edu.wlu.cs.levy.SNARLI
Class SNMatrix

java.lang.Object
  |
  +--edu.wlu.cs.levy.SNARLI.SNMatrix

public class SNMatrix

extends java.lang.Object

This class provides glue code for matrix operations in SNARLI. You do not need to call any of these methods to use SNARLI. This class is provided to show you the methods you need to implement if you want to use a different implementation for matrices.

Constructor Summary

SNMatrix(double[][] x)
Constructs a matrix using a 2D array of double-precision floating-point values.

SNMatrix(int m, int n)
Constructs a matrix of all zeros.

Method Summary

void	add(edu.wlu.cs.levy.SNARLI.SNMatrix a) Adds a matrix to this matrix.
static edu.wlu.cs.levy.SNARLI.SNMatrix	arrayMultiply(edu.wlu.cs.levy.SNARLI.SNMatrix a, edu.wlu.cs.levy.SNARLI.SNMatrix b) Returns a matrix containing the products of the elements of two matrices.
edu.wlu.cs.levy.SNARLI.SNMatrix	copy() Returns a deep copy of this matrix.
void	divide(double s) Divides this matrix by a scalar.
static edu.wlu.cs.levy.SNARLI.SNMatrix	divide(edu.wlu.cs.levy.SNARLI.SNMatrix a, double s) Returns the quotient of a matrix and a scalar.
static edu.wlu.cs.levy.SNARLI.SNMatrix	gaussianNoise(int m, int n, java.util.Random r) Returns a matrix of normally distributed random values.
int	getCols() Returns the number of columns of this Matrix.
edu.wlu.cs.levy.SNARLI.SNVector	getRow(int i) Gets the row vector at a certain index of this matrix.
int	getRows() Returns the number of rows of this Matrix.
double[][]	getValues() Returns a 2D array of the values contained in this matrix.
double	max() Returns the value of the largest element of this matrix.
edu.wlu.cs.levy.SNARLI.SNVector	meanRows() Returns a vector repsenting the mean over the rows of this Matrix.
static edu.wlu.cs.levy.SNARLI.SNMatrix	multiply(edu.wlu.cs.levy.SNARLI.SNMatrix a, double s) Returns the product of a matrix and a scalar.
edu.wlu.cs.levy.SNARLI.SNVector	multiply(edu.wlu.cs.levy.SNARLI.SNVector x) Returns the product of this matrix and a vector.
edu.wlu.cs.levy.SNARLI.SNMatrix	ne(double s) Returns a matrix containing ones where this matrix's elements are not equal to a scalar, and zeros elsewhere.
void	setRow(int i, edu.wlu.cs.levy.SNARLI.SNVector x) Sets the row vector at a certain index of this matrix.
static edu.wlu.cs.levy.SNARLI.SNMatrix	subtract(edu.wlu.cs.levy.SNARLI.SNMatrix a, edu.wlu.cs.levy.SNARLI.SNMatrix b) Returns the difference between two matrices.
double	sum() Returns the sum over all elements of this Matrix.
edu.wlu.cs.levy.SNARLI.SNVector	sumRows() Returns a vector representing the sum over the rows of this Matrix.
java.lang.String	toString() Returns a String representation of this matrix.
edu.wlu.cs.levy.SNARLI.SNMatrix	transpose() Returns the transpose of this matrix.

Methods inherited from class java.lang.Object

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

Constructor Detail

SNMatrix

public SNMatrix(int m,
                int n)

Constructs a matrix of all zeros.

Parameters:

m - number of rows

n - number of columns

SNMatrix

public SNMatrix(double[][] x)

Constructs a matrix using a 2D array of double-precision floating-point values.

Method Detail

meanRows

public edu.wlu.cs.levy.SNARLI.SNVector meanRows()

Returns a vector repsenting the mean over the rows of this Matrix.

Returns:

the mean

sumRows

public edu.wlu.cs.levy.SNARLI.SNVector sumRows()

Returns a vector representing the sum over the rows of this Matrix.

Returns:

the sum

subtract

public static edu.wlu.cs.levy.SNARLI.SNMatrix subtract(edu.wlu.cs.levy.SNARLI.SNMatrix a,
                                                       edu.wlu.cs.levy.SNARLI.SNMatrix b)

Returns the difference between two matrices.

Parameters:

a - one matrix

b - the other matrix

Returns:

a - b

Throws:

java.lang.RuntimeException - if the matrices have unequal sizes

arrayMultiply

public static edu.wlu.cs.levy.SNARLI.SNMatrix arrayMultiply(edu.wlu.cs.levy.SNARLI.SNMatrix a,
                                                            edu.wlu.cs.levy.SNARLI.SNMatrix b)

Returns a matrix containing the products of the elements of two matrices.

Parameters:

a - one matrix

b - the other matrix

Returns:

a .* b

Throws:

java.lang.RuntimeException - if the matrices have unequal sizes

getRows

public int getRows()

Returns the number of rows of this Matrix.

Returns:

number of rows

getCols

public int getCols()

Returns the number of columns of this Matrix.

Returns:

number of columns

getValues

public double[][] getValues()

Returns a 2D array of the values contained in this matrix.

Returns:

values

setRow

public void setRow(int i,
                   edu.wlu.cs.levy.SNARLI.SNVector x)
            throws java.lang.RuntimeException

Sets the row vector at a certain index of this matrix. First row has index 0.

Parameters:

i - index

x - vector

Throws:

java.lang.RuntimeException - if index is out of range

getRow

public edu.wlu.cs.levy.SNARLI.SNVector getRow(int i)

Gets the row vector at a certain index of this matrix. First row has index 0.

Parameters:

i - index

Returns:

vector at row i

Throws:

java.lang.RuntimeException - if index is out of range

gaussianNoise

public static edu.wlu.cs.levy.SNARLI.SNMatrix gaussianNoise(int m,
                                                            int n,
                                                            java.util.Random r)

Returns a matrix of normally distributed random values. Values are taken from a Gaussian distribution with zero mean and standard deviation one.

Parameters:

m - number of rows

n - number of columns

Returns:

matrix of normally distributed random values

add

public void add(edu.wlu.cs.levy.SNARLI.SNMatrix a)
         throws java.lang.RuntimeException

Adds a matrix to this matrix.

Parameters:

a - the other matrix

Throws:

java.lang.RuntimeException - if the matrices have unequal sizes

divide

public static edu.wlu.cs.levy.SNARLI.SNMatrix divide(edu.wlu.cs.levy.SNARLI.SNMatrix a,
                                                     double s)

Returns the quotient of a matrix and a scalar.

Parameters:

a - the matrix

s - the scalar

Returns:

a / s

divide

public void divide(double s)

Divides this matrix by a scalar.

Parameters:

s - the scalar

transpose

public edu.wlu.cs.levy.SNARLI.SNMatrix transpose()

Returns the transpose of this matrix.

Returns:

this^T

multiply

public static edu.wlu.cs.levy.SNARLI.SNMatrix multiply(edu.wlu.cs.levy.SNARLI.SNMatrix a,
                                                       double s)

Returns the product of a matrix and a scalar.

Parameters:

a - the matrix

s - the scalar

Returns:

a * s

multiply

public edu.wlu.cs.levy.SNARLI.SNVector multiply(edu.wlu.cs.levy.SNARLI.SNVector x)

Returns the product of this matrix and a vector.

Parameters:

x - the vector

Returns:

this * x

Throws:

java.lang.RuntimeException - if number of columns of this matrix does not equal number of elements in x

sum

public double sum()

Returns the sum over all elements of this Matrix.

Returns:

the sum

max

public double max()

Returns the value of the largest element of this matrix.

Returns:

maximum value

copy

public edu.wlu.cs.levy.SNARLI.SNMatrix copy()

Returns a deep copy of this matrix.

Returns:

matrix copy

ne

public edu.wlu.cs.levy.SNARLI.SNMatrix ne(double s)

Returns a matrix containing ones where this matrix's elements are not equal to a scalar, and zeros elsewhere.

Parameters:

s - the scalar

Returns:

matrix of ones and zeros

toString

public java.lang.String toString()

Returns a String representation of this matrix.

Overrides:

toString in class java.lang.Object

Returns:

string representation