Mercurial > octave-nkf

comparison doc/interpreter/sparse.txi @ 19074:0850b5212619

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
doc: Add @nospell macro around proper names in documentation. * aspell-octave.en.pws: Remove proper names from Octave's custom dictionary. * doc/interpreter/audio.txi, doc/interpreter/diagperm.txi, doc/interpreter/diffeq.txi, doc/interpreter/emacs.txi, doc/interpreter/func.txi, doc/interpreter/geometry.txi, doc/interpreter/intro.txi, doc/interpreter/linalg.txi, doc/interpreter/numbers.txi, doc/interpreter/preface.txi, doc/interpreter/quad.txi, doc/interpreter/sparse.txi, libinterp/corefcn/colloc.cc, libinterp/corefcn/ellipj.cc, libinterp/corefcn/gammainc.cc, libinterp/corefcn/hess.cc, libinterp/corefcn/quadcc.cc, libinterp/corefcn/rand.cc, libinterp/corefcn/sqrtm.cc, libinterp/dldfcn/ccolamd.cc, libinterp/dldfcn/colamd.cc, libinterp/dldfcn/dmperm.cc, libinterp/dldfcn/symrcm.cc, scripts/general/quadgk.m, scripts/general/quadl.m, scripts/linear-algebra/commutation_matrix.m, scripts/linear-algebra/condest.m, scripts/linear-algebra/duplication_matrix.m, scripts/linear-algebra/expm.m, scripts/linear-algebra/housh.m, scripts/linear-algebra/krylov.m, scripts/linear-algebra/onenormest.m, scripts/linear-algebra/qzhess.m, scripts/optimization/fminsearch.m, scripts/optimization/glpk.m, scripts/signal/arch_fit.m, scripts/signal/bartlett.m, scripts/signal/blackman.m, scripts/signal/diffpara.m, scripts/signal/durbinlevinson.m, scripts/signal/hamming.m, scripts/signal/hanning.m, scripts/sparse/eigs.m, scripts/sparse/pcg.m, scripts/sparse/pcr.m, scripts/specfun/ellipke.m, scripts/special-matrix/gallery.m, scripts/special-matrix/invhilb.m, scripts/statistics/base/quantile.m, scripts/statistics/tests/kruskal_wallis_test.m, scripts/statistics/tests/manova.m, scripts/statistics/tests/mcnemar_test.m: Add @nospell macro around proper names in documentation.
author Rik <rik@octave.org>
date Thu, 21 Aug 2014 15:57:18 -0700
parents 9ac2357f19bc
children 38937efbee21
comparison
equal deleted inserted replaced
19073:f45b60897316 19074:0850b5212619
73 73
74 There are many different means of storing sparse matrix data. What all 74 There are many different means of storing sparse matrix data. What all
75 of the methods have in common is that they attempt to reduce the complexity 75 of the methods have in common is that they attempt to reduce the complexity
76 and storage given a priori knowledge of the particular class of problems 76 and storage given a priori knowledge of the particular class of problems
77 that will be solved. A good summary of the available techniques for storing 77 that will be solved. A good summary of the available techniques for storing
78 sparse matrix is given by Saad @footnote{Y. Saad "SPARSKIT: A basic toolkit 78 sparse matrix is given by @nospell{Saad} @footnote{Y. Saad "SPARSKIT: A basic
79 for sparse matrix computation", 1994, 79 toolkit for sparse matrix computation", 1994,
80 @url{http://www-users.cs.umn.edu/~saad/software/SPARSKIT/paper.ps}}. 80 @url{http://www-users.cs.umn.edu/~saad/software/SPARSKIT/paper.ps}}.
81 With full matrices, knowledge of the point of an element of the matrix 81 With full matrices, knowledge of the point of an element of the matrix
82 within the matrix is implied by its position in the computers memory. 82 within the matrix is implied by its position in the computers memory.
83 However, this is not the case for sparse matrices, and so the positions 83 However, this is not the case for sparse matrices, and so the positions
84 of the nonzero elements of the matrix must equally be stored. 84 of the nonzero elements of the matrix must equally be stored.
801 The band density is defined as the number of nonzero values in the band 801 The band density is defined as the number of nonzero values in the band
802 divided by the total number of values in the full band. The banded 802 divided by the total number of values in the full band. The banded
803 matrix solvers can be entirely disabled by using @dfn{spparms} to set 803 matrix solvers can be entirely disabled by using @dfn{spparms} to set
804 @code{bandden} to 1 (i.e., @code{spparms ("bandden", 1)}). 804 @code{bandden} to 1 (i.e., @code{spparms ("bandden", 1)}).
805 805
806 The QR@tie{}solver factorizes the problem with a Dulmage-Mendelsohn 806 The QR@tie{}solver factorizes the problem with a @nospell{Dulmage-Mendelsohn}
807 decomposition, to separate the problem into blocks that can be treated 807 decomposition, to separate the problem into blocks that can be treated
808 as over-determined, multiple well determined blocks, and a final 808 as over-determined, multiple well determined blocks, and a final
809 over-determined block. For matrices with blocks of strongly connected 809 over-determined block. For matrices with blocks of strongly connected
810 nodes this is a big win as LU@tie{}decomposition can be used for many 810 nodes this is a big win as LU@tie{}decomposition can be used for many
811 blocks. It also significantly improves the chance of finding a solution 811 blocks. It also significantly improves the chance of finding a solution