Mercurial > octave

comparison NEWS @ 14188:053404893d02 stable

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
update NEWS
author John W. Eaton <jwe@octave.org>
date Tue, 10 Jan 2012 16:04:55 -0500
parents 3f2a3076b9a6
children 14d602c4f6e8 e2a14d1b4eaa 73f02881260f
comparison
equal deleted inserted replaced
14187:53a9c916ba2f 14188:053404893d02
1 Summary of important user-visible changes for version 3.6: 1 Summary of important user-visible changes for version 3.6:
2 --------------------------------------------------------- 2 ---------------------------------------------------------
3 3
4 ** The PCRE library is now required to build Octave. 4 ** The PCRE library is now required to build Octave. If a pre-compiled
5 package does not exist for your system, you can find PCRE sources
6 at http://www.pcre.org
5 7
6 ** The ARPACK library is no longer distributed with Octave. 8 ** The ARPACK library is no longer distributed with Octave.
7 If you need the eigs() or svds() functions you must provide an external 9 If you need the eigs or svds functions you must provide an
8 ARPACK through a package manager or by compiling it yourself. 10 external ARPACK through a package manager or by compiling it
11 yourself. If a pre-compiled package does not exist for your system,
12 you can find the current ARPACK sources at
13 http://forge.scilab.org/index.php/p/arpack-ng
9 14
10 ** Many of Octave's binary operators (.*, .^, +, -, ...) now perform 15 ** Many of Octave's binary operators (.*, .^, +, -, ...) now perform
11 automatic broadcasting for array operations that allows you to use 16 automatic broadcasting for array operations that allows you to use
12 operator notation instead of calling bsxfun or expanding arrays (and 17 operator notation instead of calling bsxfun or expanding arrays (and
13 unnecessarily wasting memory) with repmat or similar idioms. For 18 unnecessarily wasting memory) with repmat or similar idioms. For
58 63
59 Functions now return "single" outputs for inputs of class "single". 64 Functions now return "single" outputs for inputs of class "single".
60 65
61 75% reduction in memory usage through use of logical indexing. 66 75% reduction in memory usage through use of logical indexing.
62 67
63 Random sample functions now use the same syntax as rand() and accept 68 Random sample functions now use the same syntax as rand and accept
64 a comma separated list of dimensions or a dimension vector. 69 a comma separated list of dimensions or a dimension vector.
65 70
66 Functions have been made Matlab-compatible with regard to special 71 Functions have been made Matlab-compatible with regard to special
67 cases (probability on boundaries, probabilities for values outside 72 cases (probability on boundaries, probabilities for values outside
68 distribution, etc.). This may cause subtle changes to existing 73 distribution, etc.). This may cause subtle changes to existing
69 scripts. 74 scripts.
70 75
71 negative binomial function has been extended to real, non-integer inputs. 76 negative binomial function has been extended to real, non-integer
72 discrete_inv() now returns v(1) for 0 instead of NaN. 77 inputs. The discrete_inv fucntion now returns v(1) for 0 instead of
73 nbincdf() recoded to use closed form solution with betainc(). 78 NaN. The nbincdf fucntion has bbeen recoded to use a closed form
79 solution with betainc.
74 80
75 ** strread, textscan, and textread have been completely revamped. 81 ** strread, textscan, and textread have been completely revamped.
76 82
77 They now support nearly all Matlab functionality including: 83 They now support nearly all Matlab functionality including:
78 84
401 ** The effect of comparison operators (<, >, <=, and >=) applied to 407 ** The effect of comparison operators (<, >, <=, and >=) applied to
402 complex numbers has changed to be consistent with the strict 408 complex numbers has changed to be consistent with the strict
403 ordering defined by the `max', `min', and `sort' functions. More 409 ordering defined by the `max', `min', and `sort' functions. More
404 specifically, complex numbers are compared by lexicographical 410 specifically, complex numbers are compared by lexicographical
405 comparison of the pairs `[abs(z), arg(z)]'. Previously, only real 411 comparison of the pairs `[abs(z), arg(z)]'. Previously, only real
406 parts were compared; this can be trivially achieved by wrapping the 412 parts were compared; this can be trivially achieved by converting
407 operands in real(). 413 the operands to real values with the `real' function.
408 414
409 ** The automatic simplification of complex computation results has 415 ** The automatic simplification of complex computation results has
410 changed. Octave will now simplify any complex number with a zero 416 changed. Octave will now simplify any complex number with a zero
411 imaginary part or any complex matrix with all elements having zero 417 imaginary part or any complex matrix with all elements having zero
412 imaginary part to a real value. Previously, this was done only for 418 imaginary part to a real value. Previously, this was done only for
531 matrices, especially larger collections, is now significantly more 537 matrices, especially larger collections, is now significantly more
532 efficient. This applies to both the [] operator and the 538 efficient. This applies to both the [] operator and the
533 cat/vertcat/horzcat functions. 539 cat/vertcat/horzcat functions.
534 540
535 ** It is now possible to optionally employ the xGESDD LAPACK drivers 541 ** It is now possible to optionally employ the xGESDD LAPACK drivers
536 for computing the singular value decomposition using svd(), instead 542 for computing the singular value decomposition using svd, instead
537 of the default xGESVD, using the configuration pseudo-variable 543 of the default xGESVD, using the configuration pseudo-variable
538 svd_driver. The xGESDD driver can be up to 6x times faster when 544 svd_driver. The xGESDD driver can be up to 6x times faster when
539 singular vectors are requested, but is reported to be somewhat less 545 singular vectors are requested, but is reported to be somewhat less
540 robust on highly ill-conditioned matrices. 546 robust on highly ill-conditioned matrices.
541 547
716 y = sin (x); 722 y = sin (x);
717 plot (x, y, "ydatasource", "y"); 723 plot (x, y, "ydatasource", "y");
718 for i = 1 : 100 724 for i = 1 : 100
719 pause(0.1) 725 pause(0.1)
720 y = sin (x + 0.1 * i); 726 y = sin (x + 0.1 * i);
721 refreshdata(); 727 refreshdata ();
722 endfor 728 endfor
723 729
724 works as expected. This capability has be used to introduce 730 works as expected. This capability has be used to introduce
725 stem-series, bar-series, etc., objects for better Matlab 731 stem-series, bar-series, etc., objects for better Matlab
726 compatibility. 732 compatibility.