Cholesky Decomposition (sksparse.cholmod)

Added in version 0.1.0.

Changed in version 0.5.0: Major API updates to more closely resemble the scipy.linalg.cholesky() dense interface, and incorporate more functions from the CHOLMOD MATLAB interface.

An interface to the SuiteSparse CHOLMOD package, which computes basic linear algebra operations for sparse, symmetric, positive-definite matrices.

Function Interface

cholesky

Compute the Cholesky factorization of a sparse matrix.

ldl

Compute the LDL factorization of a sparse matrix.

Object Interface

cho_factor

Compute the Cholesky factorization of a sparse matrix.

ldl_factor

Compute the LDL factorization of a sparse matrix.

CholeskyFactor

The main object used for creating and manipulating a Cholesky factor.

Symbolic Analysis

symbfact

Symbolic factorization of a sparse matrix for Cholesky or LDL.

etree

Symbolic factorization of a sparse matrix for Cholesky or LDL.

Graph Partitioning

bisect

Compute a node separator for a sparse matrix graph.

metis

Nested dissection ordering of a sparse matrix using METIS.

nesdis

Nested dissection ordering of a sparse matrix.

SeparatorTree

The separator tree of a sparse matrix graph.

Exceptions and Warnings

CholmodWarning

Base class for CHOLMOD-related warnings.

CholmodSmallDiagonalWarning

Warning for small diagonal entries.

CholmodError

Base class for CHOLMOD-related errors.

CholmodNotPositiveDefiniteError

Raised when the input matrix is not positive definite.

CholmodNotInstalledError

Raised when the CHOLMOD library is not installed.

CholmodOutOfMemoryError

Raised when CHOLMOD runs out of memory.

CholmodOverflowError

Raised when CHOLMOD encounters an integer overflow.

CholmodInvalidInputError

Raised when CHOLMOD receives invalid input.

CholmodGpuProblemError

Raised when CHOLMOD encounters a problem with CUDA.

References