Natgrid error handling

The error handling functions

Natgrid contains a set of procedures that implement a centralized error handling capability. The package invokes the C function ErrorHnd to produce all of its error messages. ErrorHnd simply invokes the ErrorLog function with the arguments specified in ErrorHnd. This allows a user to replace the error handling functionality by replacing or modifying the ErrorHnd function with one that handles errors in an application-dependent manner. All of the error handling functions are written in C; there is no way to replace the error handler using Fortran code.

Detailed description of the error functions

The error functions are:

ErrorHnd - primary error handler
ErrorLog - error logger
ErrMsg - retrieves error messages
ErrMax - retrieves the number of error messages

ErrorHnd - Primary error handler

ErrorHnd simply invokes the ErrorLog function with the arguments specified in ErrorHnd. This allows a user to replace the error handling functionality by replacing or modifying the ErrorHnd function with one that handles errors in an application-dependent manner.

Function prototype:

  void ErrorHnd (int, char *, FILE *, char *);

Argument description:

---------------------------------
              Argument | Type   
---------------------------------
void ErrorHnd (error,  | int    
               func,   | char * 
               efile,  | FILE * 
               smsg);  | char * 
---------------------------------

error: The error number, as described in the error table.
func: A string specifying the function that is invoking the error call.
efile: Specifies the file where the error message will be written. By default, error messages are written to stderr.
smsg: The error message; this will usually be ErrMsg(i) where i is an error number.

ErrorLog - error logger

Function prototype:

  void ErrorLog (int, char *, FILE *, char *);

Argument description:

--------------------------------
              Argument | Type   
--------------------------------
void ErrorLog (error,  | int    
               func,   | char * 
               efile,  | FILE * 
               smsg);  | char * 
--------------------------------

error: The error number, as described in the error table.
func: A string specifying the function that is invoking the error call.
efile: Specifies the file where the error message will be written. By default error messages are written to stderr.
smsg: The error message; this will usually be ErrMsg(i) where i is an error number.

ErrMsg - retrieves error messages

Function prototype:

  char *ErrMsg(int i)

Return value:

ErrMsg returns the error message associated with its integer argument.

Argument description:

-----------------------------------
                  Argument | Type  
-----------------------------------
char *ErrMsg (error_num    | int
                       );  | 
-----------------------------------

error_num: An error number.

ErrMax - retrieves the number of possible error messages

Function prototype:

   int ErrMax()

Return value:

The value returned is the total number of possible error messages.

Error table

The following table lists all of the error numbers and the associated error messages:

----------------------------------------------------------------------------
 Error  |                     Error message
 number |    
----------------------------------------------------------------------------
    1   |  Insufficient data in gridded region to triangulate.
    2   |  Duplicate input data coordinates are not allowed.
    3   |  Unable to open file for writing algorithmic data.
    4   |  WARNING:  The ratio of vertical to horizontal scales is too 
        |    large for meaningful gradient estimation.  Rescale the data if 
        |    gradients are required.
    5   |  WARNING:  The ratio of vertical to horizontal scales is too small 
        |    for meaningful gradient estimation.  Rescale the data if gradients 
        |    are required.
    6   |  WARNING:  The ratio of x-axis breadth to y-axis breadth of this 
        |    gridded region may be too extreme for good interpolation.  
        |    Changing the block proportions, or rescaling the x or y 
        |    coordinate may be indicated.  Gradient calculations have been 
        |    disabled.
    7   |  Unable to allocate storage for ivector.
    8   |  Unable to allocate storage for dvector.
    9   |  Unable to allocate storage for **imatrix.
   10   |  Unable to allocate storage for imatrix[].
   11   |  Unable to allocate storage for **fmatrix.
   12   |  Unable to allocate storage for fmatrix[].
   13   |  Unable to allocate storage for **dmatrix.
   14   |  Unable to allocate storage for dmatrix[].
   15   |  Unable to allocate storage for raw data.
   16   |  Unable to allocate storage for a simplex.
   17   |  Unable to allocate storage for temp.
   18   |  Unable to allocate storage for neig.
   19   |  Slopes have not been computed, set sdip.
   20   |  Row argument out of range.
   21   |  Column argument out of range.
   22   |  Aspects have not been computed, set sdip.
   23   |  Parameter name not known.
   24   |  Cannot open error file.
   25   |  Automatic scaling has been done - aspects will be distorted and 
        |    consequently are not returned.  Rescale your data manually, or 
        |    by setting magx, magy, and magz appropriately.
   26   |  Automatic scaling has been done - slopes will be distorted and 
        |    consequently are not returned.  Rescale your data manually, or 
        |    by setting magx, magy, and magz appropriately.
   27   |  Coordinate is outside of the gridded region for a single 
        |    point interpolation.
   28   |  Cannot compute aspects and slopes in conjunction with single 
        |    point interpolation mode.
   29   |  Fortran DOUBLE PRECISION entries are not supported on UNICOS.
   30   |  Error number out of range.
----------------------------------------------------------------------------