Talk:MATLAB/Archive 2

Archive 1

Archive 2

DADiSP

I have used both DADiSP and MATLAB and for numerical work and find DADiSP far more useful for signal processing whereas MATLAB is more complete for matrix manipulation and linear algebra. Now, that's an opinion or "advertisement." Adding text about DADiSP in the COMPETITION section of the MATLAB page is completely factual- no opinions or anything remotely similar to an advertisement is expressed. In fact, the Mathematica text seems serve much more as an opinion or advertisement. In any case, it seems the people who work for the MathWorks and edit this page would prefer not to mention any MATLAB competitors.

Fair enough. I know nothing about DADiSP, other than what I've learnt from a glance at their website, but it seemed to me that the program is mainly for the analysis of data series and therefore occupies a slightly different niche (I concede that the same point can be made about Mathematica). I want this article to be primarily about MATLAB itself, and not its competitors, but I may indeed have been too fast in reverting your contribution. To my defense, I can say that somebody removed the reference before and that it was re-added without edit summary. --Jitse Niesen 15:57, 22 Jun 2005 (UTC)

The competitive products mentioned in the article, like Mathematica and Mathcad, have user bases at least within an order of magnitude or two of MATLAB. DADiSP is one of literally dozens of niche competitors to MATLAB. I've been an employee of the MathWorks working on MATLAB for eight years and have never heard of it before. DADiSP is listed on the linked-to "numerical analysis#Software" page, so I feel like it's covered. Anonymous, please give more information about why you'd like to see it also listed on the MATLAB page. Also, because of my conflict of interest, I'd appreciate feedback from experienced editors on this issue. The section on competitive software grows like topsy and would quickly dominate the article if unchecked. I imagine information about Linux on the Microsoft Windows page (and vice versa) also needs periodic grooming. --Matthew Simoneau 06:54, 21 July 2005 (UTC)

The idea here is to spread factual and related information on a given topic. I am well versed in DADiSP and MATLAB as well as other tools in the technical computing space. As mentioned, DADiSP and MATLAB possess many similarities and some significant differences. The fact remains that almost anyone interested in MATLAB would also be interested in DADiSP and a handful of other tools - this page should reflect that interest. Of course, I fully understand that the purpose of this entry is to discuss MATLAB, but not exclusively. Otherwise the page devolves into a corporate dominated advertisement. In fact, one could assume the majority of readers finding this page are not familiar with MATLAB and therefore would be very interested in alternate technologies. Hence, a more prominent listing of alternatives is merited. In any case, DADiSP is hardly unknown and this information should not be censored simply because of Mr. Simoneau's ignorance of the product. I have the feeling that if Mr. Simoneau was employed by Microsoft and not The MathWorks, he would attempt to justify censoring Mozilla browsers and others on the Internet Explorer page with some sort of market share argument. Having a company employee add factual information is fine, even desirable if it is accurate and objective. However, having employees remove related and alternative information smacks of covert advertising. Adding a few lines about useful, related technologies increases the knowledge and awareness of readers without detracting from the primary purpose of the page. --Ed Miller

I was brought to this page by a post at the Village pump asking for wider views. In my opinion, the entire section on "Competition" should be removed from this article, and anything within it not already in the Numerical analysis "Software" section merged there. This is an article on one software program, which is undoubtedly one of the leaders in its field, and other programs need to be discussed and compared in the article on the genre, not here.

I have used MATLAB and also Rlab. I started the Wikipedia article on the latter. I added a link to the Rlab article to this article at that time, a long time ago by Wikipedia standards. I am now saying that link was not appropriate, or at least, that it is no longer appropriate.

As a similar situation where I have a vested interest, I wrote the article on Ghost, and also the generic article on Disk cloning. It would not be appropriate for other vendors to add reference to their competing software to the Ghost article, but very appropriate for them to do so in the Disk cloning article. I'm rather disappointed that few have done so. I have seen a question about competing software on the Talk:Ghost (software) page, and have tried to answer it honestly. --gadfium 09:16, 18 August 2005 (UTC)

As another outside view, after reading over the page, and this talk page, I agree with gadfium; the Competition section should go. I'm going to be bold and merge it into numerical analysis - if I have missed anything, feel free to add it; if there is disagreement, feel free to revert... --JesseW 04:30, 25 August 2005 (UTC)

I disagree strongly with the original post. I really cannot see any reason to remove the references to Octave, even in the See Also. MATLAB is a language of numerical analysis and scientific computing, and that although in other applications there are indeed niche products related to Matlab, octave is, by virtue of being both Free and almost as old as Matlab the best known in that community. I do not agree that we should have an ad-fest of Mathworks want-a-bes adding their own minority product, but not making any reference to octave for that reason severly detracts from the article. Billlion 19:47, 16 February 2007 (UTC)

MATLAB more convenient for numerical computing

I think that MATLAB is objectively more convenient for numerical computing than Mathematica, simply because it is made for numerical work. For instance, in MATLAB you can type "exp(-1)" and get 0.367879..., while in Mathematica it remains unevaluated. Of course, it is also possible to get a numerical approximation in Mathematica, but you need to specify that while MATLAB defaults to numerical operations. Similarly, Mathematica is superior in symbolic manipulation, which cannot be done at all in the standard version of MATLAB. --Jitse Niesen 10:53, 1 May 2005 (UTC)

I agree that MATLAB is superior for numerical computing, and that's why I use it and not Mathematica. But that's irrelevant. Who are you or I to judge which is easier? Why should we use an encyclopedia to endorse personal beliefs? --✈ James C. 16:54, 2005 May 1 (UTC)

What I am trying to say is that the superiority of MATLAB is not just a personal belief, it is what everybody (or at least almost everybody) in the field of numerical computing thinks and that is why we should mention it. --Jitse Niesen 22:22, 1 May 2005 (UTC)

I agree that everyone thinks MATLAB's better for numerical computing. I believe though that regardless of a belief's popularity, that an encyclopedia should contain only facts. This need not be reduce the helpfulness of articles, as many opinions can be factualized. For example, to support your belief that MATLAB better handles numerical computing, you stated some factual examples. With some mild restatement for style reasons, those examples can be added to the article and factually express the appropriateness of MATLAB for numerical computing. Facts need not be generalized as opinions. By sticking to pure facts, we maintain clarity and eliminate vague opinions. --✈ James C. 23:26, 2005 May 1 (UTC)

I agree that an encyclopedia should contain only facts, but I think that "MATLAB is more convenient than Mathematica" is a fact. I admit that the term convenient is rather vague, but it is possible to give a definition of convenient, and while we can argue for a long time about the details of the definition, MATLAB will be more convenient under all definitions. Of course, the examples are also facts. However, it is not that interesting that you use "exp(-1)" in MATLAB and something like "N[Exp[-1]]" in Mathematica. If we want to express the appropriateness of MATLAB for numerical computing, we should state it and not hide behind unimportant but easily verified examples. --Jitse Niesen 22:17, 2 May 2005 (UTC)

It seems to me that "convenience" is a matter of opinion no matter the context. In the case of Matlab and Mathematica, they both target marginally different markets and their syntax and capabilities reflect that. Personally, I prefer matlab, but an article entitled "MATLAB" hardly seems the place to start talking about the pros and cons of matlab and mathematica, a Comparison of Numerical Analysis Software page would be far more appropriate imo. --Quadbox 15:21, 22 March 2007 (UTC)

For the record entering Exp[-1.] gives the same result as Matlab with only one character difference.JonMcLoone 11:35, 8 June 2007 (UTC)

Syntax Section

The syntax section is laughable; I feel it should either be renamed (e.g., "Code snippets"), or, preferrably, expanded to at least include a general over view of the basic syntax, e.g., array construction, array accessing, function definition.

Octave a "near-clone"

The fact that Octave is a "near-clone" of Matlab is probably notable, but certainly highly subjective. Programs such as Sysquake or O-Matrix have also have a good Matlab compatibility among their goals. Sysquake has user-defined objects (such as lti and friends in the Control Toolbox) while Octave doesn't, for instance. --Steveaa 09:11, 1 September 2006 (UTC)

External links

I do not see why the article on MATLAB should have an external link to LabVIEW's website, and User:Dicklyon seems to agree. I'd say that the link to list of numerical analysis software is enough. -- Jitse Niesen (talk) 06:21, 3 November 2006 (UTC)

I take no position on whether a link to N.I. might be appropriate. I used to be happy using mex files to control N.I. boards and get work done. Good stuff. But it's not appropriate for any editor's contributions to consist of nothing but N.I. link spam. I told him so on both of tonight's IP addresses. Dicklyon 06:38, 3 November 2006 (UTC)

Octave

I added a comment to a section above on something called DADisP as the comment in the articles points to that. Look I work in Mathematics and Scientific computing, and in this sphere there is Matlab and Octave depending on the depth of your pocket. We need to have a proper debate about what this article says about Octave without having a revert war. Billlion 19:52, 16 February 2007 (UTC)

I think it's disingenuous to completely fail to a program by a well-known project (GNU) that is even attempting to create an alternative implementation. It should be described fairly, not shoved under the carpet. Superm401 - Talk 10:48, 17 February 2007 (UTC)

I don't think that it matters how well known the GNU project is. The important thing is how well known Octave is. In my kind of environment (academia + numerical analysis), Octave is very well known and no other MATLAB alternative comes close to it, and I thus support having Octave in the article. However, it is well possible that in other circles (perhaps data analysis) other alternatives (perhaps DADisP) are more well known. To make matters worse, this is the kind of thing that it is hard to find sources for.

We don't need a source simply to put something in a See also section. We only need to think it may interest the reader. If indeed "Octave is very well known", that is more than reason to include it, which I've done. Superm401 - Talk 01:36, 28 November 2008 (UTC)

By the way, I reverted the edit mentioning Octave partially because I thought there was a consensus not to, but mostly because I didn't like the added text because it seems to promote some connection between the availability of Octave and vendor lock-in, but it was not clear to me what this connection actually is. -- Jitse Niesen (talk) 13:57, 17 February 2007 (UTC)

I agree completely with the sentiment that it's disingenuous to not mention Octave and other similar software by name. Their absence feels like some kind of conspiracy of silence! The entry for Octave mentions Matlab prominently in describing what Octave is. The Matlab page should at least have a cross-reference to Octave. Tobin Fricke 25 May 2007

I could not recollect the name of Octave. I came to this page in the hope that the name Octave will be mentioned. But Alas! not. I think it is a good idea to have a section 'Related Software', containing atleast links to alternatives. It would have saved me time from searching further. 64.119.43.102 07:27, 28 June 2007 (UTC)

Octave falls far short of having all of the capabilities of MATLAB. It is probably worth mentioning that deep pockets aren't the only reason for deciding whether or not to use MATLAB. Some things simply CANNOT be done (as of today) using Octave. —Preceding unsigned comment added by 71.198.153.240 (talk) 06:45, 28 November 2007 (UTC)

I guess there will always be someone to criticize the fact that Matlab is commercial and to mention the single alternative she cares about, Octave, because it's free software which is good. Highly biased opinion, but I'll refrain from undoing that now. Engelec (talk) 16:45, 15 April 2008 (UTC)

Lists

I stumbled on this page in the Manual of Style: Wikipedia:Embedded list. It suggests to me that we shouldn't have a section which is just a list of MATLAB alternatives, especially since there is already a page, List of numerical analysis software, which the article already links to. I removed the list and added a comment near the link that some of these other programs are free and compatible with MATLAB. Matthew Simoneau 14:40, 20 February 2007 (UTC)

Easter Eggs

Has anyone considered adding an Easter Eggs section. There are quite a few random extras hidden in MATLAB that many people don't know about. Typing the word "why" for instance generates completely random responses. Typing "toilet" gives a graphical simulation of a toilet. I would also be interested in seeing the list grow as more Wikipedia users edit the article. —The preceding unsigned comment was added by 12.216.63.82 (talk) 00:43, 12 April 2007 (UTC).

There was an "Easter eggs" section, but someone removed it for being inappropriate. I don't know what the Wikipedia standards are for such things. Matthew Simoneau 16:43, 19 April 2007 (UTC)

Many articles have a "Trivia" section. This would seem adequate to me. --John.constantine 14:09, 7 May 2007 (UTC)

Trivia sections are officially discouraged, and should include a tag like this:

If you add one, I will remove it. Dicklyon 14:41, 7 May 2007 (UTC)

Userbox

I just discovered MATLAB's userbox. To add one to your user page, just type {{user matlab}}. Matthew Simoneau 16:38, 19 April 2007 (UTC)

Computer Algebra System

Why is this in the category for computer algebra systems? As far as I know there isn't any way to do symbolic manipulation with matlab, is there? risk 12:31, 14 October 2007 (UTC)

Yes, there is. I think it has an interface to Maple, iirc. But if it's not detailed in the article, it shouldn't be in that category. Dicklyon 16:41, 14 October 2007 (UTC)

Yes, its in an optional "toolbox", and it is a crude Maple interface. Should not be in cat. Billlion 17:02, 14 October 2007 (UTC)

Release date history

It would be nice if there was a history of different releases - i.e. the date of various major releaes. I find it confusing to know of Matlab 6.5, Rlelease 13 (I've no idea if release 13 is before or after 6.5), then 2007a. —Preceding unsigned comment added by 213.78.42.15 (talk) 13:08, 19 November 2007 (UTC)

I second this argument, I came to wikipedia page exactly to find that info and I (obviously) wasn't able to. --85.18.14.13 (talk) 17:00, 7 January 2008 (UTC) A.E.

Freemat

Perhaps, just perhaps, include Freemat in the alternatives section? It's open-source. Danuthaiduc (talk) 17:09, 21 November 2007 (UTC)

There is no alternative section, and I think it's a good thing, since it's bound to be uncomplete. Freemat is listed in List of numerical analysis software which shouldn't be replicated here. Engelec (talk) 14:47, 23 November 2007 (UTC)

Toolboxes

I think Matlab's collection of toolboxes are notable, but they're not mentioned in the article except in connection with Maple.

MusicScience (talk) 20:38, 20 January 2008 (UTC)