Profile cover photo
Profile photo
GNU Octave
high-level language for numerical computations
high-level language for numerical computations

Communities and Collections

A new version of the image package has been released, version 2.10.0,
with work mainly from Avinoam Kalma and Hartmut Gimpel. The most
user-visible changes are:

** The following functions are new:

axes2pix colorangle houghpeaks
bwpack deconvwnr otsuthresh
bwunpack houghlines

** The functions entropyfilt, rangefilt, and stdfilt were completely
incorrect and have now been fixed.

** In function regionprops, "Perimeter" was changed to be compatible
with Matlab, and "PerimterOld" was added.

** In function regionprops, properties "ConvexHull", "ConvexImage",
"ConvexArea", and "Solidity" were added.

** In function bwfill, the parameter connectivity was changed. In
previous versions, the parameter referred to background
connectivity but now refers to the foreground connectivity. This
backwards incompatible changes was done for Matlab compatibility.
Effectively, to obtain the same results as in previous versions
use a value of 8 instead of 4, and a value of 4 instead of 8 (the
default was 8).

** grayslice now expects argument V to be a vector with the actual
values used to perform threshold. Previously it expected values
in the [0 1] range which would be adjust to the correct type
internally. This change was done for Matlab compatibility (which
does the same despite their documentation stating otherwise).
Other small changes were done to be Matlab compatible in multiple
corner cases.

** The function operation "thin" in bwmorph was changed to be Matlab
compatible. The previous implementation, based on the book by
W. K. Pratt, is available under the name "thin-pratt".

** imcast now also converts images to logical type.

** wiener2 is now able to handle images with arbitrary number of

** Other functions that have been changed for smaller bugfixes,
increased Matlab compatibility, or performance:

bwdist entropy hough_line
bwmorph graythresh

Add a comment...

Post has attachment
Hi everyone,

A new patch release of the image package is out, version 2.8.1.

The image package provides a wide range of functions for image processing, from measurement of region properties,

to segmentation and geometric transformations [1].

This release fixes a regression on the imcrop function

for non-square images. Kudos to Carnë Draug.

See also the NEWS file [2]


Add a comment...

Post has attachment

A new release of the io package has been made, io-2.4.12.
It's a bug fix release, quite a few bugs have been squashed and code
improvements made with help from various contributors.



Add a comment...

Post has attachment

A new release of the arduino toolkit has been made. [1]

The package is a Basic Octave implementation of the matlab arduino
extension, allowing communication to a programmed arduino board through

This release includes:

* bug fixes for SPI, I2C and sendCommand

* updated function documentation and manual additions

* increased API compatibility with MATLAB

* implemented rotationalEncoder class

* added additional board detection for Arduino Pro/Pro Mini, Arduino
Nano and Lilypad

See the package news for all changes [2]



Add a comment...

Hello everybody,

a new release of the queueing package is available (queueing-1.2.6
[1]). This version fixes some typos in the documentation and includes a
couple of new functions. See [2] for a brief summary of user-visible


Add a comment...

Post has attachment
Hi folks,

A new release of the Symbolic package [1] is out, version 2.7.1.

This is minor release mostly for supporting SymPy-1.3; a summary of
user-visible changes is available online [2].


Add a comment...

Post has attachment

A first public release of the arduino toolkit has been made. [1]

The package is a Basic Octave implementation of the matlab arduino
extension, allowing communication to a programmed arduino board through


John Donoghue
Add a comment...

*GNU Octave 4.4.1 Released*

John W. Eaton via

The Octave developers are pleased to announce the release of GNU Octave,
version 4.4.1.

Version 4.4.1 is a bug fixing release. The bugs fixed in this release
are listed at the end of this message.

We hope you find Octave to be useful. Please help us make it even
better for the future! See

for ways you can become involved in Octave's development and

for ways to donate to the project.

The source code for Octave 4.4.1 is available for download at:

rw-r--r- 1 3003 3002 18965694 Aug 11 19:27 octave-4.4.1.tar.lz
rw-r--r- 1 3003 3002 21633768 Aug 11 19:28 octave-4.4.1.tar.xz
rw-r--r- 1 3003 3002 28626462 Aug 11 19:27 octave-4.4.1.tar.gz


e0d9556caafd83e4276af49b06c5c994 octave-4.4.1.tar.lz
3fd83526af469ee9ce6f7a21b57661c8 octave-4.4.1.tar.xz
b43bd5f4309a0c048c91af10cf8e8674 octave-4.4.1.tar.gz

Please see for mirror sites around the
world. Or you may use, which will
redirect automatically to a nearby mirror.

Binaries for Windows systems are available at

Links to binary (executable) versions for other systems will be listed
at as they become available.

As always, many people contributed to this Octave release. A complete
list of contributors may be found in the Octave manual.

About Octave:

GNU Octave is a high-level interpreted language, primarily intended for
numerical computations. It provides capabilities for the numerical
solution of linear and nonlinear problems and for performing other
numerical experiments. It also provides extensive graphics capabilities
for data visualization and manipulation. Octave is normally used
through its interactive command line interface, but it can also be used
to write non-interactive programs. The Octave language is quite similar
to Matlab so that most programs are easily portable. A full description
of Octave capabilities is available at

Summary of bugs fixed for version 4.4.1:


* disable compound binary operator optimization ([bug

* Fix initial bracketing for fzero ([bug

* Fix scrolling command window in Windows 10 ([bug

* Produce a single backtrace for global declaration overriding local

* revive 'setstr' ([bug #54417](

* 'ode15i.m': In tests, use unlikely symbol name instead of "foo".

* 'run.m': In test, use a symbol name that is unlikely to be a function.

* avoid alignment warning on solaris systems ([bug

* Improve error message if iconv is not available ([bug

* avoid error in gnulib wrapper on systems that don't have fpu
control word

* avoid unused variable warning on some systems

* avoid fprintf format warning on some systems

* eliminate unnecessary 'sun-utils.h' header file

* 'index.tst': Make tests case insensitive for 'NaN', 'NA', 'Inf'
values ([bug #54385](

* positive value from 'uname' can mean success ([bug

* add some bug numbers to the test suite

* consider any non-zero return from 'uname' to be failure ([bug

* if freetype detection fails, unset 'FT2_*' config variables ([bug

* Produce same random number stream for '"state"' initialized to
'Inf' ([bug #54342](

* avoid crash when loading classdef objects ([bug

* 'hist.m': Don't error out if called with identical 'Y' values ([bug

* Parse empty struct input correctly ([bug

* Fix regression in 4.2.0 which disallowed printing with some
Ghostscript devices.

* transform 'libiconv.dll.a' to '-liconv' for 'mkoctfile' ([bug

* 'trapz.m': Overhaul function to fix errors with different
orientations of X and Y ([bug

* Write audio OGG files in smaller chunk sizes ([bug

* Use LAPACK 'DTGSEN' subrouting for ordered-qz computation ([bug

* Fix 'blkmm' to work with empty matrices ([bug

* use octave bat file in pkg install configure/make when in windows
([bug #53938](

* 'repelem.m': Fix behavior when final repeat count is 0 ([bug

* '.hgsub': Use instead of

* move some fortran and dynamic linking configure checks to separate

* '': Reorder tests so that FPICFLAG is correctly set on
Solaris ([bug #54235](

* 'eigs.m': Decrease opts.maxit in UnconvergedEigenvalues BIST ([bug

* fix ambiguous overload build error on Solaris ([bug

* 'pkg.m': Restore ability to use file glob patterns when specifying
local pkg files ([bug #54224](

* Use '%ld' format code in 'xerbla printf' ([bug

* use 'int' instead of 'size_t' for 'xerbla' string length ([bug

* Pair 'fclose' with all 'fopen' and prevent '' test fails
([bug #54180](

* 'fsolve.m': Return 'info=1' when initial guess 'x0' is correct
([bug #53991](

* 'fsolve.m': Add BIST test for [bug

* 'fsolve.m': exit with 'info=-2' when singularity reached ([bug

* ensure that Make sees Octave exit status when running make check

* ' (intialize_xerbla_error_handler, xerbla_abort)':
New functions.

* improve warnings for global/local variable conflicts ([bug

* maybe preserve local value when declaring variable global ([bug

* fix regression in 'text' ([bug

* create valid tmp scope for 'who -file' ([bug

* 'polar.m': Correct axes limits when using negative rho values ([bug

* 'text.m': Fix regression when 'String' property is specified ([bug

* allow 'NULL mxArray*' to be returned as undefined from fcn calls
([bug #54096](

* treat 'NULL' as '[]' for 'mxSetFieldByNumber' and 'mxSetCell' ([bug

* Properly test Qt offscreen printing during manual figures generation.

* Avoid segfault with textscan ([bug

* add Qt include needed to build against Qt 5.11 ([bug

* make listing local functions for command completion work again
([bug #53925](

* Stop segfault in debug mode with readline complete ([bug

* Remove code to install default colors for qt toolkit from window
theme ([bug #53805](

* back out changes attempting to automatically call handle class

* 'residue.m': Use a pre-conditioner on final matrix division for
numerical stability ([bug #53869](

* track multiple file types '(oct|m|mex)' for private functions ([bug

* also preserve temporaries in indexing expressions ([bug

* also update load-path info if private subdir time stamp changes
([bug #53856](

* defer deletion of temporaries in argument lists ([bug

* improve test for handle class destructor

* call delete method correctly for handle class objects ([bug

* use 'size_t' for f77 char arg length params with GCC 8 ([bug

* Fix 'lrwork' dimension in svd for complex cases ([bug

* 'lines.m': Don't open empty figure or axes if none already exist
([bug #53648](


* gui: fix editor rendering on some displays, disable QScintilla
buffered drawing

Build system:

* 'm4/octave_blas_f77_func.m4': Restore 64-bit integer BLAS detection
([bug #53853](

* build: include '' in tarball ([bug

* fix test for 8-byte integer BLAS lib on big-endian systems ([bug

* 'OCTAVE_CHECK_LIB': avoid shell syntax error if not checking for lib

* build: avoid compiler warning about unused function 'abs' ([bug

* build: Work around bug in perl 5.28.0 ([bug


* doc: Add version number to top title of HTML manual for clarity.

* 'flip.m': Update docstring.

* 'setdiff.m': Fix typo in documentation ([bug

* doc: Replace instances of realroot (non-existent) with realpow
([bug #54336](

* '': Update for recently removed files.

* remove unused qtinfo source files

* doc: Remove bad calling convention for warning with a mode struct
([bug #54288](

* Document work-around for disabling OpenGL software rendering ([bug

* ' (err_nonsquare_matrix)': Clarify error message ([bug

* doc: Miscellaneous documentation fixes all over the manual ([bug

* doc: small fixes to 'size' and 'plot.txi' ([bug

* doc: Clarify description of Example 4 for 'size' ([bug

* doc: Fix malformed URL ([bug

* Fix typo in error statement from containers.Map ([bug

* doc: Use 'qcode' macro to document 'fseek' offsets ([bug

* Use correct abbreviation UTC for Coordinated Universal Time ([bug

* doc: Rewrite section on indexing for clarity ([bug

* 'md5sum.m': Fix error in sample code ([bug
Add a comment...

Post has attachment
Add a comment...

Post has attachment
Wait while more posts are being loaded