SparseMatrix

Sparse Matrix Documentation

Description

A matrix representation using an Array of Arrays representation, where each non-zero is represented using a Value-ColumnIndex pair

Methods

Method	Description
Static Methods
static SparseMatrix identity(size)	Creates a sparse identity matrix with the size.
static SparseMatrix zero(size)	Creates a zero matrix with the size.
Data Manipulation
void reserve_per_row(columns)	Pre-allocates/Reserves a number of columns per row. This should be used before inserting elements to reduce re-allocating repeatedly.
void set_dimensions(rows, columns)	Truncates or expands the matrix to fit the dimensions.
Vector2i get_dimensions()	Returns a Vector2i, with x = row_number, y = column_number of the matrix.
void set_element(row, column, value)	Inserts a non-zero element into the matrix. This is most efficient inserting left-to-right, as appending to the end of a vector does not require array rotations to make space.
double get_element(row, column)	Returns the element if it is non-zero. Otherwise returns zero.
SparseMatrix transposed()	Returns the transposed version of the matrix.
SparseMatrix clone()	Returns an identical copy of the matrix.
Array[SparseMatrixCoordinate] to_coordinate_array()	Creates an array of tuples (row, column, value), representing the non-zeros of the matrix.
DenseMatrix to_dense()	Returns the matrix as a DenseMatrix, with all zeros and non-zeros being populated.
Algebra Methods
SparseMatrix multiply_sparse(other)	Multiplies itself with another SparseMatrix, with the other on the RHS, returning a SparseMatrix.
DenseMatrix multiply_sparse_to_dense(other)	Multiplies itself with another SparseMatrix, but the return value is in the DenseMatrix format.
DenseMatrix multiply_dense(other)	Multiplies by a DenseMatrix, returning a DenseMatrix.
VectorN multiply_vector(vector)	Multiplies by a vector, returning a vector. The vector is treated like a column vector multiplied on the RHS.
SparseMatrix add_sparse(other)	Adds the other matrix and itself, returning another Sparse Matrix.
SparseMatrix subtract_sparse(other)	Subtracts the other matrix from it, returning a Sparse Matrix.
DenseMatrix add_dense(other)	Adds it to a DenseMatrix, and returns a DenseMatrix.
DenseMatrix subtract_dense(other)	Subtracts the DenseMatrix from it, returning a DenseMatrix
void multiply_scalar_in_place(scalar)	Multiplies the matrix by the scalar, in-place, without creating a copy. If a copy is desired, call matrix.clone(), and then multiply that copy.
VectorN solve_iterative_cg(B, initial_guess, max_iterations)	This performs an iterative solve using the Conjugate Gradient Method, starting from the initial_guess. This method is guaranteed to converge on Positive-Definite matrices, but may converge otherwise as well. If initial_guess is null, a zero vector is chosen by default. If max_iterations is negative, then it will be ignored.
double norm_squared()	Returns the Frobenius norm squared (equal to summing all values squared).
double norm()	Returns the Frobenius norm (equal to the square root of summing all values squared).