<html lang="en"> <head> <title>@sc{Matlab} compatibility - Frequently asked questions about Octave (with answers)</title> <meta http-equiv="Content-Type" content="text/html"> <meta name="description" content="Frequently asked questions about Octave (with answers)"> <meta name="generator" content="makeinfo 4.13"> <link title="Top" rel="start" href="index.html#Top"> <link rel="prev" href="Using-Octave.html#Using-Octave" title="Using Octave"> <link rel="next" href="Index.html#Index" title="Index"> <link href="http://www.gnu.org/software/texinfo/" rel="generator-home" title="Texinfo Homepage"> <meta http-equiv="Content-Style-Type" content="text/css"> <style type="text/css"><!-- pre.display { font-family:inherit } pre.format { font-family:inherit } pre.smalldisplay { font-family:inherit; font-size:smaller } pre.smallformat { font-family:inherit; font-size:smaller } pre.smallexample { font-size:smaller } pre.smalllisp { font-size:smaller } span.sc { font-variant:small-caps } span.roman { font-family:serif; font-weight:normal; } span.sansserif { font-family:sans-serif; font-weight:normal; } --></style> </head> <body> <div class="node"> <a name="%3cspan-class%3d%22sc%22%3eMatlab%3c%2fspan%3e-compatibility"></a> <a name="g_t_003cspan-class_003d_0022sc_0022_003eMatlab_003c_002fspan_003e-compatibility"></a> <p> Next: <a rel="next" accesskey="n" href="Index.html#Index">Index</a>, Previous: <a rel="previous" accesskey="p" href="Using-Octave.html#Using-Octave">Using Octave</a>, Up: <a rel="up" accesskey="u" href="index.html#Top">Top</a> <hr> </div> <h2 class="chapter">11 Porting programs from <span class="sc">Matlab</span> to Octave</h2> <p><a name="index-g_t_0040sc_007bMatlab_007d-compatibility-36"></a><a name="index-Compatibility-with-_0040sc_007bMatlab_007d-37"></a> People often ask <blockquote> I wrote some code for <span class="sc">Matlab</span>, and I want to get it running under Octave. Is there anything I should watch out for? </blockquote> <p class="noindent">or alternatively <blockquote> I wrote some code in Octave, and want to share it with <span class="sc">Matlab</span> users. Is there anything I should watch out for? </blockquote> <p class="noindent">which is not quite the same thing. There are still a number of differences between Octave and <span class="sc">Matlab</span>, however in general differences between the two are considered as bugs. Octave might consider that the bug is in <span class="sc">Matlab</span> and do nothing about it, but generally functionality is almost identical. If you find a difference between Octave behavior and <span class="sc">Matlab</span>, then you should send a description of this difference (with code illustrating the difference, if possible) to <a href="http://bugs.octave.org">http://bugs.octave.org</a>. <p>Furthermore, Octave adds a few syntactical extensions to <span class="sc">Matlab</span> that might cause some issues when exchanging files between <span class="sc">Matlab</span> and Octave users. As both Octave and <span class="sc">Matlab</span> are under constant development the information in this section is subject to change at anytime. <p>You should also look at the page <a href="http://octave.sourceforge.net/packages.html">http://octave.sourceforge.net/packages.html</a> and <a href="http://octave.sourceforge.net/doc/">http://octave.sourceforge.net/doc/</a> that has a function reference that is up to date. You can use this function reference to see the number of octave function that are available and their <span class="sc">Matlab</span> compatibility. <p>The major differences between Octave 3.4.N and <span class="sc">Matlab</span> R2010b are: <ul> <li>Nested Functions <p>Octave has limited support for nested functions. That is <pre class="example"> function y = foo (x) y = bar(x) function y = bar (x) y = ...; end end </pre> <p>is equivalent to <pre class="example"> function y = foo (x) y = bar(x) end function y = bar (x) y = ...; end </pre> <p>The main difference with <span class="sc">Matlab</span> is a matter of scope. While nested functions have access to the parent function's scope in <span class="sc">Matlab</span>, no such thing is available in Octave, due to how Octave essentially “un-nests” nested functions. <p>The authors of Octave consider the nested function scoping rules of <span class="sc">Matlab</span> to be more problems than they are worth as they introduce difficult to find bugs as inadvertently modifying a variable in a nested function that is also used in the parent is particularly easy. <li>Differences in core syntax There a few core <span class="sc">Matlab</span> syntaxes that are not accepted by Octave, these being <ul> <li>Some limitations on the use of function handles. The major difference is related to nested function scoping rules (as above) and their use with function handles. <li>Some limitations of variable argument lists on the LHS of an expression, though the most common types are accepted. <li><span class="sc">Matlab</span> classdef object oriented programming is not yet supported, though work is underway and when development more on to Octave 3.5 this will be included in the development tree. </ul> <li>Differences in core functions A large number of the <span class="sc">Matlab</span> core functions (ie those that are in the core and not a toolbox) are implemented, and certainly all of the commonly used ones. There are a few functions that aren't implemented, usually to do with specific missing Octave functionality (GUI, DLL, Java, ActiveX, DDE, web, and serial functions). Some of the core functions have limitations that aren't in the <span class="sc">Matlab</span> version. For example the <code>sprandn</code> function can not force a particular condition number for the matrix like <span class="sc">Matlab</span> can. <li>Just-In-Time compiler <span class="sc">Matlab</span> includes a "Just-In-Time" compiler. This compiler allows the acceleration of for-loops in <span class="sc">Matlab</span> to almost native performance with certain restrictions. The JIT must know the return type of all functions called in the loops and so you can't include user functions in the loop of JIT optimized loops. Octave doesn't have a JIT and so to some might seem slower than <span class="sc">Matlab</span>. For this reason you must vectorize your code as much as possible. The MathWorks themselves have a good document discussing vectorization at <!-- It would be nice if we had our own guide for this instead of relying --> <!-- on Matlab documentation. --> <a href="http://www.mathworks.com/support/tech-notes/1100/1109.html">http://www.mathworks.com/support/tech-notes/1100/1109.html</a>. <li>Compiler On a related point, there is no Octave compiler, and so you can't convert your Octave code into a binary for additional speed or distribution. There have been several aborted attempts at creating an Octave compiler. Should the JIT compiler above ever be implemented, an Octave compiler should be more feasible. <!-- Should we mention here any of the efforts to create a compiler? There --> <!-- used to be a dead link here to http://www.stud.tu-ilmenau.de/~rueckn/ --> <li>Graphic Handles Up to Octave 2.9.9 there was no support for graphic handles in Octave itself. In the 3.2.N versions of Octave and beyond the support for graphics handles is converging towards full compatibility. The <code>patch</code> function is currently limited to 2-D patches, due to an underlying limitation in gnuplot, but the experimental OpenGL backend is starting to see an implementation of 3-D patches. <li>GUI There are no <span class="sc">Matlab</span> compatible GUI functions. There are a number of bindings from Octave to Tcl/Tk, VTK and Zenity included in the Octave <!-- Is it too early to mention here the nascent fltk UI buttons? --> Forge project (<a href="http://octave.sourceforge.net">http://octave.sourceforge.net</a>) for example that can be used for a GUI, but these are not <span class="sc">Matlab</span> compatible. Work on a <span class="sc">Matlab</span> compatible GUI is in an alpha stage in the JHandles package <!-- Is Jhandles still usable? I thought Michael Goffioul had more or less --> <!-- already disowned it. --> (<a href="http://octave.sourceforge.net/jhandles/index.html">http://octave.sourceforge.net/jhandles/index.html</a>). This might be an issue if you intend to exchange Octave code with <span class="sc">Matlab</span> users. <li>Simulink Octave itself includes no Simulink support. Typically the simulink models lag research and are less flexible, so shouldn't really be used in a research environment. However, some <span class="sc">Matlab</span> users that try to use Octave complain about this lack. There is a similar package to simulink for the Octave and R projects available at <!-- is this project in any state of usability? --> <a href="http://www.scicraft.org/">http://www.scicraft.org/</a> <li>Mex-Files Octave includes an API to the <span class="sc">Matlab</span> MEX interface. However, as MEX is an API to the internals of <span class="sc">Matlab</span> and the internals of Octave differ from <span class="sc">Matlab</span>, there is necessarily a manipulation of the data to convert from a MEX interface to the Octave equivalent. This is notable for all complex matrices, where <span class="sc">Matlab</span> stores complex arrays as real and imaginary parts, whereas Octave respects the C99/C++ standards of co-locating the real/imag parts in memory. Also due to the way <span class="sc">Matlab</span> allows access to the arrays passed through a pointer, the MEX interface might require copies of arrays (even non complex ones). <li>Block comments Block comments denoted by "%{" and "%}" markers are supported by Octave with some limitations. The major limitation is that block comments are not supported within [] or {}. <li>Mat-File format There are some differences in the mat v5 file format accepted by Octave. <span class="sc">Matlab</span> recently introduced the "-V7.3" save option which is an HDF5 format which is particularly useful for 64-bit platforms where the standard <span class="sc">Matlab</span> format can not correctly save variables. Octave accepts HDF5 files, but is not yet compatible with the "-v7.3" versions produced by <span class="sc">Matlab</span>. <p>Although Octave can load inline function handles saved by <span class="sc">Matlab</span>, it can not yet save them. <p>Finally, Some multi-byte Unicode characters aren't yet treated in mat-files. <li>Profiler Octave doesn't have a profiler. Though there is a patch for a flat profiler, that might become a real profiler sometime in the future. See the thread <!-- Did this idea go anywhere? Should it be mentioned? --> <p><a href="http://octave.1599824.n4.nabble.com/Octave-profiler-td1641945.html#a1641947">http://octave.1599824.n4.nabble.com/Octave-profiler-td1641945.html#a1641947</a> <p>for more details. <li>Toolboxes Octave is a community project and so the toolboxes that exist are donated by those interested in them through the Octave Forge website (<a href="http://octave.sourceforge.net">http://octave.sourceforge.net</a>). These might be lacking in certain functionality relative to the <span class="sc">Matlab</span> toolboxes, and might not exactly duplicate the <span class="sc">Matlab</span> functionality or interface. <li>Short-circuit & and | operators The <code>&</code> and <code>|</code> operators in <span class="sc">Matlab</span> short-circuit when included in an if statement and not otherwise. In Octave only the <code>&&</code> and <code>||</code> short circuit. Note that this means that <pre class="example"> if (a | b) ... end </pre> <p>and <pre class="example"> t = a | b; if t ... end </pre> <p class="noindent">are different in <span class="sc">Matlab</span>. This is really a <span class="sc">Matlab</span> bug, but there is too much code out there that relies on this behaviour to change it. Prefer the || and && operators in if statements if possible. If you need to use code written for <span class="sc">Matlab</span> that depends on this buggy behaviour, you can enable it since Octave 3.4.0 with the following command: <pre class="example"> do_braindead_shortcircuit_evaluation(1) </pre> <p>Note that the difference with <span class="sc">Matlab</span> is also significant when either argument is a function with side effects or if the first argument is a scalar and the second argument is an empty matrix. For example, note the difference between <pre class="example"> t = 1 | []; ## results in [], so... if (t) 1, end ## in if ([]), this is false. </pre> <p>and <pre class="example"> if (1 | []) 1, end ## short circuits so condition is true. </pre> <p>Another case that is documented in the <span class="sc">Matlab</span> manuals is that <pre class="example"> t = [1, 1] | [1, 2, 3]; ## error if ([1, 1] | [1, 2, 3]) 1, end ## OK </pre> <p>Also <span class="sc">Matlab</span> requires the operands of && and || to be scalar values but Octave does not (it just applies the rule that for an operand to be considered true, every element of the object must be nonzero or logically true). <p>Finally, note the inconsistence of thinking of the condition of an if statement as being equivalent to <code>all(X(:))</code> when <var>X</var> is a matrix. This is true for all cases EXCEPT empty matrices: <pre class="example"> if ([0, 1]) == if (all ([0, 1])) ==> i.e., condition is false. if ([1, 1]) == if (all ([1, 1])) ==> i.e., condition is true. </pre> <p>However, <pre class="example"> if ([]) != if (all ([])) </pre> <p>because <code>samp ([]) == 1</code> because, despite the name, it is really returning true if none of the elements of the matrix are zero, and since there are no elements, well, none of them are zero. This is an example of vacuous truth. But, somewhere along the line, someone decided that if <code>([])</code> should be false. Mathworks probably thought it just looks wrong to have <code>[]</code> be true in this context even if you can use logical gymnastics to convince yourself that "all" the elements of a matrix that doesn't actually have any elements are nonzero. Octave however duplicates this behavior for if statements containing empty matrices. <li>Solvers for singular, under- and over-determined matrices <p><span class="sc">Matlab</span>'s solvers as used by the operators mldivide (\) and mrdivide (/), use a different approach than Octave's in the case of singular, under-, or over-determined matrices. In the case of a singular matrix, <span class="sc">Matlab</span> returns the result given by the LU decomposition, even though the underlying solver has flagged the result as erroneous. Octave has made the choice of falling back to a minimum norm solution of matrices that have been flagged as singular which arguably is a better result for these cases. <p>In the case of under- or over-determined matrices, Octave continues to use a minimum norm solution, whereas <span class="sc">Matlab</span> uses an approach that is equivalent to <pre class="example"> function x = mldivide (A, b) [Q, R, E] = qr(A); x = [A \ b, E(:, 1:m) * (R(:, 1:m) \ (Q' * b))] end </pre> <p class="noindent">While this approach is certainly faster and uses less memory than Octave's minimum norm approach, this approach seems to be inferior in other ways. <p>A numerical question arises: how big can the null space component become, relative to the minimum-norm solution? Can it be nicely bounded, or can it be arbitrarily big? Consider this example: <pre class="example"> m = 10; n = 10000; A = ones(m, n) + 1e-6 * randn(m,n); b = ones(m, 1) + 1e-6 * randn(m,1); norm(A \ b) </pre> <p class="noindent">while Octave's minimum-norm values are around 3e-2, <span class="sc">Matlab</span>'s results are 50-times larger. For another issue, try this code: <pre class="example"> m = 5; n = 100; j = floor(m * rand(1, n)) + 1; b = ones(m, 1); A = zeros(m, n); A(sub2ind(size(A),j,1:n)) = 1; x = A \ b; [dummy,p] = sort(rand(1,n)); y = A(:,p)\b; norm(x(p)-y) </pre> <p class="noindent">It shows that unlike in Octave, mldivide in <span class="sc">Matlab</span> is not invariant with respect to column permutations. If there are multiple columns of the same norm, permuting columns of the matrix gets you different result than permuting the solution vector. This will surprise many users. <p>Since the mldivide (\) and mrdivide (/) operators are often part of a more complex expression, where there is no room to react to warnings or flags, it should prefer intelligence (robustness) to speed, and so the Octave developers are firmly of the opinion that Octave's approach for singular, under- and over-determined matrices is a better choice that <span class="sc">Matlab</span>'s <li>Octave extensions The extensions in Octave over <span class="sc">Matlab</span> syntax are very useful, but might cause issues when sharing with <span class="sc">Matlab</span> users. A list of the major extensions that should be avoided to be compatible with <span class="sc">Matlab</span> are <ul> <li>Comments in octave can be marked with ‘<samp><span class="samp">#</span></samp>’. This allows POSIX systems to have the first line as ‘<samp><span class="samp">#! octave -q</span></samp>’ and mark the script itself executable. <span class="sc">Matlab</span> doesn't have this feature due to the absence of comments starting with ‘<samp><span class="samp">#</span></samp>’". <li>Code blocks like <code>if</code>, <code>for</code>, <code>while</code>, etc can be terminated with block specific terminations like <code>endif</code>. <span class="sc">Matlab</span> doesn't have this and all blocks must be terminated with <code>end</code>. <li>Octave has a lisp like <code>unwind_protect</code> block that allows blocks of code that terminate in an error to ensure that the variables that are touched are restored. You can do something similar with <code>try</code>/<code>catch</code> combined with ‘<samp><span class="samp">rethrow (lasterror ())</span></samp>’ in <span class="sc">Matlab</span>, however rethrow and lasterror are only available in Octave 2.9.10 and later. <span class="sc">Matlab</span> 2008a also introduced <code>OnCleanUp</code> that is similar to <code>unwind_protect</code>, except that the object created by this function has to be explicitly cleared in order for the cleanup code to run. <p>Note that using <code>try</code>/<code>catch</code> combined with ‘<samp><span class="samp">rethrow (lasterror ())</span></samp>’ can not guarantee that global variables will be correctly reset, as it won't catch user interrupts with Ctrl-C. For example <pre class="example"> global a a = 1; try _a = a; a = 2 while true end catch fprintf ('caught interrupt\n'); a = _a; rethrow (lasterror()); end </pre> <p class="noindent">compared to <pre class="example"> global a a = 1; unwind_protect _a = a; a = 2 while true end unwind_protect_cleanup fprintf ('caught interrupt\n'); a = _a; end </pre> <p>Typing Ctrl-C in the first case returns the user directly to the prompt, and the variable "a" is not reset to the saved value. In the second case the variable "a" is reset correctly. Therefore <span class="sc">Matlab</span> gives no safe way of temporarily changing global variables. <li>Indexing can be applied to all objects in Octave and not just variable. Therefore <code>sin(x)(1:10);</code> for example is perfectly valid in Octave but not <span class="sc">Matlab</span>. To do the same in <span class="sc">Matlab</span> you must do <code>y = sin(x); y = y([1:10]);</code> <li>Octave has the operators "++", "–", "-=", "+=", "*=", etc. As <span class="sc">Matlab</span> doesn't, if you are sharing code these should be avoided. <li>Character strings in Octave can be denoted with double or single quotes. There is a subtle difference between the two in that escaped characters like <code>\n</code> (newline), <code>\t</code> (tab), etc are interpreted in double quoted strings but not single quoted strings. This difference is important on Windows platforms where the "\" character is used in path names, and so single quoted strings should be used in paths. <span class="sc">Matlab</span> doesn't have double quoted strings and so they should be avoided if the code will be transferred to a <span class="sc">Matlab</span> user. </ul> </ul> </body></html>