MatSetValues#

Inserts or adds a block of values into a matrix. These values may be cached, so MatAssemblyBegin() and MatAssemblyEnd() MUST be called after all calls to MatSetValues() have been completed.

Synopsis#

#include "petscmat.h" 
PetscErrorCode MatSetValues(Mat mat, PetscInt m, const PetscInt idxm[], PetscInt n, const PetscInt idxn[], const PetscScalar v[], InsertMode addv)

Not Collective

Input Parameters#

mat - the matrix
v - a logically two-dimensional array of values
m - the number of rows
idxm - the global indices of the rows
n - the number of columns
idxn - the global indices of the columns
addv - either ADD_VALUES to add values to any existing entries, or INSERT_VALUES to replace existing entries with new values

Notes#

By default the values, v, are stored row-oriented. See MatSetOption() for other options.

Calls to MatSetValues() with the INSERT_VALUES and ADD_VALUES options cannot be mixed without intervening calls to the assembly routines.

MatSetValues() uses 0-based row and column numbers in Fortran as well as in C.

Negative indices may be passed in idxm and idxn, these rows and columns are simply ignored. This allows easily inserting element stiffness matrices with homogeneous Dirchlet boundary conditions that you don’t want represented in the matrix.

Efficiency Alert#

The routine MatSetValuesBlocked() may offer much better efficiency for users of block sparse formats (MATSEQBAIJ and MATMPIBAIJ).

Developer Note#

This is labeled with C so does not automatically generate Fortran stubs and interfaces because it requires multiple Fortran interfaces depending on which arguments are scalar or arrays.