@RolfMertig One of the main reasons I dislike ListConvolve for such signal processing applications is that you need to be explicit with all options to get the actual convolution process correctly. So an example with padded fft that demonstrates the convolution theorem would be:
@J.M. I'm aware of Goertzel's algorithm, but haven't seen that reference in particular... alas, I haven't yet done any of what I wrote in my to-do list in that answer :D
Just something I want to note. :) Goertzel can behave badly sometimes, and it's not the first thing people look at when their results look wonky... so for safety, use Reinsch's version.
@R.M I wasn't teasing you. ;) I saw the parts above where there was railing against For[].
@R.M Precisely; "use the proper tool for the job". I wouldn't use Do[] if the increments are not varying linearly, for instance...
(On the other hand, I'm not entirely immune to writing clunky code; I use While[True, (* stuff *) If[test, Break[]]] to fake an until loop whenever I need one...)
Well, I almost never use for in MATLAB either these days (in cases where it cannot be vectorized), instead resorting to more "functional" ways of doing the same
It's potentially useful in compiled code, given that one can't Return[val, fun] inside the VM. Daniel also claims to have productively used Goto; he seems to get more mileage than most out of the procedural constructs.
I don't recall that I've ever needed them, though. On the other hand, I never used a For loop in anger either.
I'm always impressed by how much Daniel and Leonid get out of procedural code... I've sort of forced myself to avoid that line of thinking when it comes to mma, so I'm routinely reminded to not be close minded.
@R.M I find the paradigm rather clumsy and hard to understand, but it has an undeniable utility in compiled code, which IIRC is mainly why Leonid uses it. I'm perpetually amazed by Daniel's procedural masterpieces, though.
Many people think they can come to a Q&A site and ask "How should I calculate my thesis' integral". And they think a bunch of idiots will work a week in solving his problem
Luís Antoni Santaló Sors (October 9, 1911 – November 22, 2001) was a Spanish mathematician.
He graduated from the University of Madrid and he studied at the University of Hamburg, where he received his Ph.D. in 1936. His advisor was Wilhelm Blaschke. Because of the Spanish Civil War, he moved to Argentina where he became a very famous mathematician.
He studied integral geometry and many other topics of mathematics and science.
He worked as a teacher in the National University of the Littoral, National University of La Plata and University of Buenos Aires.
Works
Luis Santaló published...
who was talking about the history of the Gabriel's horn problem
Edited now that the site has graduated: How do you feel about holding a Mathematica mini-contest? It seems to me that this is a great fit to our community’s overall “we find plenty of ways to do stuff” spirit, and it could help publicize our site in an effective manner.
I'd be willing to step up...
we don't have much meta activity, and I'd think this could be a great way of publicizing the website in a good way (by which I mean, not attracting more “what's wrong with this code?” questions)
@SjoerdC.deVries thanks… I wanted to get the general feeling about the idea itself, rather than propose something too specific. But I'll try to add a little
@R.M Thanx! And performance-wise, which function is better in your experience? Or is ListConvolve using Fourier and InverseFourier anyway under the hood?
@J.M. Recently I had to implement a somewhat tricky Monte-Carlo algorithm and the most readable (and closest to the paper) way of implementation in M was to use Label and Goto. It was the first time I saw no obvious alternative. In 22 years of Mathematica programming ...
@J.M. Regarding For loops: Obviously Do is more M-like, but if you have a long loop, I find it easier to read at the beginning of For[ ... from where to where to loop runs. So it all depends and I think readability (also for non-M-freaks) should be more emphasized, in general.
@RolfMertig Hmm, yes, that's another defensible use of For[], but the first thing I'd consider is if I can split off sections of the body as separate routines... basically, For[] only if "back against the wall". :)
@J.M. But there is only one person on earth who could answer that question and even change it. But he will not. So, no, let us not ask questions about language design (like why is a comment not a Mathematica expression, or why is there no useful non-commutative algebra built-in to Mathematica, etc. )
Today I asked for (and was granted) migration for a question from SO to Mathematica.SE
When the migration was done I lost some rep on SO, and as I've already capped at the destination site that rep was not gained there. Nothing serious, just a few points.
But I wonder if the same could happ...
@GraceNote Hi! We have a nominee for the oncoming election, but he lacks the experience for the job. I wonder if the nomination process does not filter candidates in some way. I don't like the idea of him being hurt. His intentions are (of course) good.
@belisarius Yes, I noticed that there's so far only one nominee, which is actually more concerning to me than this one's lack of experience
@R.M The site is too small and young for us to set the same kind of barriers we put on Stack Overflow. Which we also do because they tend to get over thirty applicants.
@belisarius It is participative, which is why I expected at least murmurs. Maybe I've been missing the murmurs, then
We were internally thinking "Maybe people are just working on their nomination speeches", but for the most part it came across as silence, which is eerie to say the least
No, no murmurs. We really don't need "speeches". There are at least 20 users we know are able to do the job quite well. But evrybody considers it a job, not a privilege, as it happens sometimes in other sites
So, people are thinking about running, though. Heavy emphasis on the thinking, but they are thinking
Back to the current nominee - I think Verbelia's approach in the comments was a decent way to start it - if you have concerns about one's experience, it's better to let them know and help them shape up. If this user seriously considers this an option, then it would be wiser to help them understand what they need to do to be considered by the community as an option, than to cut him off by imposing a mechanical restriction
@GraceNote I, for one, haven't stepped in yet for a number of reasons. First, having had the privilege for the first half year I wouldn't stand in the way of ambitious, new mods. This is a nice gang we have and if someone else wants to try, by all means, go!
Second, the workload is not so high as on SO for instance, but the pressure you feel to spend at least an hour per day on the site, spread over various check-ins is having an effect. Mathematica.SE is addictive and being a mod doesn't allow you to step back once in a while although it would oftentimes be advisable to do so career- and family-wise. So, I was thinking to let my own nomination depend on what else happens during the nomination phase.
I won't let this fantastic site go down, and will step in for sure if need be, but I' m not sure I' ll do it if there are already sufficiently qualified candidates. My guess (and fear) is that many potentials are reasoning along this line.
I think most regular users (aka the community) will serisouly consider nominating… I don't know this community as much as the R.M, Sjoerd or belisarius (I'm not a regular on chat, and I have long "lurking only" periods), but it is a wonderful community
it may be, however, that compared to other parts of SE, it is frequented by more busy/professional/otherwise-occupied types of users, which would lead to a higher barrier to people nominating
True. We have to take care not to burn certain people. Szabolcs for instance, will be back in a couple of days/weeks, but he has been doing so much in the past it became almost unhealthy. His voluntary suspension springs to mind. Better not ask him, or we risk he agrees.
@F'x Sorry, was on the phone. I can post the mod statistics, but this might not be advocated. I can say we get on average 1 or 2 flags per days, but we edit and comment like hell, roughly 1300 edits among us three and 3000 comments,
@F'x I would love to have a team of 6 mods, not sure whether that's feasible. But I can't really see how that could be a problem.
Now that the site has graduated, it is attracting a lot more attention from newcomers to Mathematica, as well as frequently receiving migrations from StackOverflow. I suppose I don't have to point out to anyone that some of these questions provide limited evidence of sincere effort on the part of...
@SjoerdC.deVries there have been arguments against it (and it's not used on many SE sites)… one of them is that you hope that the Q&A site itself becomes a standard reference at some point, which doesn't happen if you don't allow these questions
Chemistry.SE has taken another path: some of these are “not real questions” (it's not a question if you can't formulate it coherently and don't even know where to begin looking), the others should be answered once, and later closed as duplicates
A simple option, if "Not a real question" doesn't fit from whatever vagueries are present, is to solidly define what realm of questions you'd say don't belong, and make them off-topic.
@GraceNote I don't like that one much… if a newcomer asks “What is the Mathematica command to make a 3D plot?” on Mathematica SE and you tell them “This question is off-topic”, they will most likely not understand it
@F'x Will they really understand it any further if we have a close reason that says "You could find the answer elsewhere, why are you checking on Mathematica.SE?"?
@GraceNote There are some questions which are technically RTFM, but the necessary detail is in an almost hidden corner of the manual that the question is worth having as an entry point. Then there are others that are plain lazy and the answer can be found by simply typing two words from the title into the help menu (e.g. "How do I total elements in a list")... The latter one is what we don't want, but none of the close reasons fit it well.
We've been using TL to close these, but a "general reference" reason would be better.
@F'x It's not quite escalation, more of a side pass
Mostly because we had just discussed this hours ago and he had some very nice ways of phrasing it that I didn't write down word-for-word so I can't quite replicate it
@R.M I suspect you're thinking of the 'Experts Live' events. The previous Virtual Conference was larger. (And, for one of the sessions, involved me coming into the office early in the morning, then going home and getting some more sleep.)
Haha. :) I was just digesting Leonid's latest epic... if it's true that data structures make the program, and facility with them, the programmer... well, I'll have to study hard if I want to become half as good a programmer as he is.
His answer to the boggle question was what convinced me that data structures play a more important role (in larger applications) than programming smartly/neatly/cleverly/etc. A combination of both can be deadly