Mercurial > octave-libgccjit

comparison doc/interpreter/diagperm.txi @ 18812:9ac2357f19bc

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
doc: Replace "non-zero" with "nonzero" to match existing usage. Replace all occurrences in both documentation and code comments. * doc/interpreter/contrib.txi, doc/interpreter/diagperm.txi, doc/interpreter/external.txi, doc/interpreter/sparse.txi, doc/interpreter/stmt.txi, doc/interpreter/testfun.txi, doc/refcard/refcard.tex, examples/mysparse.c, libinterp/corefcn/balance.cc, libinterp/corefcn/cellfun.cc, libinterp/corefcn/data.cc, libinterp/corefcn/filter.cc, libinterp/corefcn/find.cc, libinterp/corefcn/kron.cc, libinterp/corefcn/ls-mat5.cc, libinterp/corefcn/luinc.cc, libinterp/corefcn/mappers.cc, libinterp/corefcn/oct-fstrm.cc, libinterp/corefcn/oct-fstrm.h, libinterp/corefcn/oct-iostrm.cc, libinterp/corefcn/oct-iostrm.h, libinterp/corefcn/oct-stdstrm.h, libinterp/corefcn/oct-strstrm.h, libinterp/corefcn/spparms.cc, libinterp/corefcn/toplev.cc, libinterp/corefcn/utils.cc, libinterp/dldfcn/symrcm.cc, libinterp/octave-value/ov-bool-mat.cc, liboctave/array/CSparse.cc, liboctave/array/Sparse.cc, liboctave/array/Sparse.h, liboctave/array/dSparse.cc, liboctave/numeric/randmtzig.c, liboctave/operators/Sparse-op-defs.h, scripts/help/get_first_help_sentence.m, scripts/miscellaneous/edit.m, scripts/plot/draw/pie.m, scripts/plot/draw/pie3.m, scripts/sparse/colperm.m, scripts/sparse/nonzeros.m, scripts/sparse/spdiags.m, scripts/sparse/spfun.m, scripts/sparse/spones.m, scripts/sparse/sprand.m, scripts/sparse/sprandn.m, scripts/sparse/sprandsym.m, scripts/sparse/spstats.m, scripts/sparse/svds.m, scripts/special-matrix/gallery.m, scripts/statistics/base/moment.m, scripts/statistics/tests/cor_test.m: Replace "non-zero" with "nonzero" to match existing usage.
author Rik <rik@octave.org>
date Sun, 08 Jun 2014 17:59:59 -0700
parents 200851c87444
children
comparison
equal deleted inserted replaced
18811:c53e11fab75f 18812:9ac2357f19bc
487 487
488 Making diagonal and permutation matrices special matrix objects in their own 488 Making diagonal and permutation matrices special matrix objects in their own
489 right and the consequent usage of smarter algorithms for certain operations 489 right and the consequent usage of smarter algorithms for certain operations
490 implies, as a side effect, small differences in treating zeros. 490 implies, as a side effect, small differences in treating zeros.
491 The contents of this section apply also to sparse matrices, discussed in 491 The contents of this section apply also to sparse matrices, discussed in
492 the following chapter. (@pxref{Sparse Matrices}) 492 the following chapter. (@pxref{Sparse Matrices})
493 493
494 The IEEE floating point standard defines the result of the expressions @code{0*Inf} and 494 The IEEE floating point standard defines the result of the expressions
495 @code{0*NaN} as @code{NaN}. This is widely agreed to be a good 495 @code{0*Inf} and @code{0*NaN} as @code{NaN}. This is widely agreed to be a
496 compromise. 496 good compromise. Numerical software dealing with structured and sparse matrices
497 Numerical software dealing with structured and sparse matrices (including 497 (including Octave) however, almost always makes a distinction between a
498 Octave) however, almost always makes a distinction between a "numerical zero" 498 "numerical zero" and an "assumed zero". A "numerical zero" is a zero value
499 and an "assumed zero". 499 occurring in a place where any floating-point value could occur. It is
500 A "numerical zero" is a zero value occurring in a place where any floating-point 500 normally stored somewhere in memory as an explicit value. An "assumed zero", on
501 value could occur. It is normally stored somewhere in memory as an explicit 501 the contrary, is a zero matrix element implied by the matrix structure
502 value. 502 (diagonal, triangular) or a sparsity pattern; its value is usually not stored
503 An "assumed zero", on the contrary, is a zero matrix element implied by the 503 explicitly anywhere, but is implied by the underlying data structure.
504 matrix structure (diagonal, triangular) or a sparsity pattern; its value is
505 usually not stored explicitly anywhere, but is implied by the underlying
506 data structure.
507 504
508 The primary distinction is that an assumed zero, when multiplied 505 The primary distinction is that an assumed zero, when multiplied
509 by any number, or divided by any nonzero number, 506 by any number, or divided by any nonzero number,
510 yields *always* a zero, even when, e.g., multiplied by @code{Inf} 507 yields *always* a zero, even when, e.g., multiplied by @code{Inf}
511 or divided by @code{NaN}. 508 or divided by @code{NaN}.