Jan 20, 2012: version 1.1.4 * bug fix: INT_MAX changed to Int_MAX (used for checking Int overflow). Dec 7, 2011: version 1.1.3 * fixed the Makefile to better align with CFLAGS and other standards * minor fix to KLU mexFunction Jan 25, 2011: version 1.1.2 * minor fix to "make install" * port of Tcov to Mac OSX 10.6.6 Nov 30, 2009: version 1.1.1 * added "make install" and "make uninstall" March 24, 2009, version 1.1.0 * updates to documentation * tabs expanded to spaces in all source code files Nov 1, 2007, version 1.0.1 * minor lint cleanup May 31, 2007, version 1.0 * Overview: this is the first clean release of KLU. Only one bug was fixed since in the last pre-1.0 version (see below). This release adds a 64-bit version, a better Demo, a 100% statement coverage test, new parameters and statistics in the KLU Common object, reduced memory usage, a method for limiting worst-case work in the BTF ordering, and a completely redesigned MATLAB interface. * scaling default changed from no scaling, to max row scaling * C-callable API modified for klu_malloc, klu_free, klu_realloc, klu_rcond, klu_rgrowth, klu_condest. API of other user-callable KLU functions not modified. * user ordering function prototype modified (final argument is now klu_common, not Common->user_data) * User Guide added. * KLU Demo completely rewritten. Now depends on CHOLMOD to read in its matrices, in Matrix Market format. * port to 64-bit version * reduction in memory usage, particularly when the BTF form results in many small diagonal blocks * new Common parameter (maxwork) and statistics (work, memusage, mempeak) * Makefile and object files (*.o) now placed in KLU/Lib, not KLU/Source * added klu_install.m, klu_demo.m, klu_make.m to KLU/MATLAB. * klu mexFunction now returns a struct for LU, not a lengthy list of matrices. MATLAB interface completely rewritten. * Tcov tests completely rewritten * bug fix in complex klu_z_refactor, when both btf and scaling are in use * bug fix in klu_rgrowth, when the matrix is scaled Dec 12, 2006: version 0.11 * minor MATLAB cleanup