@DavidZaslavsky for the longest time, postfix notation was my preferred form, and it still largely is. I treated it as a succession of transformations of the initial data.
I would imagine there's some penalty for having launched, so that the front page is taken up mostly by upcoming proposals (that actually need the support) rather than existing sites
One of the advantages of the Mathematica proposal was our ability to regularly communicate within the StackOverflow Mathematica chatroom. This allowed us to build and maintain a core group of users. It is this core group, that I believe helped us quickly pass through to beta after the original pr...
@Verbeia Hmm, allow me to rephrase: were the 15th hottest proposal out there, which puts us back on the first page with the minimum 15 proposals per page.
Leonid on programming is better. Nearly 100, but only 9 answers. So, it will be a little while for him. I think Mr. Wizard has a better shot at list-manipulation.
And, despite having 45 questions, front-end's highest rated user only has 50 upvotes.
He's admitting to forcing 500 to accept Heike's answer because earlier in the day, she happened to lose an accepted answer when he unknowingly wrote the same duplicate answer... and also because he believes that Heike used his answer from SO, which cannot be beat
And this is after just 2 hours of the question being posted
The 2 hours is disturbing to me. The rest, just indicates that he has a conscience. (The first time I spelled that word without spellcheck going crazy.)
Hmm... never mind. I read his comments in the moderator chatroom. He probably just didn't realize the gravity of his words... especially with the diamond (which takes a bit of getting used to)
@yoda holy cow I had no idea this was going to be taken that way. I hope I don't regret volunteering to be a moderator sooner rather than later. Yoda, I cannot make you or anyone else believe me but flatly, you have it all wrong. None of what you described motivated me. Most importantly I was not mindful of the diamond next to my name and I did not mean anything official in what I wrote.
I also did not mean to imply that because Heike had used a method that I was also fond of, which I most certainly did not invent, that somehow the answer could not be beat. In fact I specifically allowed for the probability that it could be beat if Leonid posted.
If all my comments are going to be put to this kind of scrutiny and suspicion I want no part in moderation.
In the past when I have had similar exchanges no one complained, and as far as I know the people involved were at least cordial to me. I am quite shocked that this has changed so suddenly.
It may not entirely be due to moderation... IMO any time someone writes "please accept <specific answer>" it's likely to rub some people the wrong way.
I don't know if it is the fact that I have been made a moderator that is bothering people, or if the fact that I am a moderator is simply causing people to express how they already felt. One way or another I want to, need to understand this.
I can't really say anything about this specific incident, having not been involved in it, but FWIW everybody has some hiccups when starting out as a moderator. I've seen moderators do worse things and get over it.
@DavidZ At the time I wrote that there were only two answers, mine and Heike's. I meant it only graciously, I had no ulterior motive, and I in no way was trying to force his hand.
I am quite certain that I have previously done similar on StackOverflow and that nothing bad that I was made aware of resulted.
I certainly would appreciate any guidance you can give me in this situation; I feel blindsided.
Yeah, I personally understand where you're coming from, and I wouldn't have complained. But that means I can't speak for what the people who were complaining were thinking at the time, other than what has been posted in comments.
@MrWizard I never took it that way until you confirmed in the comments that indeed that was your intention... In any case, what really took me by surprise was the fact that you were the one who regularly commented to new users (and especially to 500) that they ought to wait for 2 days to accept an answer so as to not discourage others from attempting to answer
I will not deny that I was very surprised to see you rushing to convince him to accept an answer...
@MrWizard You know that is not true because we all overwhelmingly supported you and all the other nominees
I'm absolutely positive we've got the best mod team... I was (and as you can see Sjoerd from the TL comment, rcollyer, JM and I don't know who else) surprised because this was exactly the polar opposite of the Mr.Wizard I've known
@yoda thank you for discussing this. regarding "You know that is not true" that certainly would seem to be the case which is why I feel all the more blindsided. I would not have recommended he accept that answer if he had not already accepted (my version of) that answer for the nearly-identical previous question.
By the way R.M has flagged the comments under that answer for moderator cleanup but I am scared to take action myself lest it somehow be seen as evil.
@MrWizard What I would advise you to remember is that in the long term this is not that big a deal as long as you learn from it. New mods always make mistakes and they are easily forgiven with time. As yoda said, you have a lot of support from the community, perhaps more so than the moderators on the average beta site.
@MrWizard To echo what David said, this is a site that grew from an existing community and we all trust you guys, a hell lot more than on an average site, where the mods and the users are only getting to know each other. Hell, you, Sjoerd and JM go a long way back on MathGroup too, so it even precedes SO by a good 10 years! I (and I believe everyone else too) fully trust all 3 of you
My concern was mainly because I felt that you were making erratic comments because of the events that played out earlier in the day, re: the PadRight answer. For no fault of yours, you were put in a position that is not enviable.
I felt that you were trying to overcompensate because you felt bad about one incident... I guess the larger point I was trying to make was that it's life... sometimes people make mistakes like accidentally writing a near duplicate answer (it has happened to me several times) and sometimes, the OP ends up accepting that
I am not used to my comments (often tongue in cheek) being connected to authority. If nothing else I have already learned that I will have to conduct myself differently (which is sad, as I have enjoyed being able to take a lighthearted approach to SE).
@MrWizard If my experience is any indication, with time, you and the community come to some sort of equilibrium, and you'll get to know when it's appropriate to make jokes (e.g. in TL) and when it's not.
@MrWizard Let me just be clear that I would've brought this up with you even if you were not a moderator... it was not the comment in isolation that was the concern; it was the comment seen in light of earlier events (especially since you took it to heart) that was concerning. Again the concern was not that you're abusing your powers (you're not!) but rather that you might be taking things a bit too seriously...
There might be legitimate reasons to request that an OP change answer, but it always ends up grating someone the wrong way. Recently, I did ask someone to change their answer to a more well rounded answer with examples (and I did not have any stake in the game... strictly a 3rd party), and I don't think it came off very well to the other person...
@yoda Do you see the pinned message about the syntax highlighter on the right? I don't see it here on the right. When I search for it, the system says it's already pinned.
Click here to install and test @halirutan's syntax highlighter on Mma.SE. With Firefox, you need to install Greasemonkey first, with Chrome it works out of the box. Please test, this will eventually become part of the site! Problem reports go here.
gaaah! It's being stubborn. While I figure it out, we'll have this:
Click here to install and test @halirutan's syntax highlighter on Mma.SE. With Firefox, you need to install Greasemonkey first, with Chrome it works out of the box. Please test, this will eventually become part of the site! Problem reports go here
@yoda I figured that since font attributes are really a front end affair I should let the front end figure out what the right value was instead of the kernel by wrapping it in Dynamic.
@MrWizard without wanting to enter this whole mess (I personally did not find your comments particularly inappropriate, but I can see why yoda got worried), let me encourage you to experiment with Compile in v7. It is actually trivial to get something much faster than what Heike and you did if you use imperative code and compile (as you can see in the other answers), even if you do not compile to C.
@MrWizard indeed, but if what you need is speed, it is often possible to avoid coding C by using compile. It's not an alternative to mathematica but an alternative to explicit C
A quickie for you guys: nothing more compact than f @@@ Partition[{a, b, c, d}, 2, 1] for producing {f[a, b], f[b, c], f[c, d]}, no? (No, I don't want to use Developer`PartitionMap[].)
What's a simple way to evaluate to do the following for generic a f[x]? Let f[x_]:=Sin[x] and x=1. I would like to start with HoldForm[f[x]] and end up with HoldForm[Sin[x]]
It needs to work for any f[x], and the variable needs to be x. It has to work even if x has a value.
It could be that the initial excitement is sort of over, most active people may be a little burned out, and generally this may be a start of the next phase, perhaps going to a steady-state. This has to happen at some point anyways, IMO. But it is hard to judge by only a couple of days, particularly weekend.
mathematica.stackexchange.com/questions/1643/… <---- I wonder why this got a downvote. True, it is a duplicate, and it needed to be closed, but it is a reasonable question. Downvotes and close votes shouldn't be mixed.
@halirutan Here's a somewhat unusual example where both default patterns and named patterns are used. It is highlighted perfectly. mathematica.stackexchange.com/a/1572/12
I know I am going to need 8 point type in the majority of figures. SetOptions[Graphics, BaseStyle -> {FontSize -> 8}] does not seem to affect any plotting functions such as Plot. Do I really need to either SetOptions on all plotting functions or give an explicit BaseStyle for all my plots?
@halirutan Sounds great! Did you say you wrote him but he didn't reply in email yet? If that's the case perhaps he's not reading the email address you wrote to.
@Szabolcs But we already communicated over this address. But have to admin, my second email was quite long (Leonid style ;-) and I tend sometimes too, to ignore looong emais a while.
@JM I think this came up before, I should do my homework and research it ... but right now it's simpler to just copy and paste BaseStyle everywhere, as ugly as that is
@Szabolcs Well, the scratch array is what stands for the two-dimensional array in the triangular recursion.
If you look at the general formula, you have three quantities that always form a triangle in the array. The clever idea is that instead of using an explicit two-dimensional array like the one depicted, you have a one-dimensional array that is gradually overwritten by "lower left-upper right" diagonals as the algorithm proceeds.
@MrWizard I've been thinking about those algorithms, on and off, for the past five years. A good solution would be quite useful, as you might surmise from the examples I gave.
@JM Well, "functional" means that we transform the whole expression in one go instead of changing elements of a mutable array one by one. We could try to write the whole thing in a way that it could be directly translated to something that doesn't require re-allocation of arrays, but I don't think it's worth it since we don't know what Mathematica does in the background
for example, does f /@ list re-allocate list, or does it replace it with the result?
if list is not used anywhere else, it could just be smart an replace instead of make a new list, copy results, then destroy the old list
But I don't know if Mathematica is smart enough to do that, and when using this style of programming I guess we don't have much control over whether the system/compiler will do that or not
@JM Check out the code I use in creating a RationalInterpolation via Baker. It isn't the speediest thing in the world, and I don't recommend it, in general, but it isn't a bad starting point.
@JM If you show me that recursion formula with the nice triangular table, I tend to write something like the Nest in my answer when using Mathematica. If you show me the code only, I'd tend to write Do loops as well.
What I find particularly interesting is that if you look at the RationalInterpolation` package, you find that they do it via a least squares fit. So, I think that logic may be extendable to more general recursive forms.
@JM No, you misunderstood me. I was just trying to explain how I think. When I see something similar to that general recursion formula you included, I am likely to write Nest.
@JM Definitely seen it. The algorithm in Baker comes from Thiel, but covers very similar ground. The method in the package is somewhat different, though.
If you delete the checks for zero denominator, you have essentially a compact way to produce Padé approximants in wynnEpsilon[]. (Reminds me, I should write up something sometime...)
@JM A colleague sent me his code for an approximation I was attempting to use (won't finish it in time to graduate :( ), but it includes some notes on least squares fitting for Pade. Would you like a copy of the notes?
@rcollyer Circulant matrices are a very special kind of Toeplitz matrix. I do remember research in FFT methods for general Toeplitz systems, but they always have one little trip up or two.
My membership in SIAM lapsed, apparently I've spent to long in grad school to continue to qualify for reduced cost. I'll have to rejoin just for the SIAM Review.
I have ~6-8 paper boxes full of papers from my office for my research. So, there's an argument to be made for online, too. But, nothing beats paper for note taking, like outlining a particularly odd line and writing "wtf?" on the page.
@yoda Well, rcollyer brought up Padé approximants and rational interpolation. I said that the recursions for those are "lozenge recursions" (four terms) instead of "triangular recursions" (three terms).
...and since you can also get the coefficients of a Padé approximant by solving a Toeplitz system... yeah, the math is intricate, I have to agree. :D
...anyway, I will have to look at rcollyer's code; if anything, I might be able to adapt it for a possibly better implementation of Wynn $\epsilon$.
I might sound like I've already done a lot with these things, but I'm sure there's still a lot of hidden chestnuts in those... that I still don't have an effing clue.
@Szabolcs: now that I think about it, your method would indeed quickly apply to Akiyama-Tanigawa and de Casteljau. Most of the other triangular recursions would indeed be slightly problematic, though, since the multipliers depend on both loop indices.
What is the preferred way to write the name "LaTeX" on this site? A few questions use $\LaTeX$ in order to get the LaTeX name; however that looks wrong to me because it makes the letters italics, which isn't the case with the "proper" typeset name. You can get that name with $\mathrm{\LaTeX}$ whi...
