Epetra_FECrsMatrix Class Reference

#include <Epetra_FECrsMatrix.h>

Inheritance diagram for Epetra_FECrsMatrix:

[legend]Collaboration diagram for Epetra_FECrsMatrix:[legend]List of all members.

Public Types
enum	{ ROW_MAJOR = 0, COLUMN_MAJOR = 3 }
Public Member Functions
	Epetra_FECrsMatrix (Epetra_DataAccess CV, const Epetra_Map &RowMap, int *NumEntriesPerRow, bool ignoreNonLocalEntries=false)
	Epetra_FECrsMatrix (Epetra_DataAccess CV, const Epetra_Map &RowMap, int NumEntriesPerRow, bool ignoreNonLocalEntries=false)
	Epetra_FECrsMatrix (Epetra_DataAccess CV, const Epetra_Map &RowMap, const Epetra_Map &ColMap, int *NumEntriesPerRow, bool ignoreNonLocalEntries=false)
	Epetra_FECrsMatrix (Epetra_DataAccess CV, const Epetra_Map &RowMap, const Epetra_Map &ColMap, int NumEntriesPerRow, bool ignoreNonLocalEntries=false)
	Epetra_FECrsMatrix (Epetra_DataAccess CV, const Epetra_CrsGraph &Graph, bool ignoreNonLocalEntries=false)
	Epetra_FECrsMatrix (const Epetra_FECrsMatrix &src)
virtual	~Epetra_FECrsMatrix ()
Epetra_FECrsMatrix &	operator= (const Epetra_FECrsMatrix &src)
int	SumIntoGlobalValues (int numIndices, const int *indices, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	SumIntoGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	SumIntoGlobalValues (int numIndices, const int *indices, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	SumIntoGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	InsertGlobalValues (int numIndices, const int *indices, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	InsertGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	InsertGlobalValues (int numIndices, const int *indices, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	InsertGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	ReplaceGlobalValues (int numIndices, const int *indices, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	ReplaceGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	ReplaceGlobalValues (int numIndices, const int *indices, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	ReplaceGlobalValues (int numRows, const int *rows, int numCols, const int *cols, const double *const *values, int format=Epetra_FECrsMatrix::ROW_MAJOR)
int	SumIntoGlobalValues (const Epetra_IntSerialDenseVector &indices, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	SumIntoGlobalValues (const Epetra_IntSerialDenseVector &rows, const Epetra_IntSerialDenseVector &cols, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	InsertGlobalValues (const Epetra_IntSerialDenseVector &indices, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	InsertGlobalValues (const Epetra_IntSerialDenseVector &rows, const Epetra_IntSerialDenseVector &cols, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	ReplaceGlobalValues (const Epetra_IntSerialDenseVector &indices, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	ReplaceGlobalValues (const Epetra_IntSerialDenseVector &rows, const Epetra_IntSerialDenseVector &cols, const Epetra_SerialDenseMatrix &values, int format=Epetra_FECrsMatrix::COLUMN_MAJOR)
int	GlobalAssemble (bool callFillComplete=true)
void	setIgnoreNonLocalEntries (bool flag)

---

Detailed Description

Epetra Finite-Element CrsMatrix. This class provides the ability to input finite-element style sub-matrix data, including sub-matrices with non-local rows (which could correspond to shared finite-element nodes for example). This class inherits Epetra_CrsMatrix, and so all Epetra_CrsMatrix functionality is also available.

It is intended that this class will be used as follows:

• Construct with either a map or graph that describes a (non-overlapping) data distribution.
• Input data, including non-local data, using the methods InsertGlobalValues(), SumIntoGlobalValues() and/or ReplaceGlobalValues().
• Call the method GlobalAssemble(), which gathers all non-local data onto the owning processors as determined by the map provided at construction. Users should note that the GlobalAssemble() method has an optional argument which determines whether GlobalAssemble() in turn calls FillComplete() after the data-exchange has occurred. If not explicitly supplied, this argument defaults to true.

Sub-matrix data, which is assumed to be a rectangular 'table' of coefficients accompanied by 'scatter-indices', can be provided in three forms:

• Fortran-style packed 1-D array.
• C-style double-pointer, or list-of-rows.
• Epetra_SerialDenseMatrix object.

In all cases, a "format" parameter specifies whether the data is laid out in row-major or column-major order (i.e., whether coefficients for a row lie contiguously or whether coefficients for a column lie contiguously). See the documentation for the methods SumIntoGlobalValues() and ReplaceGlobalValues().

Important notes:

1. Since Epetra_FECrsMatrix inherits Epetra_CrsMatrix, the semantics of the Insert/SumInto/Replace methods are the same as they are on Epetra_CrsMatrix, which is:
   • InsertGlobalValues() inserts values into the matrix only if the graph has not yet been finalized (FillComplete() has not yet been called). For non-local values, the call to InsertGlobalValues() may succeed but the GlobalAssemble() method may then fail because the non-local data is not actually inserted in the underlying matrix until GlobalAssemble() is called.
   • SumIntoGlobalValues() and ReplaceGlobalValues() only work for values that already exist in the matrix. In other words, these methods can not be used to put new values into the matrix.

---

Constructor & Destructor Documentation

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  Epetra_DataAccess  CV, const Epetra_Map &  RowMap, int *  NumEntriesPerRow, bool  ignoreNonLocalEntries = false )

Constructor.

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  Epetra_DataAccess  CV, const Epetra_Map &  RowMap, int  NumEntriesPerRow, bool  ignoreNonLocalEntries = false )

Constructor.

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  Epetra_DataAccess  CV, const Epetra_Map &  RowMap, const Epetra_Map &  ColMap, int *  NumEntriesPerRow, bool  ignoreNonLocalEntries = false )

Constructor.

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  Epetra_DataAccess  CV, const Epetra_Map &  RowMap, const Epetra_Map &  ColMap, int  NumEntriesPerRow, bool  ignoreNonLocalEntries = false )

Constructor.

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  Epetra_DataAccess  CV, const Epetra_CrsGraph &  Graph, bool  ignoreNonLocalEntries = false )

Constructor.

Epetra_FECrsMatrix::Epetra_FECrsMatrix (  const Epetra_FECrsMatrix &  src  )

Copy Constructor.

virtual Epetra_FECrsMatrix::~Epetra_FECrsMatrix (   )  [virtual]

Destructor.

---

Member Function Documentation

int Epetra_FECrsMatrix::GlobalAssemble (  bool  callFillComplete = true  )

Gather any overlapping/shared data into the non-overlapping partitioning defined by the Map that was passed to this matrix at construction time. Data imported from other processors is stored on the owning processor with a "sumInto" or accumulate operation. This is a collective method -- every processor must enter it before any will complete it. Parameters: callFillComplete  option argument, defaults to true. Determines whether GlobalAssemble() internally calls the FillComplete() method on this matrix. Returns: error-code 0 if successful, non-zero if some error occurs

int Epetra_FECrsMatrix::InsertGlobalValues (  const Epetra_IntSerialDenseVector &  rows, const Epetra_IntSerialDenseVector &  cols, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Insert a general sub-matrix into the global matrix. For square structurally-symmetric sub-matrices, see the other overloading of this method. Parameters: rows  List of row-indices. rows.Length() must be the same as values.M(). cols  List of column-indices. cols.Length() must be the same as values.N(). values  Sub-matrix of coefficients. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::InsertGlobalValues (  const Epetra_IntSerialDenseVector &  indices, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Insert a square structurally-symmetric sub-matrix into the global matrix. For non-square sub-matrices, see the other overloading of this method. Parameters: indices  List of scatter-indices. indices.Length() must be the same as values.M() and values.N(). values  Sub-matrix of coefficients. Must be square. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::InsertGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Insert a C-style table (double-pointer, or list of lists) of coefficients into the matrix. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List of column-numbers (scatter-indices) for the sub-matrix. values  Rectangular sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::InsertGlobalValues (  int  numIndices, const int *  indices, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Insert a C-style table (double-pointer, or list of lists) of coefficients into the matrix. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  Square sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::InsertGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Insert a Fortran-style table (single-dimensional packed-list) of coefficients into the matrix. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List of column-numbers (scatter-indices) for the sub-matrix. values  List, length numRows*numCols. Rectangular sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

int Epetra_FECrsMatrix::InsertGlobalValues (  int  numIndices, const int *  indices, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Insert a Fortran-style table (single-dimensional packed-list) of coefficients into the matrix. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  List, length numIndices*numIndices. Square sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

Epetra_FECrsMatrix& Epetra_FECrsMatrix::operator= (  const Epetra_FECrsMatrix &  src  )

Assignment operator

int Epetra_FECrsMatrix::ReplaceGlobalValues (  const Epetra_IntSerialDenseVector &  rows, const Epetra_IntSerialDenseVector &  cols, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Use a general sub-matrix into the global matrix to replace existing values. For square structurally-symmetric sub-matrices, see the other overloading of this method. Parameters: rows  List of row-indices. rows.Length() must be the same as values.M(). cols  List of column-indices. cols.Length() must be the same as values.N(). values  Sub-matrix of coefficients. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::ReplaceGlobalValues (  const Epetra_IntSerialDenseVector &  indices, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Use a square structurally-symmetric sub-matrix to replace existing values in the global matrix. For non-square sub-matrices, see the other overloading of this method. Parameters: indices  List of scatter-indices. indices.Length() must be the same as values.M() and values.N(). values  Sub-matrix of coefficients. Must be square. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::ReplaceGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Copy C-style table (double-pointer, or list of lists) of coefficients into the matrix, replacing any coefficients that may already exist at the specified row/column locations. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List of column-numbers (scatter-indices) for the sub-matrix. values  Rectangular sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::ReplaceGlobalValues (  int  numIndices, const int *  indices, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Copy C-style table (double-pointer, or list of lists) of coefficients into the matrix, replacing any coefficients that may already exist at the specified row/column locations. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  Square sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::ReplaceGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Copy Fortran-style table (single-dimensional packed-list) of coefficients into the matrix, replacing any coefficients that may already exist at the specified row/column locations. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List, of column-numbers (scatter-indices) for the sub-matrix. values  List, length numRows*numCols. Rectangular sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

int Epetra_FECrsMatrix::ReplaceGlobalValues (  int  numIndices, const int *  indices, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Copy a Fortran-style table (single-dimensional packed-list) of coefficients into the matrix, replacing any coefficients that may already exist at the specified row/column locations. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  List, length numIndices*numIndices. Square sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

void Epetra_FECrsMatrix::setIgnoreNonLocalEntries (  bool  flag  )  [inline]

Set whether or not non-local data values should be ignored. By default, non-local data values are NOT ignored.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  const Epetra_IntSerialDenseVector &  rows, const Epetra_IntSerialDenseVector &  cols, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Sum a general sub-matrix into the global matrix. For square structurally-symmetric sub-matrices, see the other overloading of this method. Parameters: rows  List of row-indices. rows.Length() must be the same as values.M(). cols  List of column-indices. cols.Length() must be the same as values.N(). values  Sub-matrix of coefficients. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  const Epetra_IntSerialDenseVector &  indices, const Epetra_SerialDenseMatrix &  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Sum a square structurally-symmetric sub-matrix into the global matrix. For non-square sub-matrices, see the other overloading of this method. Parameters: indices  List of scatter-indices. indices.Length() must be the same as values.M() and values.N(). values  Sub-matrix of coefficients. Must be square. format  Optional format specifier, defaults to COLUMN_MAJOR.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Sum C-style table (double-pointer, or list of lists) of coefficients into the matrix, adding them to any coefficients that may already exist at the specified row/column locations. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List of column-numbers (scatter-indices) for the sub-matrix. values  Rectangular sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  int  numIndices, const int *  indices, const double *const *  values, int  format = Epetra_FECrsMatrix::ROW_MAJOR )

Sum C-style table (double-pointer, or list of lists) of coefficients into the matrix, adding them to any coefficients that may already exist at the specified row/column locations. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  Square sub-matrix of coefficients, provided in a 2-D array, or double-pointer. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is ROW_MAJOR.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  int  numRows, const int *  rows, int  numCols, const int *  cols, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Sum a Fortran-style table (single-dimensional packed-list) of coefficients into the matrix, adding them to any coefficients that may already exist at the specified row/column locations. Parameters: numRows  Number of rows in the sub-matrix. rows  List of row-numbers (scatter-indices) for the sub-matrix. numCols  Number of columns in the sub-matrix. cols  List of column-numbers (scatter-indices) for the sub-matrix. values  List, length numRows*numCols. Rectangular sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

int Epetra_FECrsMatrix::SumIntoGlobalValues (  int  numIndices, const int *  indices, const double *  values, int  format = Epetra_FECrsMatrix::COLUMN_MAJOR )

Sum a Fortran-style table (single-dimensional packed-list) of coefficients into the matrix, adding them to any coefficients that may already exist at the specified row/column locations. Parameters: numIndices  Number of rows (and columns) in the sub-matrix. indices  List of scatter-indices (rows and columns) for the sub-matrix. values  List, length numIndices*numIndices. Square sub-matrix of coefficients, packed in a 1-D array. Data is packed either contiguously by row or by column, specified by the final parameter 'format'. format  Specifies whether the data in 'values' is packed in column-major or row-major order. Valid values are Epetra_FECrsMatrix::ROW_MAJOR or Epetra_FECrsMatrix::COLUMN_MAJOR. This is an optional parameter, default value is COLUMN_MAJOR.

---

The documentation for this class was generated from the following file:

• Epetra_FECrsMatrix.h

---