jsat.linear

Class SparseMatrix

java.lang.Object

jsat.linear.Matrix

jsat.linear.SparseMatrix

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable

public class SparseMatrix
extends Matrix

Creates a new Sparse Matrix where each row is backed by a sparse vector.

This implementation does not support the QR or lup() decompositions.
transposeMultiply(jsat.linear.Matrix, jsat.linear.Matrix, java.util.concurrent.ExecutorService) currently does not use multiple cores.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
SparseMatrix(int rows, int cols) Creates a new sparse matrix
SparseMatrix(int rows, int cols, int rowCapacity) Creates a new sparse matrix
SparseMatrix(SparseVector[] rows) Creates a new Sparse Matrix backed by the given array of SpareVectors.

Method Summary

Modifier and Type	Method and Description
void	changeSize(int newRows, int newCols) This method alters the size of a matrix, either adding or subtracting rows from the internal structure of the matrix.
SparseMatrix	clone()
int	cols() Returns the number of columns stored in this matrix
double	get(int i, int j) Returns the value stored at at the matrix position Ai,j
Vec	getRowView(int r) Obtains a vector that is backed by this, at very little memory cost.
void	increment(int i, int j, double value) Alters the current matrix at index (i,j) to be equal to Ai,j = Ai,j + value
boolean	isSparce() Returns true if the matrix is sparse, false otherwise
Matrix[]	lup()
Matrix[]	lup(java.util.concurrent.ExecutorService threadPool)
void	multiply(Matrix B, Matrix C) Alters the matrix C to be equal to C = C+A*B
void	multiply(Matrix B, Matrix C, java.util.concurrent.ExecutorService threadPool) Alters the matrix C to be equal to C = C+A*B
void	multiply(Vec b, double z, Vec c) If this matrix is Am x n, and b has a length of n, and c has a length of m, then this will mutate c to store c = c + A*b*z
void	multiplyTranspose(Matrix B, Matrix C) Alters the matrix C to be equal to C = C+A*BT
void	multiplyTranspose(Matrix B, Matrix C, java.util.concurrent.ExecutorService threadPool) Alters the matrix C to be equal to C = C+A*BT
void	mutableAdd(double c) Alters the current matrix to store the value A+c
void	mutableAdd(double c, java.util.concurrent.ExecutorService threadPool) Alters the current matrix to store the value A+c
void	mutableAdd(double c, Matrix B) Alters the current matrix to store the value A+c*B
void	mutableAdd(double c, Matrix B, java.util.concurrent.ExecutorService threadPool) Alters the current matrix to store the value A+c*B
void	mutableMultiply(double c) Alters the current matrix to be equal to A*c
void	mutableMultiply(double c, java.util.concurrent.ExecutorService threadPool) Alters the current matrix to be equal to A*c
void	mutableTranspose() Transposes the current matrix in place, altering its value.
long	nnz() Returns the number of non zero values stored in this matrix.
Matrix[]	qr()
Matrix[]	qr(java.util.concurrent.ExecutorService threadPool)
int	rows() Returns the number of rows stored in this matrix
void	set(int i, int j, double value) Sets the value stored at at the matrix position Ai,j
void	swapRows(int r1, int r2) Alters the current matrix by swapping the values stored in two different rows.
void	transpose(Matrix C) Overwrites the values stored in matrix C to store the value of A'
void	transposeMultiply(double c, Vec b, Vec x) Alters the vector x to be equal to x = x + A'*b*c
void	transposeMultiply(Matrix B, Matrix C) Alters the matrix C so that C = C + A'*B
void	transposeMultiply(Matrix B, Matrix C, java.util.concurrent.ExecutorService threadPool) Alters the matrix C so that C = C + A'*B
void	zeroOut() Alters the current matrix so that all values are equal to zero.

Methods inherited from class jsat.linear.Matrix

add, add, add, add, canBeMutated, canMultiply, copyTo, diag, diagMult, diagMult, equals, equals, eye, getColumn, getColumnView, getRow, isSquare, isSymmetric, isSymmetric, multiply, multiply, multiply, multiply, multiply, multiplyTranspose, multiplyTranspose, mutableAdd, mutableAdd, mutableSubtract, mutableSubtract, mutableSubtract, mutableSubtract, mutableSubtract, mutableSubtract, OuterProductUpdate, OuterProductUpdate, pascal, random, sameDimensions, subtract, subtract, subtract, subtract, toString, transpose, transposeMultiply, transposeMultiply, transposeMultiply, updateRow

Methods inherited from class java.lang.Object

getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

SparseMatrix

public SparseMatrix(int rows,
                    int cols,
                    int rowCapacity)

Creates a new sparse matrix

Parameters:

rows - the number of rows for the matrix

cols - the number of columns for the matrix

rowCapacity - the initial capacity for non zero values for each row

SparseMatrix

public SparseMatrix(SparseVector[] rows)

Creates a new Sparse Matrix backed by the given array of SpareVectors. Altering the array of any object in it will also alter the this matrix.

Parameters:

rows - the array to back this SparseMatrix

SparseMatrix

public SparseMatrix(int rows,
                    int cols)

Creates a new sparse matrix

Parameters:

rows - the number of rows for the matrix

cols - the number of columns for the matrix

Method Detail

mutableAdd

public void mutableAdd(double c,
                       Matrix B)

Description copied from class: Matrix

Alters the current matrix to store the value A+c*B

Specified by:

mutableAdd in class Matrix

Parameters:

c - the scalar constant to multiple B by

B - the matrix to add to this

mutableAdd

public void mutableAdd(double c,
                       Matrix B,
                       java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the current matrix to store the value A+c*B

Specified by:

mutableAdd in class Matrix

Parameters:

c - the scalar constant to multiple B by

B - the matrix to add to this

threadPool - the source of threads to do computation in parallel

mutableAdd

public void mutableAdd(double c)

Description copied from class: Matrix

Alters the current matrix to store the value A+c

Specified by:

mutableAdd in class Matrix

Parameters:

c - the scalar constant to add to this

mutableAdd

public void mutableAdd(double c,
                       java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the current matrix to store the value A+c

Specified by:

mutableAdd in class Matrix

Parameters:

c - the scalar constant to add to this

threadPool - the source of threads to do computation in parallel

multiply

public void multiply(Vec b,
                     double z,
                     Vec c)

Description copied from class: Matrix

If this matrix is A_{m x n}, and b has a length of n, and c has a length of m, then this will mutate c to store c = c + A*b*z

Specified by:

multiply in class Matrix

Parameters:

b - the vector to be treated as a colum vector

z - the constant to multiply the A*b value by.

c - where to place the result by addition

multiply

public void multiply(Matrix B,
                     Matrix C)

Description copied from class: Matrix

Alters the matrix C to be equal to C = C+A*B

Specified by:

multiply in class Matrix

Parameters:

B - the matrix to multiply this with

C - the matrix to add the result to

multiply

public void multiply(Matrix B,
                     Matrix C,
                     java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the matrix C to be equal to C = C+A*B

Specified by:

multiply in class Matrix

Parameters:

B - the matrix to multiply this with

C - the matrix to add the result to

threadPool - the source of threads to do computation in parallel

mutableMultiply

public void mutableMultiply(double c)

Description copied from class: Matrix

Alters the current matrix to be equal to A*c

Specified by:

mutableMultiply in class Matrix

Parameters:

c - the scalar constant to multiply by

mutableMultiply

public void mutableMultiply(double c,
                            java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the current matrix to be equal to A*c

Specified by:

mutableMultiply in class Matrix

Parameters:

c - the scalar constant to multiply by

threadPool - the source of threads to do computation in parallel

lup

public Matrix[] lup()

Specified by:

lup in class Matrix

lup

public Matrix[] lup(java.util.concurrent.ExecutorService threadPool)

Specified by:

lup in class Matrix

qr

public Matrix[] qr()

Specified by:

qr in class Matrix

qr

public Matrix[] qr(java.util.concurrent.ExecutorService threadPool)

Specified by:

qr in class Matrix

mutableTranspose

public void mutableTranspose()

Description copied from class: Matrix

Transposes the current matrix in place, altering its value. Only valid for square matrices

Specified by:

mutableTranspose in class Matrix

transpose

public void transpose(Matrix C)

Description copied from class: Matrix

Overwrites the values stored in matrix C to store the value of A'

Specified by:

transpose in class Matrix

Parameters:

C - the matrix to store the transpose of the current matrix

transposeMultiply

public void transposeMultiply(Matrix B,
                              Matrix C)

Description copied from class: Matrix

Alters the matrix C so that C = C + A'*B

Specified by:

transposeMultiply in class Matrix

Parameters:

B - the matrix to multiply by

C - the matrix to add the result to

transposeMultiply

public void transposeMultiply(Matrix B,
                              Matrix C,
                              java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the matrix C so that C = C + A'*B

Specified by:

transposeMultiply in class Matrix

Parameters:

B - the matrix to multiply by

C - the matrix to place the results in

threadPool - the source of threads to do computation in parallel

transposeMultiply

public void transposeMultiply(double c,
                              Vec b,
                              Vec x)

Description copied from class: Matrix

Alters the vector x to be equal to x = x + A'*b*c

Specified by:

transposeMultiply in class Matrix

Parameters:

c - the scalar constant to multiply by

b - the vector to multiply by

x - the vector the add the result to

getRowView

public Vec getRowView(int r)

Description copied from class: Matrix

Obtains a vector that is backed by this, at very little memory cost. Mutations to this vector will alter the values stored in the matrix, and vice versa.

Overrides:

getRowView in class Matrix

Parameters:

r - the row to obtain a view of

Returns:

a vector backed by the specified row of the matrix

get

public double get(int i,
                  int j)

Description copied from class: Matrix

Returns the value stored at at the matrix position A_i,j

Specified by:

get in class Matrix

Parameters:

i - the row, starting from 0

j - the column, starting from 0

Returns:

the value at A_i,j

set

public void set(int i,
                int j,
                double value)

Description copied from class: Matrix

Sets the value stored at at the matrix position A_i,j

Specified by:

set in class Matrix

Parameters:

i - the row, starting from 0

j - the column, starting from 0

value - the value to place at A_i,j

increment

public void increment(int i,
                      int j,
                      double value)

Description copied from class: Matrix

Alters the current matrix at index (i,j) to be equal to A_i,j = A_i,j + value

Overrides:

increment in class Matrix

Parameters:

i - the row, starting from 0

j - the column, starting from 0

value - the value to add to the matrix coordinate

rows

public int rows()

Description copied from class: Matrix

Returns the number of rows stored in this matrix

Specified by:

rows in class Matrix

Returns:

the number of rows stored in this matrix

cols

public int cols()

Description copied from class: Matrix

Returns the number of columns stored in this matrix

Specified by:

cols in class Matrix

Returns:

the number of columns stored in this matrix

isSparce

public boolean isSparce()

Description copied from class: Matrix

Returns true if the matrix is sparse, false otherwise

Specified by:

isSparce in class Matrix

Returns:

true if the matrix is sparse, false otherwise

swapRows

public void swapRows(int r1,
                     int r2)

Description copied from class: Matrix

Alters the current matrix by swapping the values stored in two different rows.

Specified by:

swapRows in class Matrix

Parameters:

r1 - the first row to swap

r2 - the second row to swap

zeroOut

public void zeroOut()

Description copied from class: Matrix

Alters the current matrix so that all values are equal to zero.

Specified by:

zeroOut in class Matrix

clone

public SparseMatrix clone()

Specified by:

clone in class Matrix

nnz

public long nnz()

Description copied from class: Matrix

Returns the number of non zero values stored in this matrix. This is mostly useful for sparse matrices.

Overrides:

nnz in class Matrix

Returns:

the number of non zero values stored in this matrix.

changeSize

public void changeSize(int newRows,
                       int newCols)

Description copied from class: Matrix

This method alters the size of a matrix, either adding or subtracting rows from the internal structure of the matrix. Every resize call may cause a new allocation internally, and should not be called for excessive changing of a matrix. All added rows/ columns will have values of zero. If a row / column is removed, it is always the bottom/right most row / column removed. Values of the removed rows / columns will be lost.

Specified by:

changeSize in class Matrix

Parameters:

newRows - the new number of rows, must be positive

newCols - the new number of columns, must be positive.

multiplyTranspose

public void multiplyTranspose(Matrix B,
                              Matrix C)

Description copied from class: Matrix

Alters the matrix C to be equal to C = C+A*B^T

Specified by:

multiplyTranspose in class Matrix

Parameters:

B - the matrix to multiply this with

C - the matrix to add the result to

multiplyTranspose

public void multiplyTranspose(Matrix B,
                              Matrix C,
                              java.util.concurrent.ExecutorService threadPool)

Description copied from class: Matrix

Alters the matrix C to be equal to C = C+A*B^T

Specified by:

multiplyTranspose in class Matrix

Parameters:

B - the matrix to multiply this with

C - the matrix to add the result to

threadPool - the source of threads to do computation in parallel