type specified.
\end{itemize}



\funcdef{ex_put_prop_array}
{int~exoid,
ex_entity_type~obj_type,
char~*prop_name,
int~*values}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type obj_type \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{char* prop_name \R{}}]
The name of the property for which the values will be stored. Maximum
length of this string is \param{MAX_STR_LENGTH}.

\item[{int* values \R{}}]
An array of property values.
\end{parameters}

For an example of code to write an array of object properties, refer
to the description for \cfuncref{ex_put_prop_names}.




\subsection{Read Object Property Array}

The function \cfuncref{ex_get_prop_array} reads an array of integer
property values for all element blocks, node sets, or side sets. The
order of the values in the array correspond to the order in which the
element blocks, node sets, or side sets were introduced into the
file. Before this function is invoked, memory must be allocated for
the returned array of(\cmd{num_elem_blk}, \cmd{num_node_sets}, or
{num_side_sets}) integer values.


This function can be used in place of
\cfuncref{ex_get_elem_blk_ids},
\cfuncref{ex_get_node_set_ids}, and
\cfuncref{ex_get_side_set_ids}
to get element block, node set, and side set IDs, respectively, by
requesting the property name ``ID.'' One should also note that this
same function can be accomplished with multiple calls to
\cfuncref{ex_get_prop}.

In case of an error, \cfuncref{ex_get_prop_array} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:


\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid object type specified.

 \item a warning value is returned if a property with the specified
 name is not found.
\end{itemize}


\funcdef{ex_get_prop_array}
{int~exoid,
ex_entity_type~obj_type,
char~*prop_name,
int~*values}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_entity_type obj_type \R{}}]
Type of object; use one of the following options:\\

\begin{tabular}{ll}
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\param{EX_ELEM_MAP}  &  Element Map entity type \\
\param{EX_NODE_MAP}  &  Node Map entity type \\
\param{EX_EDGE_MAP}  &  Edge Map entity type \\
\param{EX_FACE_MAP}  &  Face Map entity type \\
\end{tabular}

\item[{char* prop_name \R{}}]
The name of the property (maximum length of \param{MAX_STR_LENGTH})
for which the values are desired.

\item[{int* values \W{}}]
Returned array of property values.
\end{parameters}

For an example of code to read an array of object properties, refer to
the description for \cfuncref{ex_get_prop_names}.

\section{Results Data}

This section describes functions which read and write analysis results
data and related entities. These include results variables (global,
elemental, and nodal), element variable truth table, and simulation
times.

\subsection{Write Results Variables Parameters}

The function \cfuncref{ex_put_variable_param} writes the number of global,
nodal, or element variables that will be written to the database.

In case of an error, \cfuncref{ex_put_variable_param} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item invalid variable type specified.

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item this routine has already been called with the same variable
 type; redefining the number of variables is not allowed.

 \item a warning value is returned if the number of variables
is specified as zero.
\end{itemize}

\funcdef{ex_put_variable_param}{
int~exoid,
ex_entity_type~var_type,
int~num_vars}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type var_type \R{}}]
Variable indicating the type of variable which is described. Use one
of the following options:\\

\begin{tabular}{ll}
\param{EX_GLOBAL}     &  Global entity type \\
\param{EX_NODAL}      &  Nodal entity type  \\
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\end{tabular}

\item[{int num_vars \R{}}]
The number of \cmd{var_type} variables that will be written to the
database.
\end{parameters}

For example, the following code segment initializes the data file to
store global variables:

\begin{lstlisting}
int num_glo_vars, error, exoid;

/* write results variables parameters */
num_glo_vars = 3;

error = ex_put_variable_param (exoid, EX_GLOBAL, num_glo_vars);
\end{lstlisting}

\subsection{Read Results Variables Parameters}

The function \cfuncref{ex_get_variable_param} reads the number of global,
nodal, or element variables stored in the database.

In case of an error, \cfuncref{ex_get_variable_param} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid variable type specified.
\end{itemize}


\funcdef{ex_get_variable_param}
{int~exoid,
ex_entity_typevar_type,
int~*num_vars}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{ex_entity_type var_type \R{}}]
Variable indicating the type of variable which is described. Use one
of the following options:\\

\begin{tabular}{ll}
\param{EX_GLOBAL}     &  Global entity type \\
\param{EX_NODAL}      &  Nodal entity type  \\
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\end{tabular}


\item[{int* num_vars \W{}}]
Returned number of \cmd{var_type} variables that are stored in the
database.
\end{parameters}
As an example, the following coding will determine the number of
global variables stored in the data file:

\begin{lstlisting}
int num_glo_vars, error, exoid;

/* read global variables parameters */
error = ex_get_variable_param(exoid, EX_GLOBAL, &num_glo_vars);
\end{lstlisting}

\subsection{Write Results Variables Names}

The function \cfuncref{ex_put_variable_names} writes the names of the
results variables to the database. The names are
\param{MAX_STR_LENGTH}-characters in length. The function
\cfuncref{ex_put_variable_param} must be called before this function is
invoked.

In case of an error, \cfuncref{ex_put_variable_names} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item invalid variable type specified.

 \item \cfuncref{ex_put_variable_param} was not called previously or was
 called with zero variables of the specified type.

 \item \cfuncref{ex_put_variable_names} has been called previously for the
 specified variable type.
\end{itemize}

\funcdef{ex_put_variable_names}
{int~exoid,
ex_entity_type~var_type,
int~num_vars,
char~**var_names[]}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_entity_type var_type \R{}}]
Variable indicating the type of variable which is described.
Use one of the following options:\\

\begin{tabular}{ll}
\param{EX_GLOBAL}     &  Global entity type \\
\param{EX_NODAL}      &  Nodal entity type  \\
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\end{tabular}

\item[{int num_vars \R{}}]
The number of \cmd{var_type} variables that will be written
to the database.

\item[{char** var_names \R{}}]
Array of pointers to \cmd{num_vars} variable names.
\end{parameters}

The following coding will write out the names associated with the
nodal variables:
\begin{lstlisting}
int num_nod_vars, error, exoid;
char *var_names[2];

/* write results variables parameters and names */
num_nod_vars = 2;

var_names[0] = "disx";
var_names[1] = "disy";

error = ex_put_variable_param (exoid, EX_NODAL, num_nod_vars);
error = ex_put_variable_names (exoid, EX_NODAL, num_nod_vars, var_names);
\end{lstlisting}

\subsection{Read Results Variable Names}

The function \cfuncref{ex_get_variable_names} reads the names of the
results variables from the database. Memory must be allocated for the
name array before this function is invoked. The names are
\param{MAX_STR_LENGTH}-characters in length.

In case of an error, \cfuncref{ex_get_variable_names} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:
\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item invalid variable type specified.

 \item a warning value is returned if no variables of the specified
 type are stored in the file.
\end{itemize}

\funcdef{ex_get_variable_names}
{int~exoid,
ex_entity_type~var_type,
int~num_vars,
char~*var_names[]}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{ex_entity_type var_type \R{}}]
Variable indicating the type of variable which is described. Use one
of the following options:\\

\begin{tabular}{ll}
\param{EX_GLOBAL}     &  Global entity type \\
\param{EX_NODAL}      &  Nodal entity type  \\
\param{EX_NODE_SET}  &  Node Set entity type \\
\param{EX_EDGE_BLOCK}&  Edge Block entity type \\
\param{EX_EDGE_SET}  &  Edge Set entity type \\
\param{EX_FACE_BLOCK}&  Face Block entity type \\
\param{EX_FACE_SET}  &  Face Set entity type \\
\param{EX_ELEM_BLOCK}&  Element Block entity type \\
\param{EX_ELEM_SET}  &  Element Set entity type \\
\param{EX_SIDE_SET}  &  Side Set entity type \\
\end{tabular}

\item[{int num_vars \R{}}]
The number of \cmd{var_type} variables that will be read
from the database.

\item[{char** var_names \W{}}]
Returned array of pointers to \cmd{num_vars} variable names.
\end{parameters}


As an example, the following code segment will read the names of the
nodal variables stored in the data file:

\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_nod_vars;
char *var_names[10];

/* read nodal variables parameters and names */
error = ex_get_variable_param(exoid, EX_NODAL, &num_nod_vars);
for (i=0; i < num_nod_vars; i++) {
   var_names[i] = (char *) calloc ((MAX_STR_LENGTH+1), sizeof(char));
}
error = ex_get_variable_names(exoid, EX_NODAL, num_nod_vars, var_names);
\end{lstlisting}

\subsection{Write Time Value for a Time Step}

The function \cfuncref{ex_put_time} writes the time value for a
specified time step.

Because time values are floating point values, the application code
must declare the array passed to be the appropriate type (``float'' or
``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_put_time} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.
\end{itemize}


\funcdef{ex_put_time}
{int~exoid,
int~time_step,
void~*time_value}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number. This is essentially a counter that is
incremented only when results variables are output to the data
file. The first time step is 1.

\item[{void* time_value \R{}}]
The time at the specified time step.
\end{parameters}

The following code segment will write out the simulation time value at
simulation time step n:

\begin{lstlisting}
int error, exoid, n;
float time_value;

/* write time value */
error = ex_put_time (exoid, n, &time_value);
\end{lstlisting}


\subsection{Read Time Value for a Time Step}

The function \cfuncref{ex_get_time} reads the time value for a
specified time step.

Because time values are floating point values, the application code
must declare the array passed to be the appropriate type (``float'' or
``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_time} returns a negative number;
a warning will return a positive number. Possible causes of errors
include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item no time steps have been stored in the file.
\end{itemize}

\funcdef{ex_get_time}
{int~exoid,
int~time_step,
void~*time_value}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number. This is essentially an index (in the time
dimension) into the global, nodal, and element variables arrays stored
in the database. The first time step is 1.

\item[{void* time_value \W{}}]
Returned time at the specified time step.
\end{parameters}

As an example, the following coding will read the time value stored in
the data file for time step n:

\begin{lstlisting}
int n, error, exoid;
float time_value;

/* read time value at time step 3 */
n = 3;
error = ex_get_time (exoid, n, &time_value);
\end{lstlisting}

\subsection{Read All Time Values}

The function \cfuncref{ex_get_all_times} reads the time values for all
time steps. Memory must be allocated for the time values array before
this function is invoked. The storage requirements (equal to the
number of time steps) can be determined by using the
\cfuncref{ex_inquire} or \cfuncref{ex_inquire_int} routines.

Because time values are floating point values, the application code
must declare the array passed to be the appropriate type (``float'' or
``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_all_times} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}

 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item no time steps have been stored in the file.
\end{itemize}


\funcdef{ex_get_all_times}{
int~exoid,
void~*time_values}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{void* time_values \W{}}]
Returned array of times. These are the time values at all time steps.
\end{parameters}

The following code segment will read the time values for all time
steps stored in the data file:

\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_time_steps;
float *time_values;

/* determine how many time steps are stored */
num_time_steps = ex_inquire_int(exoid, EX_INQ_TIME);

/* read time values at all time steps */
time_values = (float *) calloc(num_time_steps, sizeof(float));

error = ex_get_all_times(exoid, time_values);
\end{lstlisting}

\subsection{Write Element Variable Truth Table}

The function \cfuncref{ex_put_elem_var_tab} writes the \exo{} element
variable truth table to the database. The element variable truth table
indicates whether a particular element result is written for the
elements in a particular element block. A 0 (zero) entry indicates
that no results will be output for that element variable for that
element block. A non-zero entry indicates that the appropriate results
will be output.

Although writing the element variable truth table is optional, it is
encouraged because it creates at one time all the necessary
\code{NetCDF} variables in which to hold the \exo{} element variable
values. This results in significant time savings. See
Appendix~\ref{app:efficiency} for a discussion of efficiency issues.

The function \cfuncref{ex_put_variable_param} must be called before
this routine in order to define the number of element variables.

In case of an error, \cfuncref{ex_put_elem_var_tab} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item the specified number of element blocks is different than the
 number specified in a call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_elem_block} not called previously to specify
 element block parameters.

 \item \cfuncref{ex_put_variable_param} not called previously to specify
 the number of element variables or was called but with a different
 number of element variables.

 \item \cfuncref{ex_put_elem_var} previously called.
\end{itemize}


\funcdef{ex_put_elem_var_tab}
{int~exoid,
int~num_elem_blk,
int~num_elem_var,
int~**elem_var_tab}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int num_elem_blk \R{}}]
The number of element blocks.

\item[{int num_elem_var \R{}}]
The number of element variables.

\item[{int elem_var_tab[num_elem_blk,num_elem_var] \R{}}]
A 2-dimensional array (with the \cmd{num_elem_var} index
cycling faster) containing the element variable truth table.

\end{parameters}
The following coding will create, populate, and write an
element variable truth table to an opened \exo{} file (NOTE:
all element variables are valid for all element blocks in this
example.):

\begin{lstlisting}
int *truth_tab, num_elem_blk, num_ele_vars, error, exoid;

/* write element variable truth table */
truth_tab = (int *)calloc((num_elem_blk*num_ele_vars), sizeof(int));

for (i=0, k=0; i < num_elem_blk; i++) {
   for (j=0; j < num_ele_vars; j++) {
      truth_tab[k++] = 1;
   }
}
error = ex_put_elem_var_tab(exoid, num_elem_blk, num_ele_vars,
                            truth_tab);
\end{lstlisting}

\subsection{Read Element Variable Truth Table}

The function \cfuncref{ex_get_elem_var_tab} reads the \exo{} element
variable truth table from the database. For a description of the truth
table, see the usage of the function
\cfuncref{ex_put_elem_var_tab}. Memory must be allocated for
the truth table(\cmd{num_elem_blk} $\times$ \cmd{num_elem_var} in length)
before this function is invoked. If the truth table is not stored in
the file, it will be created based on information in the file and then
returned.

In case of an error, \cfuncref{ex_get_elem_var_tab} returns a
negative number; a warning will return a positive number.  Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item the specified number of element blocks is different than the
 number specified in a call to \cfuncref{ex_put_init}.

 \item there are no element variables stored in the file or the
 specified number of element variables doesn't match the number
 specified in a call to \cfuncref{ex_put_variable_param}.
\end{itemize}


\funcdef{ex_get_elem_var_tab}
{int~exoid,
int~num_elem_blk,
int~num_elem_var,
int~*elem_var_tab}

\begin{parameters}
\item[{int exoid \R{}}]
exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int num_elem_blk \R{}}]
The number of element blocks.

\item[{int num_elem_var \R{}}]
The number of element variables.

\item[{int elem_var_tab[num_elem_blk,num_elem_var}{]} {\W{}}]
Returned 2-dimensional array (with the \cmd{num_elem_var} index cycling
faster) containing the element variable truth table.
\end{parameters}

As an example, the following coding will read the element
variable truth table from an opened \exo{} file:

\begin{lstlisting}
int *truth_tab, num_elem_blk, num_ele_vars, error, exoid;

truth_tab = (int *) calloc ((num_elem_blk*num_ele_vars),
                            sizeof(int));

error = ex_get_elem_var_tab (exoid, num_elem_blk, num_ele_vars,
                             truth_tab);
\end{lstlisting}

\subsection{Write Element Variable Values at a Time Step}

The function \cfuncref{ex_put_elem_var} writes the values of a single
element variable for one element block at one time step. It is
recommended, but not required, to write the element variable truth
table (with \cfuncref{ex_put_elem_var_tab} before this function is
invoked for better efficiency. See Appendix~\ref{app:efficiency} for a
discussion of efficiency issues.

Because element variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_put_elem_var} returns a negative
number; a warning will return a positive number.  Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item invalid element block ID.

 \item \cfuncref{ex_put_elem_block} not called previously to specify
 parameters for this element block.

 \item \cfuncref{ex_put_variable_param} not called previously specifying
 the number of element variables.

 \item an element variable truth table was stored in the file but
 contains a zero (indicating no valid element variable) for the
 specified element block and element variable.
\end{itemize}



\funcdef{ex_put_elem_var}
{int~exoid,
int~time_step,
int~elem_var_index,
int~elem_blk_id,
int~num_elem_this_blk,
void~*elem_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number, as described under \cfuncref{ex_put_time}.
This is essentially a counter that is incremented only when results
variables are output. The first time step is 1.

\item[{int elem_var_index \R{}}]
The index of the element variable. The first variable has
an index of 1.

\item[{int elem_blk_id \R{}}]
The element block ID.

\item[{int num_elem_this_blk \R{}}]
The number of elements in the given element block.

\item[{void* elem_var_vals \R{}}]
Array of \cmd{num_elem_this_blk} values of the \cmd{elem_var_index}\th{}
element variable for the element block with ID of \cmd{elem_blk_id}
at the \cmd{time_step}\th{} time step.
\end{parameters}

The following coding will write out all of the element variables for a
single time step {n} to an open \exo{} file:

\begin{lstlisting}
int num_ele_vars, num_elem_blk, *num_elem_in_block,error,
    exoid, n, *ebids;

/* write element variables */
for (k=1; k <= num_ele_vars; k++) {
   for (j=0; j < num_elem_blk; j++) {
      float *elem_var_vals = (float *)
         calloc(num_elem_in_block[j], sizeof(float));

         for (m=0; m < num_elem_in_block[j]; m++) {
            /* simulation code fills this in */
            elem_var_vals[m] = 10.0;
         }

      error = ex_put_elem_var (exoid, n, k, ebids[j],
                               num_elem_in_block[j], elem_var_vals);
      free (elem_var_vals);
   }
}
\end{lstlisting}

\subsection{Read Element Variable Values at a Time Step}

The function \cfuncref{ex_get_elem_var} reads the values of a single
element variable for one element block at one time step. Memory must
be allocated for the element variable values array before this
function is invoked.

Because element variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.


In case of an error, \cfuncref{ex_get_elem_var} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item variable does not exist for the desired element block.

 \item invalid element block.
\end{itemize}

\funcdef{ex_get_elem_var}
{int~exoid,
int~time_step,
int~elem_var_index,
int~elem_blk_id,
int~num_elem_this_blk,
void~*elem_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number, as described under \cfuncref{ex_put_time}, at
which the element variable values are desired. This is essentially an
index (in the time dimension) into the element variable values array
stored in the database. The first time step is 1.

\item[{int elem_var_index \R{}}]
The index of the desired element variable. The first variable
has an index of 1.

\item[{int elem_blk_id \R{}}]
The desired element block ID.

\item[{int num_elem_this_blk \R{}}]
The number of elements in this element block.
\end{parameters}

{void* elem_var_vals \W{}}


Returned array of \cmd{num_elem_this_blk} values of the \cmd{elem_var_index}\th{}
element variable for the element block with ID of \cmd{elem_blk_id}
at the \cmd{time_step}\th{} time step.


As an example, the following code segment will read the
\cmd{var_index}\th{} element variable at one time step stored in an
\exo{} file:

\begin{lstlisting}
int num_elem_blk, error, exoid, *num_elem_in_block,  step, var_ind;

int *ids = (int *) calloc(num_elem_blk, sizeof(int));
error = ex_get_elem_blk_ids (exoid, ids);

step = 1; /* read at the first time step */
for (i=0; i < num_elem_blk; i++) {
   float *var_vals = (float *) calloc (num_elem_in_block[i], sizeof(float));
   error = ex_get_elem_var (exoid, step, var_ind, ids[i],
                            num_elem_in_block[i], var_vals);
   free (var_values);
}
\end{lstlisting}


\subsection{Read Element Variable Values through Time}

The function \cfuncref{ex_get_elem_var_time} reads the values of an
element variable for a single element through a specified number of
time steps. Memory must be allocated for the element variable values
array before this function is invoked.

Because element variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_elem_var_time} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file not initialized properly with call to \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_elem_block} not called previously to specify
 parameters for all element blocks.

 \item variable does not exist for the desired element or results
 haven't been written.
\end{itemize}


\funcdef{ex_get_elem_var_time}
{int~exoid,
int~elem_var_index,
int~elem_number,
int~beg_time_step,
int~end_time_step,
void~*elem_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int elem_var_index \R{}}]
The index of the desired element variable. The first variable has an
index of 1.

\item[{int elem_number \R{}}]
The internal ID (see Section~\ref{s:enm}) of the desired
element. The first element is 1.

\item[{int beg_time_step \R{}}]
The beginning time step for which an element variable value is
desired. This is not a time value but rather a time step number, as
described under \cfuncref{ex_put_time}. The first time step is 1.

\item[{int end_time_step \R{}}]
The last time step for which an element variable value is desired. If
negative, the last time step in the database will be used. The first
time step is 1.

\item[{void* elem_var_vals \W{}}]
Returned array of(\cmd{end_time_step} {-} \cmd{beg_time_step} +
1) values of the \cmd{elem_number}\th{} element for the \cmd{elem_var_index}\th{}
element variable.
\end{parameters}

For example, the following coding will read the values of the
\cmd{var_index}\th{} element variable for element number 2 from the first
time step to the last time step:

\begin{lstlisting}
/* determine how many time steps are stored */
int num_time_steps = ex_inquire_int(exoid, EX_INQ_TIME);

/* read an element variable through time */
float *var_values = (float *) calloc (num_time_steps, sizeof(float));
int var_index = 2;

int elem_num = 2;

int beg_time =  1;
int end_time = -1;

int error = ex_get_elem_var_time (exoid, var_index, elem_num,
                              beg_time, end_time, var_values);
\end{lstlisting}

\subsection{Write Global Variables Values at a Time Step}

The function \cfuncref{ex_put_glob_vars} writes the values of all the
global variables for a single time step. The function
\cfuncref{ex_put_variable_param} must be invoked before this call is made.

Because global variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_glob_vars} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item \cfuncref{ex_put_variable_param} not called previously specifying
 the number of global variables.
\end{itemize}

\funcdef{ex_put_glob_vars}
{int~exoid,
int~time_step,
int~num_glob_vars,
void~*glob_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number, as described under \cfuncref{ex_put_time}.
This is essentially a counter that is incremented when results
variables are output. The first time step is 1.

\item[{int num_glob_vars \R{}}]
The number of global variables to be written to the database.

\item[{void* glob_var_vals \R{}}]
Array of \cmd{num_glob_vars} global variable values for
the \cmd{time_step}\th{} time step.
\end{parameters}

As an example, the following coding will write the values of all the
global variables at one time step to an open \exo{} II file:

\begin{lstlisting}
int num_glo_vars, error, exoid, time_step;

float *glob_var_vals

/* write global variables */
for (j=0; j < num_glo_vars; j++) {
   /* application code fills this array */
   glob_var_vals[j] = 10.0;
}
error = ex_put_glob_vars (exoid, time_step, num_glo_vars,
                          glob_var_vals);
\end{lstlisting}

\subsection{Read Global Variables Values at a Time Step}

The function \cfuncref{ex_get_glob_vars} reads the values of all the
global variables for a single time step. Memory must be allocated for
the global variables values array before this function is invoked.

Because global variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_glob_vars} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item no global variables stored in the file.

 \item a warning value is returned if no global variables are stored
 in the file.
\end{itemize}


\funcdef{ex_get_glob_vars}
{int~exoid,
int~time_step,
int~num_glob_vars,
void~*glob_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step, as described under \cfuncref{ex_put_time}, at
which the global variable values are desired. This is essentially
an index (in the time dimension) into the global variable values
array stored in the database. The first time step is 1.

\item[{int num_glob_vars \R{}}]
The number of global variables stored in the database.

\item[{void* glob_var_vals \W{}}]
Returned array of \cmd{num_glob_vars} global variable values
for the \cmd{time_step}\th{} time step.
\end{parameters}

The following is an example code segment that reads all the global
variables at one time step:

\begin{lstlisting}
int num_glo_vars, time_step;

int error = ex_get_variable_param (idexo, EX_GLOBAL, &num_glo_vars);
float *var_values = (float *) calloc (num_glo_vars, sizeof(float));
error = ex_get_glob_vars (idexo, time_step, num_glo_vars,
                          var_values);
\end{lstlisting}



\subsection{Read Global Variable Values through Time}

The function \cfuncref{ex_get_glob_var_time} reads the values of a
single global variable through a specified number of time
steps. Memory must be allocated for the global variable values array
before this function is invoked.

Because global variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_glob_var_time} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
or \cfuncref{ex_open}

 \item specified global variable does not exist.

 \item a warning value is returned if no global variables
are stored in the file.
\end{itemize}


\funcdef{ex_get_glob_var_time}
{int~exoid,
int~glob_var_index,
int~beg_time_step,
int~end_time_step,
void~*glob_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int glob_var_index \R{}}]
The index of the desired global variable. The first variable
has an index of 1.

\item[{int beg_time_step \R{}}]
The beginning time step for which a global variable value is
desired. This is not a time value but rather a time step number, as
described under \cfuncref{ex_put_time}. The first time step is 1.

\item[{int end_time_step \R{}}]
The last time step for which a global variable value is desired. If
negative, the last time step in the database will be used. The first
time step is 1.

\item[{void* glob_var_vals \W{}}]
Returned array of (end_time_step - beg_time_step +
1) values for the \cmd{glob_var_index}$^{th}$ global variable.
\end{parameters}

The following is an example of using this function:

\begin{lstlisting}
#include "exodusII.h"
int error, exoid, num_time_steps, var_index;
int beg_time, end_time;

float *var_values;

/* determine how many time steps are stored */
num_time_steps = ex_inquire_int(exoid, EX_INQ_TIME);

/* read the first global variable for all time steps */
var_index = 1;
beg_time = 1;
end_time = -1;

var_values = (float *) calloc (num_time_steps, sizeof(float));

error = ex_get_glob_var_time(exoid, var_index, beg_time,
                             end_time, var_values);
\end{lstlisting}

\subsection{Write Nodal Variable Values at a Time Step}

The function \cfuncref{ex_put_nodal_var} writes the values of a single
nodal variable for a single time step. The function
\cfuncref{ex_put_variable_param} must be invoked before this call is made.

Because nodal variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_put_nodal_var} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:

\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item data file opened for read only.

 \item data file not initialized properly with call to
 \cfuncref{ex_put_init}.

 \item \cfuncref{ex_put_variable_param} not called previously specifying
 the number of nodal variables.
\end{itemize}

\funcdef{ex_put_nodal_var}
{int~exoid,
int~time_step,
int~nodal_var_index,
int~num_nodes,
void~*nodal_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step number, as described under \cfuncref{ex_put_time}. This
is essentially a counter that is incremented when results variables
are output. The first time step is 1.

\item[{int nodal_var_index \R{}}]
The index of the nodal variable. The first variable has an index of 1.

\item[{int num_nodes \R{}}]
 The number of nodal points.

\item[{void* nodal_var_vals \R{}}]
Array of \cmd{num_nodes} values of the \cmd{nodal_var_index}\th{} nodal
variable for the \cmd{time_step}\th{} time step.
\end{parameters}

As an example, the following code segment writes all the nodal
variables for a single time step:

\begin{lstlisting}
int num_nod_vars, num_nodes, error, exoid, time_step;

/* write nodal variables */
float *nodal_var_vals = (float *) calloc(num_nodes, sizeof(float));
for (k=1; k <= num_nod_vars; k++) {
   for (j=0; j < num_nodes; j++) {
      /* application code fills in this array */
      nodal_var_vals[j] = 10.0;
   }
   error = ex_put_nodal_var(exoid, time_step, k, num_nodes,
                            nodal_var_vals);
}
\end{lstlisting}

\subsection{Read Nodal Variable Values at a Time Step}

The function \cfuncref{ex_get_nodal_var} reads the values of a single
nodal variable for a single time step. Memory must be allocated for
the nodal variable values array before this function is invoked.


Because nodal variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed in
\cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_nodal_var} returns a negative
number; a warning will return a positive number. Possible causes of
errors include:
\begin{itemize}
 \item data file not properly opened with call to \cfuncref{ex_create}
 or \cfuncref{ex_open}

 \item specified nodal variable does not exist.

 \item a warning value is returned if no nodal variables are stored in
 the file.
\end{itemize}

\funcdef{ex_get_nodal_var}
{int~exoid,
int~time_step,
int~nodal_var_index,
int~num_nodes,
void~*nodal_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create}
or \cfuncref{ex_open}.

\item[{int time_step \R{}}]
The time step, as described under \cfuncref{ex_put_time}, at which the
nodal variable values are desired. This is essentially an index (in
the time dimension) into the nodal variable values array stored in the
database. The first time step is 1.

\item[{int nodal_var_index \R{}}]
The index of the desired nodal variable. The first variable
has an index of 1.

\item[{int num_nodes \R{}}]
The number of nodal points.

\item[{void* nodal_var_vals \W{}}]
Returned array of \cmd{num_nodes} values of the \cmd{nodal_var_index}\th{}
nodal variable for the \cmd{time_step}\th{} time step.
\end{parameters}

For example, the following demonstrates how this function would be
used:

\begin{lstlisting}
/* read the second nodal variable at the first time step */
int time_step = 1;
int var_index = 2;

float *var_values = (float *) calloc (num_nodes, sizeof(float));
error = ex_get_nodal_var(exoid, time_step, var_index, num_nodes,
                         var_values);
\end{lstlisting}


\subsection{Read Nodal Variable Values through Time}

The function \cfuncref{ex_get_nodal_var_time} reads the values of a
nodal variable for a single node through a specified number of time
steps. Memory must be allocated for the nodal variable values array
before this function is invoked.

Because nodal variables are floating point values, the application
code must declare the array passed to be the appropriate type
(``float'' or ``double'') to match the compute word size passed
in \cfuncref{ex_create} or \cfuncref{ex_open}.

In case of an error, \cfuncref{ex_get_nodal_var_time} returns a
negative number; a warning will return a positive number. Possible
causes of errors include:

\begin{itemize}
 \item specified nodal variable does not exist.

 \item a warning value is returned if no nodal variables are stored in
 the file.
\end{itemize}

\funcdef{ex_get_nodal_var_time}
{int~exoid,
int~nodal_var_index,
int~node_number,
int~beg_time_step,
int~end_time_step,
void~*nodal_var_vals}

\begin{parameters}
\item[{int exoid \R{}}]
\exo{} file ID returned from a previous call to \cfuncref{ex_create} or
\cfuncref{ex_open}.

\item[{int nodal_var_index \R{}}]
The index of the desired nodal variable. The first variable has an
index of 1.

\item[{int node_number \R{}}]
The internal ID (see  Section~\ref{s:nnm}) of the desired
node. The first node is 1.

\item[{int beg_time_step \R{}}]
The beginning time step for which a nodal variable value
is desired. This is not a time value but rather a time step number,
as described under \cfuncref{ex_put_time}. The first time step
is 1.

\item[{int end_time_step \R{}}]
The last time step for which a nodal variable value is desired. If
negative, the last time step in the database will be used. The first
time step is 1.

\item[{void* nodal_var_vals \W{}}]
Returned array of(\cmd{end_time_step} {-} \cmd{beg_time_step} +1) values
of the \cmd{node_number}\th{} node for the \cmd{nodal_var_index}\th{} nodal
variable.
\end{parameters}

For example, the following code segment will read the values
of the first nodal variable for node number one for all time
steps stored in the data file:
\begin{lstlisting}
#include "exodusII.h"
int node_num, beg_time, end_time, error, exoid;

/* determine how many time steps are stored */
int num_time_steps = ex_inquire_int(exoid, EX_INQ_TIME);

/* read a nodal variable through time */
float *var_values = (float *) calloc (num_time_steps, sizeof(float));

int var_index = 1; node_num = 1; beg_time = 1; end_time = -1;
error = ex_get_nodal_var_time(exoid, var_index, node_num, beg_time,
                              end_time, var_values);

\end{lstlisting}