27

Note: There seems to be a lot of support for having our own blog, and we have several volunteers, so I've added the tag to this post to draw the attention of the SE team and hopefully set up a blog.


Update: Here's some info on creating/requesting a blog.


Several StackExchange sites have a blog, including some beta sites.

We can request a blog as well. The idea came up before, but it was in the form of using an external blog provider so it was downvoted.

In chat discussions, several people expressed interest in having a blog for this site or contributing to it, but nothing concrete was started.

The purpose of this meta thread is to find out if there is enough interest and enough contributors to start a blog on BlogOverflow.

Those of you who would like to contribute to a blog occasionally, please post an answer below and tell us in a few words what you have in mind.

If we have enough contributors, we can go ahead and request a blog. I think even a low volume blog (say, a post every month or six weeks) would be valuable.

  • 2
    Maybe someone should write a post about how to ask a good question about Mathematica. What measures to take, how to reduce the problem, find the bottleneck and construct a helpful minimal example. As the number of newcomers is increasing, such a short post could be useful. – István Zachar May 28 '12 at 18:12
  • 5
    Once the blog exists, is there somewhere that draft posts get posted for comment and editing? Even if I don't have time to write my own posts, if there's one thing I can do it's copyedit/tighten up drafting on other people's prose. I pretty much do it for a living nowadays. – Verbeia May 29 '12 at 4:50
  • @Verbeia I would really appreciate if you could do that with mine. If there's no draft view feature in the blog system (though there should be one), we can always communicate by other means. – Szabolcs May 29 '12 at 20:17
  • 3
    @Verbeia: The blog system has drafts, and you can share them. (I know from the Cooking.SE blog) – derobert May 30 '12 at 15:24
  • Would the blog have MathJax and code pretty-printing? Maybe more importantly, would it be possible to get CDF embedding code into that blog? It seems to be WordPress based, so that should in principle be possible, right? – Jens May 30 '12 at 16:07
  • 5
    It would also help to get some Wolfram employees to write posts for this blog, to give it credibility. Most recently, I thought @Yu-Sung Chang writes in a very educational style that would be great for a blog. – Jens May 30 '12 at 16:21
  • 1
    @Jens I agree, and I did comment on his last post suggesting it as a blog topic. – Szabolcs May 30 '12 at 16:22
  • 2
    I will try to come up with some ideas when the blog is there. – Sjoerd C. de Vries Jun 1 '12 at 21:17

12 Answers 12

34

I would like to contribute to the blog too. A couple of topics that are within my reach are:

  • Making the transition from MATLAB to Mathematica (series) — aimed at users who are repeatedly told and fed the myth that Mathematica is slower and only good for computing the 1000th digit of π (this comparison, which is a top hit for "MATLAB vs Mathematica" doesn't help either, when it really is a case study in how not to program in Mathematica).

    The series will cover the basics of the different programming paradigms and how functional and rule-based should be favoured over procedural in Mathematica, working with and manipulating matrices and cells in the former vs. regular lists and irregular/non-rectangular lists in the latter, eventually moving on to covering the functionality provided by the toolboxes (e.g., image processing, curve fitting, optimization, statistics, etc.).

  • Creating publication quality figures using Mathematica (series, co-authored) — There are several posts here on rasterizing/anti-aliasing/grayscale plots/legends/printers points, etc., and it would be nice to expand on those as a recurring series. I'd prefer if the original answerer(s) wrote the article, but I'd be willing to write/co-write if they cannot (and if I can spare the time). I could also write an article on using LevelScheme, which is an equally good system for creating good looking figures (I know that rcollyer uses it actively, so perhaps this can be a collaborative post).

  • 6
    +1 on posting about transitioning from matlab to mma – JxB May 28 '12 at 6:25
  • 1
    That comparison is extremely annoying. I wrote an email to the author in my master student days, asking for the data files so I could improve the code, but got no reply. Maybe, as a master, I sounded 'too excited' ... – Szabolcs May 28 '12 at 6:46
  • 2
    Great idea! Being perhaps the only one here with dual expert-level Matlab and Mathematica background, you are in a unique position to debunk these myths of Matlab numerical superiority in terms of speed etc.+1. – Leonid Shifrin May 28 '12 at 8:37
  • 1
    Yes, we are in dire need of a clear comparison of Mathematica vs. MATLAB that can serve as a base to convince/attract people to use Mathematica. – István Zachar May 28 '12 at 10:34
  • 2
    I don't recall ever seeing that MATLAB v. Mathematica post before. That's some of the worst mma code I've seen in a long time. – rcollyer May 29 '12 at 1:09
  • @rcollyer you probably never searched for "Matlab vs mathematica" ;) – rm -rf May 29 '12 at 1:10
  • No, no I haven't. – rcollyer May 29 '12 at 1:11
  • GREAT idea!! Would be very useful to some people I know. I'll be cheering – Rojo May 29 '12 at 1:24
  • 1
    I could help with some of the publication-quality figures stuff, based on this, this, this, this and my current project (to the extent that I can if I run into IP issues). – Verbeia May 29 '12 at 4:34
  • 5
    I'd be willing to help with the figure creation idea. There are some things I haven't posted as answers that could be better for a blog. Ultimately, that brings up the question whether this blog post (and others like it) could be accompanied by a MMA package that people could download to conveniently use the insights of the post. – Jens May 30 '12 at 14:39
  • 3
    A series on transitioning from MATLAB to Mathematica would be great, and very useful, but posts on Mathematica vs MATLAB ("which is better") I would rather avoid. They'd be too controversial. Let's just leave out any 'which is better' type comparison, and focus on practical things. – Szabolcs May 30 '12 at 21:05
  • 2
    @Szabolcs I don't think we have to be particularly politically correct towards Matlab here, as this is a site devoted to Mathematica. My main beef with Matlab (and I did all computations for my Bachelor's thesis using it, and have dealt with it more recently as well) is that in many cases, using it confines one to a mindset which prevents creative and novel solutions. I also believe that its total domination in certain areas is harmful for those areas, and promoting healthy competition is good IMO. But I give it credit where it's due - it is a very practical tool for certain kinds of problems. – Leonid Shifrin May 31 '12 at 9:49
  • 3
    Actually showing areas where Matlab is better than Mathematica would also help Mathematica, by showing where it has room for improvement. Also a biased comparison would not really serve Mathematica, because even you say "Mathematica is better here" even where it isn't, you'll likely get many people trying it, seeing that for their problem (which happens to be one which works better on Matlab) Mathematica performs worse and make the conclusion that Mathematica always will perform worse, and everyone saying differently must be either clueless or a shill. So if Matlab is better at ... – celtschk Jun 2 '12 at 18:34
  • 1
    ... some specific problem, it also serves Mathematica better to openly say so. Otherwise you only remove credibility of statements that Mathematica is better in areas where Mathematica is better. Don't get me wrong: There's nothing wrong with highlighting Mathematica's strengths. However, it is equally important to also admit any weaknesses. Discalimer: I've never used Matlab, so I don't know about any of its strengths and weaknesses, or how they compare to Mathematica. I just wanted to counter the idea that a biased blog entry would only hurt Matlab fans. It would hurt Mathematica as well. – celtschk Jun 2 '12 at 18:41
  • 3
    @celtschk, say that to Stephen and his marketing department – Rojo Jun 6 '12 at 13:13
29

I would like to write on something, but not yet sure on what :). Here are some topics which came to mind now:

  • Mathematica evaluation sequence, tools of evaluation control, and some practical examples of their use.

  • Advanced programming tools, such as higher-order functions, closures and macros, and how they help design modular programs.

  • Scoping as a practical programming tool.

  • A war story: computing one indefinite integral ( using hybrid symbolic / numeric approaches in Mathematica to obtain new analytical results, particular example is in the theory of Bessel functions).

  • Mathematica - Java interop: effective use of JLink and other tools to get the best of both worlds

  • Linked lists in Mathematica - a heavily underrated tool.

  • 7
    You had my vote at "I would like to write..." :-) – Mr.Wizard May 26 '12 at 19:14
  • 1
    @Mr.Wizard And you had mine on infix :) – Leonid Shifrin May 26 '12 at 19:15
  • 2
    It's all interesting. Right now, I'd like to hear more about the Bessel functions. – Jens May 30 '12 at 15:21
  • @Jens I can imagine most folks here are already quite tired of the programming stuff usually coming from me :) This is one ghost from the past, when I was still doing some real things :) I will keep your preference in mind, perhaps will make this the first one. It is basically ready (I have a notebook with the complete thing), just needs to be put into a blog post. – Leonid Shifrin May 30 '12 at 15:30
  • I have to second Jens preferences, it sounds like a harrowing tale. :) – rcollyer Jun 4 '12 at 1:35
  • These are all excellent ideas but for me the last two (war story and Java interop) sound like both the most interesting suggestions and also the most difficult topics to deal with in the Q&A format. Strong +1. – Oleksandr R. Jun 6 '12 at 17:34
  • @OleksandrR. Thanks :) Given such a strong reaction, it will be hard to keep up to the expectations, but I will do my best :) – Leonid Shifrin Jun 6 '12 at 19:30
  • Leonid, I have to agree with you on linked lists: I never know when they are useful in mma, and what sort of speed ups I can get by using them. – rcollyer Jun 11 '12 at 14:29
  • @rcollyer Actually, my own appreciation of their usefulness progressed as I accumulated more cases where I was able to successfully use them, and most of these cases originated from questions asked on SO and then here. It is rather recently that I finally came to a conclusion that they are not just occasionally useful but can be used in many more situations than I anticipated before. They give some of the best performance possible with the top-level code, code is easy to write and feels natural, is often short and elegant, and free of side effects - what's not to like? – Leonid Shifrin Jun 11 '12 at 14:36
  • So, Blog first, developer's conference later? :) – rcollyer Jun 11 '12 at 14:40
  • @rcollyer Sounds like a plan :) The order depends on whichever comes first, I hope the blog. Actually, I was thinking of a series of posts on underrated programming techniques in Mathematica programming, but since I already listed quite a few things above, I decided to stay modest and only list the linked lists case so far :) – Leonid Shifrin Jun 11 '12 at 14:43
  • That's why I only have one thing listed, at the moment: time, and several things are not in a state where I'm ready to show them to the world. :) – rcollyer Jun 11 '12 at 14:57
  • @rcollyer Yes, I see your point. I listed more because some of them I have mostly ready, either as some parts of the new book I am working on (which actually slowed down to a halt in the last half a year, which is sad), or, like with the Bessel stuff, things from the past which I happened to have revisited lately. – Leonid Shifrin Jun 11 '12 at 15:00
  • Leonid, you're probably very busy since you haven't been seen in 2 days. I was just checking up to see if you have any blogs in the pipeline (now that new groom Szabolcs will be away for a while). I left a message in the blog chatroom. Please don't feel pressured... there is no hurry. – rm -rf Aug 3 '12 at 15:49
  • Thanks @R.M.. I am rather busy indeed. It should not take long to write a blog post on Bessel functions, since I have that material already used in a presentation, but I will only have time for it in a week or something. – Leonid Shifrin Aug 3 '12 at 16:33
27

I would suggest a series of posts on "obscure Mathematica function or option of the week". It could cover things like the level specifications in Join and Flatten, non-obvious uses of Inner and Outer, the finer points of UnitStep, Boole and DiracDelta etc, as well as some of the less well known functions, such as those in the model-fitting functionality.

I am not suggesting that I would be the person to write those posts - work is getting a bit busy and I need to focus on some other projects. But I think they would be useful complements to the main Q&A aspect of the site, and the official documentation. Maybe I could do one or two.

25

I could do a few posts as well.

  • I was thinking of a post explaining how graphics work, the different types of units that Mathematica uses (plot coordinates, scaled coordinates, offset coordinates), and also discuss exporting publication quality graphics. (This is related to @R.M.'s second topic, it could be part of that series.)

  • I'd like to do a post on Mathematica and generative art. Here's an old book with some fun images. This one needs more work before I can get started on it though.

  • A tutorial-like post on extending Mathematica with C/C++ (MathLink, LibraryLink, etc.) (Possibly as a co-authored series.)

Finally, I think it's important that we have blog posts for audiences with all levels of Mathematica knowledge, both beginner and advanced.

  • +1. I like your choice of topics. The last one is probably closest to my area of "professional" interest, but I am also very curious about the first two. – Leonid Shifrin May 28 '12 at 8:39
  • +1 for having beginner topics as well. – enedene Jun 1 '12 at 9:26
23

I would be interested in writing an article describing some terse coding methods and syntax tricks. I think a blog would be a good place to assemble and describe things like ##&[]. It could also be a good place to make the case for ~infix~ much to everyone's delight. ;^)

  • 2
    I can't wait to read the comments on your post! – Dr. belisarius May 26 '12 at 17:49
  • 2
    I think all comments to that blog post should also be required to use infix. – Leonid Shifrin May 26 '12 at 18:37
  • 2
    will it also be a terse article? ;) – rm -rf May 27 '12 at 15:02
  • 1
    @Leonid infix is the closest to the natural form of English; I propose instead that any criticisms be written in match-form or whatever form is preferred by the critic. lol – Mr.Wizard May 28 '12 at 1:56
  • 6
    I object. The purpose of such a post the usefulness of such a form to highlight is. :) – rcollyer May 29 '12 at 1:23
  • 2
    Some Star Wars characters come to mind as prime syntactic judge material. – Yves Klett Jun 5 '12 at 12:33
  • 1
    @YvesKlett it took some thought to put together my prior comment. – rcollyer Jun 5 '12 at 13:34
  • @rcollyer You should improve your mindset then :) – Dr. belisarius Jun 7 '12 at 7:24
23

I would like to contribute too. There are a few answers of mine such as the word cloud one which I would like to elaborate and I think a blog would be a good place for that. I can't guarantee that my posts will be as insightful as Leonid's or Mr.Wizard's are indubitably going to be, but I can guarantee pretty pictures at least.

  • 7
    You are too modest. I am sure your blog posts will be as exceptional as your answers are. – Leonid Shifrin May 27 '12 at 7:09
  • 1
    I was hoping you'd write a blog post on the word cloud, it's one of my favourite answers here! – Szabolcs May 27 '12 at 9:43
  • 2
    Gimme them pwetty picturez :-) – Yves Klett Jun 5 '12 at 12:23
13

I think there are three topics where I both feel confident enough and have the code almost ready at hand to contribute to the blog:

  • Building graphical user interfaces: how to create new controllers, how to put together a fully-fledged GUI. Recently I had to design a lot of these things, and during this excursion I've learnt a lot that might be useful for others too.
  • Dealing with reaction systems: how to get from a system of (chemical) reactions through differential equations and numerical integration to visualizing the behaviour of the system. Applying Mathematica to Systems Biology problems.
  • Advanced numerical differetial equation solving: how to track equilibrium, perturb differential equation systems and perform other tricks during numerical DE solving.

I hope even the more scientific topics can be presented in such a general, domain-independent and easy-to-understand way that can help people to apply the solutions/ideas to their own fields. And of course all three readily lend themselves for fancy dynamic visualization and embedded CDF content.

  • +1, Great suggestions! – Leonid Shifrin Jun 25 '12 at 18:47
  • Thanks, hopefully I will have the time to do this. By the way, I must have felt some disturbance in the Force to post this answer just right before Szabolcs has announced the blog. – István Zachar Jun 25 '12 at 19:00
  • Well, having R.M. around, we al probably became more sensitive to the Force :-) – Leonid Shifrin Jun 25 '12 at 19:09
  • I think that building graphical user interfaces in Mathematica is always a pain, so any article about this would be really great !! – crazzymath May 28 '13 at 14:14
11

One idea I had last night for a series was to have various people write about what's in their tool bag and how they customize their Mathematica environment.

Our local newspaper has a few regular features like this, such as Studio Visit (arts & entertainment) and Gearhead of the Week (vintage automobiles).

Potentially this is something that a lot of people could do, and with a bit less effort than writing about a topic in detail.

  • That's a good idea, I saw that you have a docked cell toolbar in some screenshots you posted. – Szabolcs Jun 15 '12 at 15:15
11

There are several entries I'd like to make as time permits:

  • Begin and BeginPackage versus c++ namespaces. Their behaviors can be surprising.

and others as they occur to me. Or, more correctly, as I think I have something complete enough to discuss.

6

I'd like to suggest a post on profiling Mathematica programs.

There is the WB profiler, but right now I'm just using Print@AbsoluteTiming allover my code. Probably there has to be a better way :)

This is just a topic suggestion -- I'm not planning to write it.

  • Are you planning to write a post on this, or is this just a topic suggestion? – Szabolcs Jun 27 '12 at 12:17
  • @Szabolcs Just a topic suggestion. – Ajasja Jun 27 '12 at 12:58
  • I think many of us would love to see a non-WB-based profiling solution (see this post), though writing this code would be a huge undertaking. – István Zachar Jun 28 '12 at 9:11
  • @Istvan Have you seen a light-weight profiler I posted here? – Leonid Shifrin Jun 28 '12 at 23:25
  • @Ajasja +1, a really good topic IMO (although I personally don't use a profiler :-)) – Leonid Shifrin Jun 28 '12 at 23:26
  • @Leonid: Yes, I've seen it back in the time (and even voted for it!), but strangely I could not find it yesterday and thought that I might have seen it somewhere else, and there is no chance to find that site over the internet anymore. Thanks for the link! I really have to start bookmarking these posts... – István Zachar Jun 29 '12 at 8:21
6

The process of making a tetris game in Mathematica

Note: so far it's just an intent, nothing has been coded, and tetris is only the best game idea that came up in the chatroom. I'm open to (and wishing for) suggestions, votes or requests

I am one of those guys that, even when I might have a good understanding of the Mathematica language, I can't read it fluently except when it's well documented, very modular, or too straightforward. In most cases, breaking it down takes some work. That's why I don't usually answer questions that require understanding a big chunk of code before knowing what the OP wants.

I once in a while go to the demonstrations project and see a few cool things. Then if I want to check out the code, I find huge Manipulates that put me off reading them. The closed question on the breakout game is no different.

All in all, I feel that GUI construction, stuff about the front end, and dynamic ineractivity, are one of the biggest weak spots of most of us in the community, and I'd like to see that change.

So I thought about doing a blog post on a similar game, showing the process of thought and creation of a tetris game, trying to make it helpful to those that find doing this kinds of things in Mathematica requires huge chunks of incomprehensible code, and keeping track of 73 things at the same time.

I don't consider myself an expert in the area either of dynamic interactivity or design in general, but I feel confident enough about being able to work this and write interesting stuff, even though I don't know exactly what it will be.

  • I just saw that this intent probably overlaps a little with the first @IstvanZachar's proposal... Humm – Rojo Jul 6 '12 at 0:39
  • Not at all: I think the dynamic interactivity leaves a LOT of room for experimentation and research, and game designing is something that was completely ignored here so far (that's why I did not vote for the closing of the breakout game). I would love to see your ideas/solutions in this field. We can even join forces. – István Zachar Jul 6 '12 at 10:28
3

The 10 most useful commands in Mathematica

If you had to choose 10 Mathematica commands to take on a deserted island, which would it be? Personally I had until recently strong feeling against WRI's "I am the language with most keyword" attitude (may be the beginning of Alzheimer!). I sort of changed slightly my mind, but still I think it would be useful to direct beginners towards the real building blocks of Mathematica.

To me it should probably include Map, but it would be great to post a blog on which people could vote and argue why they believe this or that function should enter the top ten.

We could have a top 100 as well or a top 10 in the new version of mathematica...

  • FWIW, if memory serves, Theo Gray made a list like this in one of his old books... OTOH, the version at the time was version 2.2! – J. M. is away Oct 17 '12 at 16:57
  • @J.M. right; but I would be interested by more recently added functions too! – chris Oct 17 '12 at 18:09
  • Oh, certainly; I just wanted to mention that there's precedent, so it would be nice to have one for the blog... – J. M. is away Oct 17 '12 at 18:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .