Wolfram Mathematica - 2012-02-24

« first day (37 days earlier) ← previous day next day → last day (4430 days later) »

12:14 AM

@rcollyer I'll look for it. Are you familiar with the Eisbock?

@Rojo that's really cool; I put a star by it to draw some attention.

1 hour later…

1:43 AM

@MrWizard I was thinking it myself for exactly the same reason, but didn't get a chance to do so.

@MrWizard No, I'm not.

2:44 AM

@MrWizard might I ask why this question was closed and not migrated to Mathematics?

@rcollyer Don't know what reasons Mr.Wiz had, but if I may offer an opinion, I think it's good decision and in general, a good policy to follow.

Unregistered user + new user + basic question which is most likely a dupe on math.se... better close here and direct him there

I can see the reasoning, but I'm not sure it sits well with me. (And, I don't know why it doesn't.)

Migrations were initially meant for good questions that happened to be in the wrong place or to preserve answers that were already given... now it has become a shipping lane for crappy questions (not saying this one is) and more often than not, such questions are dupes. Instead of unnecessarily pissing off other site mods by migrating such dupes, better point these new users there.. then it's entirely their problem as to how they deal with the user and the question.

That reasoning I can live with.

I would actually support a migrate nothing out of here unless it is stellar policy... but that's upto the mods anyway, since they're the only ones that can migrate to other sites

2 hours later…

4:52 AM

Apparently, mma is just not good enough for him, so he's looking for Ruby equivalents, again. :P

5:32 AM

0

Q: Are questions about doing symbolic math in Wolfram Alpha on topic here?

Sometimes, users ask questions on Computational Science about doing something in Wolfram Alpha. As far as I can tell, Wolfram Alpha uses Mathematica for its math engine. Are those questions on topic here?

5 hours later…

10:26 AM

Does anyone know how to prevent a FlipView from flickering on flip? I can use a TabView as a workaround, but then I need to aim on tiny tabs with the mouse every time I flip.

@Szabolcs Can you give an example? I don't see any flickering.

@Heike It turns out it's because the expression I'm showing takes a long time to render. Rasterizing it made it much better. The difference between TabView and FlipView is that the former shows the old image until the new one is rendered while the latter shows a white background.

@Verbeia Do you have any new statistics for last week? It seems the visit count has plateaued and the question rate has stabilized as well.

11:04 AM

@Szabolcs Feel free to edit or add stuff: meta.mathematica.stackexchange.com/q/268/187

0

Q: Including google-code-prettify language extension for Mathematica into this site

As most of the users may know, we developed a language extension for google-code-prettify to support the correct highlighting of Mathematica code on mathematica.stackexchange.com. Currently, this extension can be tested by installing it as local JavaScript-plugin into your browser. A detailed dis...

feature-request

@halirutan passed it to a dev

@yoda thx

11:47 AM

@Szabolcs sorry, haven't been near the computer much

user image

user image

user image

user image

user image

user image

I think the question rate is starting to turn up in the last day or two.

You're right that visits have been bumping around between 670 and 710. But we have more and more avid users, and more than a third of people who have ever registered have visited in the last couple of days.

The biggest source of incoming users might be SE's "hot questions" list. The people who come from there might only be curious and may never come back.

@Szabolcs maybe, but they wouldn't register

There was a rather badly formulated question on MathGroup the other day about performance optimization:

@Szabolcs I think that accounts for the slight bump of a few days ago in the visits/day graph

Oh, about MathGroup, I wanted to say that I replied with lots of links to here (as well as elsewhere), but it didn't show up. It seems it did:

groups.google.com/d/topic/comp.soft-sys.math.mathematica/…

11:56 AM

Of the 584 users in the stats, {{"moderator", 19}, {"registered", 546}, {"unregistered", 19}}

Unregistered users only count if they have asked a question etc

I always lose posts on MathGroup. The lack of proper threading is really annoying (and pretty much impossible to solve in a coupled newgroups / mailing-list)

@Szabolcs That's one of the reasons I unsubscribed from the mailing list

Maybe this message was overkill, but my point is that I think it is worth cross-referencing MG and Mma.SE. It'll help Google.

Every post is prefixed with a unique id so it's impossible to sort them in a mailing box

When a link here obviously helps, it's worth including it in the MG reply too, like here (I already replied, summarized a solution, and put a link to Mma.SE's version of the question)

12:03 PM

user image

Here's one I haven't done before, a "barcode" of when users were last seen. I'm using FillingStyle -> Directive[AbsoluteThickness[0.4], Opacity[0.3]]. Notice the dark strip in the past few hours. People are coming back.

Anyway it's bedtime here. Good night !

12:20 PM

Good night!

2 hours later…

2:37 PM

@Szabolcs Thanks, but your answer was pretty good

There's a related question here

8

Q: Mathematica: Matlab like figure plot

How do you obtain a figure similar to Matlab's figures? That is, how can you make Mathematica have a white background inside the axes and some other color outside the axes. See the following figure: Here is an attempt to replicate the looks of a Matlab figure Framed@Plot[Sin[x], {x, 1, 10}, ...

mathematica plot

3:02 PM

@Eli My answer was "leaky" (colour leaking outside the frame) and does not add anything to yours. If I were reading the question as an outsider, I'd prefer to have only what you wrote and no clutter. So I deleted mine :-) Above 4000 I don't care about rep score any more (at least not while we're still in beta).

And even if I did, in this case deleting would be what benefitted the site and community more.

@rcollyer I am following yoda's logic and experience on this. Maybe not quite to the same degree, as I think it is appropriate to migrate a question even it is not "stellar."

Good day, Szabolcs.

3:15 PM

@MrWizard I was just curious, it seemed like it should have been migrated, instead of closed. Since it wasn't, I was wondering about the reasoning behind that choice for my own edification. @yoda gave a sound explanation, and I'm duly ejimicated now. :)

@Verbeia thanks for the stats. The slight uptick in ?'s/day is nice to see alongside with a growth in avid users.

3:35 PM

@MrWizard Do people still often use "good day"? I have never lived in an English speaking country so I can't really tell apart nuances.

1

Q: Numerics with mathematica

From time to time, I would like to use Mathematica purely numerically, e.g., plotting a function which is defined as an integral which cannot be solve numerically or a solution of a differential equation, or ... . It turns out that mathematica is rather efficient (I would say comparable to matlab...

He is right that one problem with Mathematica is that one needs to learn a lot to be able to use if effectively.

People are sometimes amazed when I do "magic" with Mathematica, but when they try it themselves they often get demotivated after finding that they can't immediately work effectively with it.

Is Mathematica like Chinese characters?

3:54 PM

@Szabolcs A former office-mate of mine is Chinese, and he made an interesting point about Chinese characters: it is more like a programming language than English (or, other western tongue) is. A character in Chinese conveys semantic meaning, not pronunciation, while the characters in the Latin, Cyrillic, or Greek alphabets carry pronunciation, not semantic meaning. So, walking through Tokyo airport, he could read the signs, but could not pronounce them.

In imperial China, this worked out well, as the government could send out proclamations, and everyone could read them despite have many regional dialects. This would not have been possible in a western style language.

@Szabolcs Have to remember to use that ...

@rcollyer Please don't ever say "Western style language" ;-) this is a property of the writing system and not the language itself. Also, it is not really true that a Chinese character conveys meaning, actually it's a lot more complex than that. If you're interested, look up the book of John DeFrancis, especially this chapter

@Szabolcs allow me to rephrase: western style writing system. Better? :)

W Hannas's books have a lot of very interesting information as well, but I don't agree with his main point (i.e. that the writing system inhibits creativity). I just think learning Chinese characters is a waste of time.

The reason I'm learning them is that currently that's the only viable approach to learning Chinese.

@Szabolcs Which writing system inhibits creativity?

@Szabolcs why learn Chinese then?

4:07 PM

@Heike The Chinese one, according to some. As I said I disagree with that point, but apart from that the books are an interesting read

@Szabolcs Probably not "often" but it's heard, and I use it.

@Szabolcs I don't think it bypasses "the restrictive intermediary of speech," just that their meanings remain relatively standard, so that someone who knows one dialect can read a message written by someone who knows another dialect. Of course, both would have to stick to a standard set of symbols, or the usefulness is lost.

@yoda I have my reasons but it is definitely not the "beauty" of Chinese characters. Those are what kept me from starting for a long time. I used to have lots of Chinese colleagues, so it seemed like a better choice than Greek (which I could only practice in writing). Also, I find the language quite interesting.

I knew a guy who learnt it because he thought they would take over the world (at least, economically) in 20 years... he didn't want to be jobless then

We could always go the FireFly route, and keep both English and Chinese.

4:11 PM

@rcollyer If you read the two chapters I linked to, you'll see how that is not completely true. It is a common misconception that Chinese characters represent meanings. They are in fact very much tied to one language (or one dialect of Chinese if we prefer the term "dialect"). Of course it's true that someone who knows Chinese can guess a lot about a Japanese text, but they can't read it completely and one of the reasons it's intelligbile is the huge influence of CHinese on Japanese.

I know someone who learnt Chinese because he was invited to give a lecture there.

@Heike That would be a very good reason for me :-)

Part of his training was to put stickers with the Chinese word on all the object in his house.

@Szabolcs my info comes second hand from a Chinese office-mate. So, I'm unable to debate the details.

@Heike That really works... I learnt Russian for a while and sticking post-its on stuff all over the house helped

4:16 PM

@rcollyer Let's don't, just look at the DeFrancis chapter, it's really interesting. BTW I found that it is often quite difficult to get accurate information from Chinese people about this.

I had trouble getting around the fact that in cursive Russian (and some in typeset too), m is T and y is u, c is s and so on...

@yoda Maybe I should try that with French or German words. I've had several years of both in school but I've forgotten most of it.

I used to collect stamps as a kid... suddenly, CCCP made a whole lot more sense after learning Russian.

@yoda To be fair, another reason could be women ;-) but I started learning before that.

@Szabolcs aha ;) Well, learning before certainly helps :)

4:17 PM

@Szabolcs Never a bad reason to learn something, but it could inhibit the process.

@Szabolcs Plan on reading it. Thanks.

That is true. I was a lot more diligent in my attempts to learn Greek

my Chinese is messy ... and broken.

heh, that's a good excuse to spend more time together trying to fix your language ;)

@Szabolcs Hmm, reminds me of my high school spanish final ... I rushed through it as my girlfriend got done before me ...

I passed. But, barely.

I know exactly one Chinese Mandarin word and that's due to an open university course that was broadcast on national television many years ago.

Which one?

4:21 PM

Hello

你好 - nihao?

I should get something done today. So, later all.

user image

Yes, I as well.

That's for you @Szabolcs

4:21 PM

shi shi shi :D

oh, I take it you've seen it :D

@Szabolcs Yes that's the one. That was the title of the course.

knee how?

yes, but with the right tones

@yoda But I only know four characters from there, and I can guess which one is the lion (as it has the animal radical).

Lion-Eating Poet in the Stone Den

The Lion-Eating Poet in the Stone Den () is a 92-character modern poem written in Classical Chinese by Yuen Ren Chao (1892–1982), in which every syllable has the sound shi (in different tones) when read in modern Mandarin Chinese. It is a famous example of constrained writing. The sentence "Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo" is a well-known example of this type of writing in English. The text, although written in Classical Chinese, can be easily comprehended by most educated readers. However, changes in pronunciation over 2,500 years resulted in a large d...

4:26 PM

Another problem for me is that I am not good with languages, so I have a somewhat unusual approach to learning: understand grammar rules and focus on writing. Then theoretically one can use the grammar rules in conjunction with a dictionary to build sentences and write a few emails. Theoretically. In practice it does not work. But it is a very good exercise.

With Chinese I can't really practice by reading/writing because the writing system is more difficult than the spoken language. And the grammar is rather minimal and de-emphasized in teaching materials (even though there are some very important points)

So it was like the most difficult possible choice. :-)

People who have a talent to learn languages can pick them up really quickly by listening and trying to speak. That is my weakest point.

@Szabolcs So, basically you study languages like a computer language...

@Szabolcs I have the same problem. I remember stuff by deriving things myself, not from learning by heart. This works great for mathematics, but is less useful for languages.

personally, i'm terrible at languages. in highschool i was learning two extra languages at once and would always manage to piss off both teachers by mixing the two together

seems my brain operates in "english" and "foreign language" mode

I had a French teacher who studied Italian as well as French at university. Years later she decided to learn Romanian, but apparently Italian and Romanian were so similar that she kept mixing them up. In the end she could speak neither Italian nor Romanian any more.

@EliLansey Not really, doing that is just the starting step. Then I write simple texts and get people to correct them for me so I can "get" the idiomatic usage

4:36 PM

@Heike but... in most languages other than English, words and meanings can be derived if one knows the basics

@EliLansey I used to terribly mess up the English word "weird" while I was learning German... in German, ee sound is written ie, and I kept writing the word as "wierd"

@yoda But some things can't be derived such as tenses of irregular verbs

@yoda hell, i write the word as wierd half the time anyway, and i'm a native english speaker :)

@EliLansey lol

@Heike yeah, that's true

i also love how there are rules, exceptions to rules, and then exceptions to the exceptions

In Hungarian it is possible to derive a lot. It is an agglutinative language, meaning that each grammatical function has a suffix which can be attached to the word. When two things need to be indicated, we attach two suffixes. For example, plural is -k and object (similar to the accusative case) is -t. For plural objects we attach both --> apple = alma, almák, almát, almákat. This works when deriving new words too: different suffixes change the meaning in different ways ...

... and they can be accumulated.

4:40 PM

i am very thankful that i don't need to learn english as a second langues

@Heike that's not true in all languages. hebrew, for example, has very few roots, and you get the rest by applying a set of conjugation rules

to "verbinate" nouns, and the like

In Greek one ending indicates several things (e.g. both the grammatical case and the number), and what ending needs to be used depends on the word itself as well.

So don't be scared when you hear than Hungarian has 18 cases, it's not nearly as bad as the 4 cases of Greek :-)

@EliLansey "verbing nouns weirds language" :-)

(there are other difficulties though)

@Heike you have no idea

Hebrew language

Hebrew () (, ', ) is a West Semitic language of the Afroasiatic language family. Culturally, it is considered by Jews and other religious groups as the language of the Jewish people, although other Jewish languages had originated among diaspora Jews, and the Hebrew language was also used by non-Jewish groups, such as the ethnically related Samaritans. Modern Hebrew is spoken by most of the eight million people in Israel, while Classical Hebrew has been used for prayer or study in Jewish communities around the world. The language is attested from the 10th century BCE to the late Secon...

like, there are on the order of 100 root words

compared to ~10^6 english words

and the rest you get by prefix, infix, and suffix combinations

So what you're saying is Hebrew is like Mathematica v1, and English is like Mathematica v8?

4:48 PM

I think in Dutch it's more common to turn verbs into nouns.

@EliLansey But the English seem to have kept every word ever encountered in the past 1600 years or so.

@Heike this is true. and has picked up words from other languages much more than hebrew

although, tracing hebrew from the old testament, through today, you can see the introduction of various foreign words

@yoda Ha! Yeah, I guess something like that.

@EliLansey They never bothered to update the spelling of those words either.

@Heike Which explains the "I before E except after C or in any of the following list of exceptions"

@EliLansey I never learned that rule in school.

@yoda no, but let me know if u figure out anything

i also need to do integrals like that

4:57 PM

bessels are generally a monster when it comes to integrating them... at least, for me.

Are a and b real?

No, they are in general complex

@yoda can i guess what you're doing??? :)

no

@yoda ok, fine. that's what im doing, and it's a pain

@yoda what's your field? i want to see this paper when it comes out!

5:10 PM

not very fun...

@yoda intersting. can you let me know if/when this paper is published?

@EliLansey yeah, sure

thanks

All, I just had this question migrated here:

8

Q: Mathematica: Matlab like figure plot

How do you obtain a figure similar to Matlab's figures? That is, how can you make Mathematica have a white background inside the axes and some other color outside the axes. See the following figure: Here is an attempt to replicate the looks of a Matlab figure Framed@Plot[Sin[x], {x, 1, 10}, ...

For merging with:

7

Q: Changing the background color of a framed plot

I frequently generate framed plots like this: Plot[Sin[x], {x, 0, 2 \[Pi]}, Frame -> True] Is there an easy way to change the background color of the framed region only? Specifying the Background option changes the background color of the entire plot, not just the framed region: Plot[Si...

graphics plotting

they're similar, but not quite the same

5:15 PM

pre-emptive aye!

They seemed similar enough not to have both. I hope this was not incorrect. What do you think?

(I am learning on the job here after all.)

I prefer the second one to be the master though... it's a more natural question to ask rather than "matlab like plot"

true

(the aye was for merging)

but i imagine the "matlab like plot" might be a common query?

not sure, though

5:17 PM

Okay. As a rule how should I select the master?

I mean, the old one, the new one, the "best" one, the most concise one?

i think the new one is the most precise, and better question

but the old one may draw in viewers

There is no rule... usually, older questions are already answered and you end up merging the new to the old. However, the master should in general be the clearer question. The answers don't matter because you'll have the same set of answers whichever way you merge

In this case, I think merging old to new is the better solution... some laws of causality might be violated, but that's ok :)

thing is, they're not exact duplicates

what's the procedure for similar, but not idential questions?

@MrWizard I knew I had seen that question on StackOverflow but I couldn't find it again.

Eli, I see that now. The old one is more specific and potentially more in line with the existing questions it has.

I wish I had caught this sooner and done the merge then, but I was slow.

Heike, what do you think should be done now?

5:21 PM

I think the second one is clearer.

Okay, going with the new Q as master.

Please wipe the off-topic & obsolete comments on the old one... I guess from the meta question (which was spurred by my flagging, but not on my answer), my take away was that people don't generally like comments to be deleted, but... let's admit it, we were all loose tongued and chatty on SO since we weren't active in chat. We're more disciplined now and keep such comments to this room

Well I just realized I robbed Simon of an Accept. I feel that by nature of precedence it should belong to him. Eli, if you disagree please speak up.

@yoda I'll take a look at it.

@MrWizard That always happens... one person loses an accept. The OP here accepted Eli's answer, so not sure why he should give it up either...

@MrWizard i have no idea how it works. his question was accepted first.

but it was for a different question

beats me

as simon's answer stands, it is an excellent answer for the earlier question, less so for this question (sort of inverse question)

but, i don't really care

5:28 PM

The fact that this is not an exact duplicate is a complication. Let me put that aside for the moment. If I was on the ball I could have closed the new question before you answered it. Is that the optimal solution?

@MrWizard no clue. i'm fairly new to this SE/SO business. i don't know the rules

Maybe we should add a comment that the question was merged with a previous question on stackoverflow?

I don't know what the best way to respect everyone and keep the site clean is. I'm new to moderating.

Heike, agreed. Do you feel like doing that, or should it come from me?

@MrWizard I think it's better coming from you

How about lets forget about reputation scores and who answered first for a moment and let's try to make it a useful question? Merging would involve editing too, so edit the question and the answers so they fit each other and are as useful as possible for future visitors. (It's a lot of work though)

5:30 PM

@Szabolcs be my guest. :^)

@Szabolcs I think that makes the most sense.

That is a good point however.

I agree with szabolcs here. Perhaps the best solution is to edit the question to also include the inverse — no need for it to be a second question. And @MrWizard If you feel bad for robbing simon of a +15, remember that you can vote for him again here =)

in any case, i need to go now

Luckily most answerers of this question are regulars here so hopefully it won't offend anyone.

5:32 PM

Bye, Eli.

@yoda it's actually not the points but the Accept that bothers me. I just feel like it was his solution, and he should get credit. Of course we cannot preserve a perfect record of who created what, and if we did I'd lose 90% of my Accepts because someone somewhere (MathGroup, etc.) had the method first.

@Szabolcs Probably not, unless those people are trying to save up for extra privileges on SO ;-)

The migration took the SO rep points away anyway

I asked, and I ask again: "If I was on the ball I could have closed the new question before you answered it. Is that the optimal solution?"

@MrWizard Why did it get migrated at all?

@Szabolcs that's what I meant.

5:35 PM

@Szabolcs because I asked for it. :-/

I am trying to do what a moderator should. If at any time I screw up, please tell me, and I will try to learn. Did I screw up?

@MrWizard I don't know ... I think that the version of the question and the answer here were more clear and more generally useful. It could have been taken as an opportunity to have a cleaner version of the question here than on SO. But migration makes sense as well.

@MrWizard No you didn't.

@MrWizard If you feel strongly that the migration was a mistake and that the original should've been left on SO, ping Rebecca Chernoff about it

Actually, you can undo migrations (with an SO mod too) yourself.. but merges need dev intervention

@yoda I don't. But, I do want to respect everyone else's views as well.

so if you feel strongly, go for it. We can have just this question here as a "clean" version

@MrWizard @yoda Well, I edited the MATLAB part into the question. Now the answers make more sense even without touching them.

5:44 PM

@Szabolcs simple, clean, effective. Thank you.

In the future, if I see a new question on this site and it is very nearly a duplicate of a question on SO, which was answered well, should I try to migrate and merge ASAP?

@MrWizard I would think that would be a good option. If someone has something to add, they can still do so

Although, I wouldn't suggest merging right away.

Why not?

Close as a duplicate and wait for a few days to see how the community responds... perhaps it could've been an over sight.. perhaps OP might've wanted to edit and make his question not-a-dupe after seeing the other question, etc.

Reordering lines :-)

:)

The thing is that closing/migrations etc can be handled at the moderator level. Unmerging requires a dev, and they won't be too happy to do this often

5:54 PM

@acl It looks like a Windows version of Julia will happen (if the system becomes popular).

@MrWizard Also, could it be retagged? It shows up as untagged, and for some reason I can't edit it.

@rcollyer I'll look at it.

For those who care, we've just overtaken the French on the site directory.

@Szabolcs thanks

@MrWizard cool, thanks.

5:56 PM

@rcollyer In questions? You can sort by many different metrics. Here it's sorted by visits per day.

@Szabolcs in the Default sorting mode. Which is ordered in some odd way that seems to be a loose "site ranking."

Try this.

Astonomy has much less visitors per day, and so does Philosophy.

And again, we have a hot question at the top of the list. The background colour question. I wonder how hotness is calculated and why this site is always hovering near the top.

@rcollyer If I sort by questions per day we appear to be the top beta site.

@Szabolcs Cue confused answerer/random migration

@MrWizard Absolutely. I like the default sort, though, as it can be viewed as an overall site ranking. And, I like jumping over the longer lived sites as we go.

Although, if the uptick we're seeing in questions per day remains, we could be in very, very good shape. :)

Early graduation? (I mean soon after 90 days).

6:05 PM

Do you ever get worried about the future of SE? I can't even imagine where they get the money from to pay so many employees. I read the Wolfram blog post about MathGroup, and even MG seems to have cost a lot of money to run. Based on past experience all great sites start to impose restrictions, and become not so great or not so free eventually. Like Last.fm ...

Is this sustainable over say, 20 years?

Is it a platform one can count on?

That's an interesting question. Currently, they get a lot of cash from venture capitalists.

How much they earn from advertising, I don't know.

I assume they'll ask for their money back eventually. But I really don't know how these things work.

@yoda said it several times that sites usually graduate after the site traffic starts growing. That is what is needed for a site to be able to generate money from advertisements.

Depending on the contract, the VCs either get stock and/or bonds. Bonds usually have less requirements, but usually the VCs require a portion of the company, sometimes a board seat, or 4.

@Szabolcs I'm looking for hopeful signals in very little data. :)

What is a stock? What is a bond?

Stock: quite literally a percentage of the company.

6:11 PM

I have some vague ideas about the meaning of stock, but have no idea about "bond"

Bond: a loan.

Both, can be traded on a secondary market.

the first is buying a portion of the company, the second is lending to the company and hoping to get back more?

AFAIK, only SO brings good money. Then, perhaps SU, Gaming and SF. They also make good money from careers.se (companies pay to advertise and recruit, etc.)

@Szabolcs Yes.

@yoda SO is actually earning money already? That's encouraging. I see little advertising and almost none that interests me.

6:14 PM

@Szabolcs most ads are removed after you cross 200 rep or something... but you forget one thing... 90-95% of SO's traffic is from random folks Googling :) Us registered users are just seeding the site

@yoda Which makes the whole model brilliant.

try viewing it in incognito mode in Chrome or private browsing in FF

I actually have data to analyze. Cool. Later folks.

@yoda It will probably look the same, if I'm honest.

Bye! :-)

@yoda I checked. They're advertising Careers 90% of the time. Occasionally they're advertising other SE sites. The only way I could get other ads was looking at the top voted questions in the C# tag. Other popular tags (e.g. java) did not give me any non-SE-related ads.

1 hour later…

7:38 PM

Does anyone else think this is too localized? Basically, the OP has solved their problem, but I don't think it's worth keeping and asking him to post his solution as an answr

Or is it just me that has misread the question and it in fact is a good Q (it has 1 upvote as of writing)?

@MrWizard Could you please edit the title of this question and remove the "Mathematica:" part (perhaps rephrase it)? Having that term is redundant on our site and I can't edit it since it's locked

I'm somewhat torn. I would have expected a couple of answers on how to accomplish his goal, or a rewrite of his code. I'd consider doing it as a SparseArray and ArrayRules to accomplish it. It is more specific than the usual list-manipulations, though.

8:20 PM

@yoda I think the OP has only solved part of the problem. As far as I understand, he wants to swap two elements in an array. He got halfway there by doing Si[[i,j]]=Extract[Si,pos], but still needs to set the the entry at position pos to the old value of Si[[i,j]].

9:00 PM

@yoda It only shows "Closed" for me; are merged questions always Locked as well?

9:47 PM

From the docs of FullGraphics: "FullGraphics[g] should display the same as g, though it may have a different internal structure." should --- so unsure, and indeed it doesn't!

9:57 PM

Did you see the W|A background today? wolframalpha.com

Do you see how it could be generated? The rule is not obvious to me.

@Szabolcs Interesting pattern. I don't see the pattern either.

10:35 PM

Wow, someone went to a lot of trouble writing a confetti simulation:

5

A: How can this confetti code be improved to include shadows and gravity?

How can this code be improved, for example, by including shadows, raytracing or the effects of gravity to make it more realistic? I felt that this question deserved an answer. The one I describe here is to create a set of confetti "agents" that respond in quasi-physical ways to external for...

@Heike :D I was just joining to post the same

user image

Yes. I am impressed.

Did you run the code? The one that produces this exact output is not included

I haven't tried the code yet

rendering ...

10:47 PM

I tried it and I get one without any shadows, and strange little lines instead of squares.

@MrWizard Only evaluate the last display (just read the explanation)

Pardon, which part don't I execute?

Ah, never mind.

You got it then :-)

I am curious, what is your AbsoluteTiming to export confetti4a.gif?

I won't do it again, sorry

10:53 PM

I have a feeling that his rotate function could be written as something like frame.RotationMatrix[Norm[Alpha], Alpha]

It is long.

That bad? :^)

Mine is 54.3 seconds.

Wow, 12 votes in such a short time, we must have lots of lurkers around this room :P

@MrWizard All frames or only the ones he exported?

just these: Export["confetti4a.gif", frames[[141 ;; 210]]] // AbsoluteTiming

40.47 with GPU antialiasing intentionally turned off

@MrWizard 47 with AA on

10:59 PM

Good machine. :-)

I'm not even going to try on my 2 year old laptop

@MrWizard It is not mine, it is borrowed for a couple of weeks. Mine would be 3 times slower... This is just an i5 2.4 GHz

I am running at 4.6 GHz. I wonder if Export was improved in v8.

Overclocked?

Yes. i5-2500K.

Some people have them on phase-change cooling at 5.4 or some craziness like that.

11:02 PM

My flatmate tried to overclock his but the OS started to crash and do weird stuff, even though it never went above 50 C according to the sensors. But then he doesn't really know how to overclock, so he gave up.

(he does have some rather big cooler though)

There is really very little to do with these new chips. It's not even really overclocking as they are designed for it (the K CPUs) and the system basically regulates itself. The original Core 2 Duo line was a lot more interesting to play with. I had a 1.8 running a 3.34.

hello @PFonseca

@Heike I'm sorry but this time you've been outdone (re: confetti).

@MrWizard I'm fully aware of that

Maybe she will use whuber's framework to do something more exciting ;-)

It just means I'll have to up my ante next time.

I did try to get a CFD model out of you. ;-)

11:07 PM

@Szabolcs I have a feeling his code could be streamlines slightly.

No doubt, but at the same time the Export is by far the bottleneck.

He's basically solving an ODE

Indeed, Export is the bottleneck.

@MrWizard Graphics rendering is too slow. Once I found that the slowest step isn't even the actual rendering, but 1. converting Graphics stuff to GraphicsBox stuff and 2. compressing it before sending it to the front end for rasterization

@Szabolcs Is that with or without shadow?

Feb 11 at 19:23, by Szabolcs

This is the Koch curve code:

11:10 PM

@Szabolcs I remember that. Was that in an answer, or here in chat?

Way to preempt me. :-p

Well jolly; I just crashed Mathematica trying to export the full set of frames: "The front end encountered an error while processing a "CallPacket" packet. Error in MLTransferExpression"

did one of those confetti get stuck in your processor?

@Heike Generating all 241 frames took 40 seconds here. Exporting 70 frames took 40 seconds as well.

:D

ouch

@Szabolcs How can that be?

I wonder if applying Yves Klett's trick would improve the rendering time by much.

11:18 PM

Which trick is that?

Oh, you mean in the same thread?

yes

Which trick?

glueing all polygons into one big polygon

and using VertexColors to render the different colors

Oh

11:39 PM

I just posted a question that I hope will be interesting. I also hope it still behaves that way in version 8.

11:59 PM

@rcollyer I'd be astonished if VCs took standard bonds, i.e. senior debt. Yes, you get paid out first in a bankruptcy, but there is no upside if the firm wins big. Equity (stocks) is a share of the company and thus the profits. Maybe they took some kind of convertible debt instrument or preferred share: you are more likely to get a return in the early years (they get paid even if shareholders get no dividend) and at some point you get to convert to shares.

« first day (37 days earlier) ← previous day next day → last day (4430 days later) »

Transcript for

Feb '1224

Wolfram Mathematica

Welcome! This is the main Mathematica chat room for mathematic...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile