dragon.matrix
Class AbstractSparseMatrix

java.lang.Object
  |
  +--dragon.matrix.AbstractMatrix
        |
        +--dragon.matrix.AbstractSparseMatrix

All Implemented Interfaces:

Matrix, java.io.Serializable, SparseMatrix

Direct Known Subclasses:

AbstractFlatSparseMatrix, AbstractGiantSparseMatrix, AbstractSuperSparseMatrix

public abstract class AbstractSparseMatrix

extends AbstractMatrix

implements SparseMatrix, java.io.Serializable

Abstract Sparse matrix implements the interface of sparse matrix which can be called by any data type's sparse matrix such as double and integer

Copyright: Copyright (c) 2005

Company: IST, Drexel University

Version:

1.0

Author:

Davis Zhou

See Also:

Serialized Form

Field Summary

protected boolean	isFinalized
protected boolean	mergeMode
protected boolean	miniMode

Fields inherited from class dragon.matrix.AbstractMatrix

cellDataLength, columnBase, columns, rowBase, rows, transposeMatrix

Constructor Summary

AbstractSparseMatrix(boolean mergeMode, boolean miniMode, int cellDataLength)

Method Summary

double	cosine(int rowA, int rowB) This method treats each row a vector and return the cosine similarity of two rows.
boolean	finalizeData() It is equal to calling finalizeData(true)
boolean	genCooccurrenceMatrix(IntSparseMatrix outputCooccurMatrix) The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row and the j-th row.
boolean	genCooccurrenceMatrix(SparseMatrix matrixY, int minOccurrence, IntSparseMatrix outputCooccurMatrix) The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row in the current matrix and the j-th row in the matrixY.
boolean	genCooccurrenceMatrix(SparseMatrix matrixY, IntSparseMatrix outputCooccurMatrix) The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row in the current matrix and the j-th row in the matrixY.
static boolean	genTranslationMatrix(IntSparseMatrix inputCooccurMatrix, DoubleSparseMatrix outputTransMatrix)
int	getCooccurrenceCount(int rowA, int rowB) If two cells in the same column have non-zero scores, they will be counted as co-occurred.
double	getDouble(int row, int column) Return the score stored in the specified cell.
int	getInt(int row, int column) Return the score stored in the specified cell.
Cell	getNonZeroCellInColumn(int column, int index) This method returns the index-th non-zero cell in the given row.
double	getNonZeroDoubleScoreInColumn(int column, int index) This method returns the double score of the index-th non-zero cell in the given column.
double	getNonZeroDoubleScoreInRow(int row, int index) This method returns the double score of the index-th non-zero cell in the given row.
double[]	getNonZeroDoubleScoresInColumn(int column) If the given column has n non-zero cells, this method returns a n-length double array.
double[]	getNonZeroDoubleScoresInRow(int row) If the given row has n non-zero cells, this method returns a n-length double array.
int	getNonZeroIntScoreInColumn(int column, int index) This method returns the integer score of the index-th non-zero cell in the given column.
int	getNonZeroIntScoreInRow(int row, int index) This method returns the integer score of the index-th non-zero cell in the given row.
int[]	getNonZeroIntScoresInColumn(int column) If the given column has n non-zero cells, this method returns a n-length integer array.
int[]	getNonZeroIntScoresInRow(int row) If the given row has n non-zero cells, this method returns a n-length integer array.
int	getNonZeroNumInColumn(int column) If the transposed matrix is not set yet, this method will automatically generate the transposed matrix.
int	getNonZeroRowInColumn(int column, int index) This method returns the row index of the index-th non-zero cell in the given column.
int[]	getNonZeroRowsInColumn(int column) If the given column has n non-zero cells, this method returns a n-length integer array.
boolean	isFinalized() If this method returns true, one can not add data to the sparse matrix any more.
Matrix	transpose() if one has set the transposed matrix by calling the method setTranspose, this method simply return that transposed matrix.

Methods inherited from class dragon.matrix.AbstractMatrix

columns, getBaseColumn, getBaseRow, getCellDataLength, getTranspose, rows, setTranspose

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface dragon.matrix.SparseMatrix

add, createCell, createCell, createSparseMatrix, finalizeData, flush, getCell, getNonZeroCellInRow, getNonZeroColumnInRow, getNonZeroColumnsInRow, getNonZeroNum, getNonZeroNumInRow

Methods inherited from interface dragon.matrix.Matrix

close, columns, getBaseColumn, getBaseRow, getCellDataLength, getTranspose, rows, setTranspose

Field Detail

isFinalized

protected boolean isFinalized

mergeMode

protected boolean mergeMode

miniMode

protected boolean miniMode

Constructor Detail

AbstractSparseMatrix

public AbstractSparseMatrix(boolean mergeMode,
                            boolean miniMode,
                            int cellDataLength)

Method Detail

isFinalized

public boolean isFinalized()

Description copied from interface: SparseMatrix

If this method returns true, one can not add data to the sparse matrix any more.

Specified by:

isFinalized in interface SparseMatrix

Returns:

true if finalized

finalizeData

public boolean finalizeData()

Description copied from interface: SparseMatrix

It is equal to calling finalizeData(true)

Specified by:

finalizeData in interface SparseMatrix

Returns:

true if finalizing successfully

getNonZeroCellInColumn

public Cell getNonZeroCellInColumn(int column,
                                   int index)

Description copied from interface: SparseMatrix

This method returns the index-th non-zero cell in the given row. The index starts from zero. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroCellInColumn in interface SparseMatrix

Parameters:

column - the index of the column

index - the index-th non-zero cell

Returns:

the cell

getNonZeroNumInColumn

public int getNonZeroNumInColumn(int column)

Description copied from interface: SparseMatrix

If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes for the first call.

Specified by:

getNonZeroNumInColumn in interface SparseMatrix

Parameters:

column - the index of the column

Returns:

the number of non-zero cells in the given column.

getNonZeroRowInColumn

public int getNonZeroRowInColumn(int column,
                                 int index)

Description copied from interface: SparseMatrix

This method returns the row index of the index-th non-zero cell in the given column. The index starts from zero. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroRowInColumn in interface SparseMatrix

Parameters:

column - the index of the column

index - the index-th non-zero cell

Returns:

the row index of the specified cell.

getNonZeroRowsInColumn

public int[] getNonZeroRowsInColumn(int column)

Description copied from interface: SparseMatrix

If the given column has n non-zero cells, this method returns a n-length integer array. Each element of the array stores the row index of the corresponding non-zero cell. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroRowsInColumn in interface SparseMatrix

Parameters:

column - the index of the column

Returns:

the array of non-zero rows in the given column.

getDouble

public double getDouble(int row,
                        int column)

Description copied from interface: Matrix

Return the score stored in the specified cell. If the score is integer-typed, it will be coverted to a double.

Specified by:

getDouble in interface Matrix

Parameters:

row - the row index of the cell

column - the column index of the cell

Returns:

the score stored in the cell

getInt

public int getInt(int row,
                  int column)

Description copied from interface: Matrix

Return the score stored in the specified cell. If the score is double-typed, it will be coverted to an integer.

Specified by:

getInt in interface Matrix

Parameters:

row - the row index of the cell

column - the column index of the cell

Returns:

the score stored in the cell

getNonZeroDoubleScoreInColumn

public double getNonZeroDoubleScoreInColumn(int column,
                                            int index)

Description copied from interface: SparseMatrix

This method returns the double score of the index-th non-zero cell in the given column. The index starts from zero. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroDoubleScoreInColumn in interface SparseMatrix

Parameters:

column - the index of the column

index - the index-th non-zero cell

Returns:

the double score

getNonZeroDoubleScoresInColumn

public double[] getNonZeroDoubleScoresInColumn(int column)

Description copied from interface: SparseMatrix

If the given column has n non-zero cells, this method returns a n-length double array. Each element of the array stores the double score of the corresponding non-zero cell. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroDoubleScoresInColumn in interface SparseMatrix

Parameters:

column - the index of the column

Returns:

the array of non-zero double scores in the given column.

getNonZeroIntScoreInColumn

public int getNonZeroIntScoreInColumn(int column,
                                      int index)

Description copied from interface: SparseMatrix

This method returns the integer score of the index-th non-zero cell in the given column. The index starts from zero. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroIntScoreInColumn in interface SparseMatrix

Parameters:

column - the index of the column

index - the index-th non-zero cell

Returns:

the integer score

getNonZeroIntScoresInColumn

public int[] getNonZeroIntScoresInColumn(int column)

Description copied from interface: SparseMatrix

If the given column has n non-zero cells, this method returns a n-length integer array. Each element of the array stores the integer score of the corresponding non-zero cell. If the transposed matrix is not set yet, this method will automatically generate the transposed matrix. Thus, it may take several minutes.

Specified by:

getNonZeroIntScoresInColumn in interface SparseMatrix

Parameters:

column - the index of the column

Returns:

the array of non-zero integer scores in the given column.

getNonZeroDoubleScoreInRow

public double getNonZeroDoubleScoreInRow(int row,
                                         int index)

Description copied from interface: SparseMatrix

This method returns the double score of the index-th non-zero cell in the given row. The index starts from zero.

Specified by:

getNonZeroDoubleScoreInRow in interface SparseMatrix

Parameters:

row - the index of the row.

index - the index-th non-zero cell

Returns:

the double score

getNonZeroIntScoreInRow

public int getNonZeroIntScoreInRow(int row,
                                   int index)

Description copied from interface: SparseMatrix

This method returns the integer score of the index-th non-zero cell in the given row. The index starts from zero.

Specified by:

getNonZeroIntScoreInRow in interface SparseMatrix

Parameters:

row - the index of the row.

index - the index-th non-zero cell

Returns:

the integer score

getNonZeroDoubleScoresInRow

public double[] getNonZeroDoubleScoresInRow(int row)

Description copied from interface: SparseMatrix

If the given row has n non-zero cells, this method returns a n-length double array. Each element of the array stores the double score of the corresponding non-zero cell.

Specified by:

getNonZeroDoubleScoresInRow in interface SparseMatrix

Parameters:

row - the index of row

Returns:

the array of non-zero double scores in the given row

getNonZeroIntScoresInRow

public int[] getNonZeroIntScoresInRow(int row)

Description copied from interface: SparseMatrix

If the given row has n non-zero cells, this method returns a n-length integer array. Each element of the array stores the integer score of the corresponding non-zero cell.

Specified by:

getNonZeroIntScoresInRow in interface SparseMatrix

Parameters:

row - the index of row

Returns:

the array of non-zero integer scores in the given row

genCooccurrenceMatrix

public boolean genCooccurrenceMatrix(IntSparseMatrix outputCooccurMatrix)

Description copied from interface: SparseMatrix

The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row and the j-th row.

Specified by:

genCooccurrenceMatrix in interface SparseMatrix

Parameters:

outputCooccurMatrix - the output cooccurrence matrix

Returns:

true if successfully

genCooccurrenceMatrix

public boolean genCooccurrenceMatrix(SparseMatrix matrixY,
                                     IntSparseMatrix outputCooccurMatrix)

Description copied from interface: SparseMatrix

The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row in the current matrix and the j-th row in the matrixY. The current matrix and the input matrix should have the same number of columns.

Specified by:

genCooccurrenceMatrix in interface SparseMatrix

Parameters:

matrixY - another sparse matrix

outputCooccurMatrix - the output cooccurrence matrix

Returns:

true if successfully

genCooccurrenceMatrix

public boolean genCooccurrenceMatrix(SparseMatrix matrixY,
                                     int minOccurrence,
                                     IntSparseMatrix outputCooccurMatrix)

Description copied from interface: SparseMatrix

The cell (i,j) in the output cooccurrence matrix stores the cooccurrence count of the i-th row in the current matrix and the j-th row in the matrixY. The current matrix and the input matrix should have the same number of columns. If the cooccurrence count less than the given threshold, it will be set to zero.

Specified by:

genCooccurrenceMatrix in interface SparseMatrix

Parameters:

matrixY - another sparse matrix

minOccurrence - the minimum cooccurrence count

outputCooccurMatrix - the output cooccurrence matrix

Returns:

true if successfully

getCooccurrenceCount

public int getCooccurrenceCount(int rowA,
                                int rowB)

Description copied from interface: Matrix

If two cells in the same column have non-zero scores, they will be counted as co-occurred.

Specified by:

getCooccurrenceCount in interface Matrix

Parameters:

rowA - the index of the first row

rowB - the index of the second row

Returns:

the cooccurrence count

cosine

public double cosine(int rowA,
                     int rowB)

Description copied from interface: Matrix

This method treats each row a vector and return the cosine similarity of two rows.

Specified by:

cosine in interface Matrix

Parameters:

rowA - the row index of the first vector

rowB - the row index of the second vector

Returns:

the cosine similarity of two given rows.

genTranslationMatrix

public static boolean genTranslationMatrix(IntSparseMatrix inputCooccurMatrix,
                                           DoubleSparseMatrix outputTransMatrix)

transpose

public Matrix transpose()

Description copied from interface: Matrix

if one has set the transposed matrix by calling the method setTranspose, this method simply return that transposed matrix. Otherwise, it will generate the transposed matrix online. So this method may take several minutes.

Specified by:

transpose in interface Matrix

Overrides:

transpose in class AbstractMatrix

Returns:

the transposed matrix