SparseMatrices

A set of functions to deal with sparse matrices. For first instance, the least squares method over sparse matrixes. (University team project)

API documentation

void QuickSort

Applies the quicksort algorithm to the vector and gives the indices modified.

Input

• double *vec: the vector itself
• int first, int last: indices which limits the vector that it is going to sort (it can sort a subvector)

Output

• double *vec: the sorted vector (by reference)
• int *indices: the indices of the sorted vector (by reference)

double *gaussSeidel

Solves an Ax = b system where A is a sparse matrix and b a is a vector.

Input

• matrizSparse matriz: a matrix of type matrizSparse
• double *b: the vector b of the system

Output

• double x: the solution vector

Type matrizSparse

This is how sparse matrices are stored

• int nfil: number of rows
• int ncol: number of colums
• int nza: number of non-zero values
• int *ifil: row values
• int *icol: colum values
• double *xval: values of the matrix

It is designed in such way that if A is the sparse matrix, for each i in [0, nza) we have A in ifil[i], icol[i] is xval[i]

void matrizPorMatriz

Computes A*B, where A and B are sparse matrices

Input

• const matrizSparse matrizA
• const matrizSparse matrizB

Output

• matrizSparse *matrizAB: result of the computation matrizA*matrizB (passed by reference)

double *matrizPorVector

Computes A*b where A is a sparse matrix and b is a vector.

Input

• const matrizSparse *A
• double *b

Output

• double *resultado: the result vector of the computation

matrizSparse matrizTraspuesta

Trasposes a sparse matrix

Input

• const matrizSparse matriz

Output

• matrizSparse matrizSparseT: matriz^t