I mean we have Length, for the number of items at the first level of a list. Dimensions for the size of each level of a list, and ArrayDepth for the number of full lists that are nested.
@JM They may have (internally) aliased it to ArrayDepth[]
Other renaming mishaps: Trace[] used to be Debug[]. It ought to be fine, but when they made the matrix trace a built-in function, they now had to use the symbol Tr[] for it. Yech.
In retrospect, it makes sense to simply deprecate and then keep the symbols in the system. Any legacy code that exists still works. But I'd keep the deprecation warning on it honestly.
@JM The one advantage there is Tr(A) is used in mathematics.
Did you notice the "black out" on Wikipedia?
It's not a real black out because you can get around it by disabling JavaScript.
Also the Accumulate[] of version 1 functions way differently from today's Accumulate[]. So much for the claim of "old code will still more or less work in new versions."
@MikeBantegui Oh, it was a real big banner on there yesterday. SOPA's a really bad thing...
@MikeBantegui I suppose, it is after all consistent with Det[]... but then you have MatrixRank[], but just plain NullSpace[]...
Heh. Going from five to six was a bit of a shock for me, I'd say.
For a while I persisted on using five even when most other people were raving about six. It felt to me that they put everything including the kitchen sink in the new version.
...and thus it would bring my barely-managing computer to its knees.
I knew I wouldn't have the patience to fiddle with JS and all that, which is why I opened all 5-10 tabs of wiki articles that I would need for tomorrow. Now if only I can plan the rest of my stuff in advance...
@yoda There's NoScript, or if you've generally enabled JS then you might want to have a look at YesScript, which is pretty much the opposite (forbidding certain pages, allowing per default)
To be honest I can't imagine a closed beta with 100 active people can get over the 15 questions a day over a longer period, but I might be wrong since it's my first SXC beta
@David long story short, the first mma proposal was created when there was no community – just a couple of users here and there using it. Slowly we started growing and SO had a nice community built up and are interested in the proposal. Then they slam it shut for no reason and give silly reasons like "had the marks of early death" written on it, and all other bogus drivel. Users get enraged and start a new proposal to throw pie in SE's face
I suppose, but I've found that concepts used in a specific instance often turn out to be useful in something unrelated you're working on months later...
@David it's more like users will enter and ask something and not be turned away by a ghost town if there's prior content on the site. In addition, you also want them to like what they see, so there's the need for good content.
When reading through the documentation, you often encounter the phrases DownValues, UpValues, SubValues, and OwnValues. I am trying to understand the distinction between the four of them, and how, for a given Symbol, I can use them. Also, are they processed differently internally?
Well, "is there a better way to do X" could mean cleaner code or faster code or lucid code... as long as it doesn't mean shortest code, I think it should be ok
y'know, I had a question that I've been meaning to ask for a long time. I floated it in chat and Brett said that he wished mma had that functionality from MATLAB
It's basically the ability to control your number of output arguments.
I haven't used gallery, but I wouldn't be surprised if it had different output args based on the inputs used
@yoda Yeah, I was envisioning something like MatrixData["Clement"][6] would correspond to gallery('clement',6), for instance. And then you could also do something like MatrixData["Clement", "SingularValues"][6].
('cause for a lot of those matrices, they have explicit decompositions, eigenvalues, singular values...)
So, I don't know what's a better question: "how do I systematically write a curated data function?", or "can you make a curated data function out of a bunch of separate routines?"...
@JM umm... meh :) I dunno, it was something I used to think about at first, but I never do these days. MATLAB's "function" is more of the 'handy script I can reuse' kind, whereas I see mma's functions as closer to the mathematical concept. If I don't know what a function's input is, I'm now lost :)
I can generally slip in between the two modes of thinking pretty seamlessly, but there are days when I just blankout in one after heavy work in the other
I was meaning to write an answer, but I'll probably do later... but in short, we've come up with our own ad hoc conventions on SO
if it's a long block of code that shows the output at the end, then people generally leave it alone, but separate by a blank line.
If it's input and output interspersed like you would have in a testing environment, some comment out the output lines so that you can still copy-paste into your notebook
I tend to delete those lines if they don't add much value
Sometimes, the output itself is important (e.g., result of MatrixForm), in which case, I post a screenshot of the output
@Szabolcs enjoy :) but please make it a ticker feed. I'm sure we'll all be on top of everything for a few days and dropping them as chat posts will disrupt the flow
I used the code below (which is a sample from this gist containing more similar code) in my answer to my own question about Mandelbrot-like sets for functions other than the simple quadratic on Math.SE to generate this image:
cosineEscapeTime =
Compile[{{c, _Complex}},
Block[{z = c, n = 2...
is something I could answer immediately as I've solved the exact same thing (and this is the sort of thing I do with mma). by the time I woke up, the question was an hour old and had already two comprehensive answers...
@acl one of the criteria of beta success is the answer ratio - so if you have an answer that adds something to the discussion that isn't an edit of someone else's, then posting it is a good idea.
@acl I think it's okay build on the existing answers and post an improved version. Do that by all means! (Well if the improvement is tiny and you don't want to write up a whole answer, just comment ...)
@PrashantBhate This site is in private beta now, so for 1-2 weeks only the original committers can access it. For after that, we discussed this, and it seems this is discouraged by SE. It's not even certain that the site will not fail. Let me find the meta post for you\
@PrashantBhate In the beta phase, which lasts at least 3 months (but usually longer), they won't. After that it may happen for some questions, I heard old TeX questions still get migrated from SO to TeX.SE occasionally
@PrashantBhate If a beta site fails and it gets closed, all question from it will be lost. So they prefer not to migrate anything until the site is out of beta.
everyone is out of votes and the question scores look pathetic compared to yesterday. Once the UTC day rolls over, everyone's going to be rushing for the next wave of upvotes :)
That's the best kind. Post something and wake up to see upvotes... unlike the unfortunate ones who happen to be in an SE traffic timezone, and sit and continuously hit refresh to see where the score is ;)
Each site gets a twitter account that will tweet posts from the site. We're running into Twitter's length restraints though. We have 10 characters to work with. Mathematica is 11. |:
@RebeccaChernoff When do you think we should think about the logo? In a couple of months only? Mathematica's logo was made with Mathematica itself, and it'd be nice to make something non-identical (so we don't step on trademarks and don't have something Mathematica-version specific) but easily recognizable as Mathematica related. Perhaps have a logo contest on meta. I'm sure people would love it. See also mathworld.wolfram.com/Spikey.html
@RebeccaChernoff Agreed. It has been a good first day, I think...
@Szabolcs ...and it's dual to the dodecahedron. So, either of those two could be a motif, if you still want something vaguely recognizable as related...
We are already working on tagging (question 3), and have had discussions in the past on SO about on/off topic (question 1). Also, there's a question today on meta that is perfect for the FAQ (question 2). So, I think we're making a good start.
I'm not completely sure ... I have seen ' used as a substitute more than once, and I've definite seen more than one broken comment on SO, but couldn't find the latter now.
I remember searching for this myself, and it was not so easy to find. Now we have an answer on meta but I thought it'd be good to add this info to a place where people are the most likely to look for it (i.e FAQ)
@RebeccaChernoff Whenever we make a piece of text containing [ or ] into a link, they get escaped --- even if they are in code. These characters are very common in Mathematica code. Please see here: mathematica.stackexchange.com/a/175/12 Is SE aware of this issue? Of course it can be fixed manually, but it'd be nice to make the editor smarter and not make this mistake, if possible.
@RebeccaChernoff I fixed it up, but you can see the edit history.
There was a suggestion in the piratepad discussion to have a "core tag", which applies to the whole area, and more specific tags. The more specific ones were suggested to be dynamic, manipulate, demonstrations, by Brett. I agree that this categorization will be useful. Now should we just use dynamic as the core tag too, or there should be dynamic-interactivity? I like the latter better, but I think people will keep using dynamic anyway, ...
so it's maybe a better decision to just go with dynamic
I tagged my own question only with dynamic, even though Dynamic[] doesn't appear anywhere in it (but it is about dynamic interactivity)
@rcollyer It'd be good to have a meta post about several tags maybe? Anyway, can you post about this single one,linking to the read-only version of the piratepad where Brett has commented?
@yoda and everyone else. I've edited this question a bit now. I don't know if these questions should be welcome or not, but I'd personally find it useful to allow them. What do you think?
Note: this is a somewhat subjective question asking about "best-practices" and what different ways are there to achieve this. I already have a solution which satisfies most requirements I have, but it would be tremendously helpful to at least hear from others who have made palettes in the past. ...
The fact is, we don't share nearly as much code as "normal" programmers, many of us don't know what's idiomatic. Just a few people telling me how they did something similar in the past would be very helpful.
When reading through the documentation, you often encounter the phrases DownValues, UpValues, SubValues, and OwnValues. I am trying to understand the distinction between the four of them, and how, for a given Symbol, I can use them. Also, are they processed differently internally?
I used the code below (which is a sample from this gist containing more similar code) in my answer to my own question about Mandelbrot-like sets for functions other than the simple quadratic on Math.SE to generate this image:
cosineEscapeTime =
Compile[{{c, _Complex}},
Block[{z = c, n = 2...
