« first day (401 days earlier)      last day (4097 days later) » 

12:00 AM
that also satisfies my needs
both examples are way better than my clunky procedural thinking, TY!
 
No problem
 
Ohh, [[]] is Part not Extract :D
 
acl
12:17 AM
@AdamDreaver however the original way you wrote it allows you to change the structure without changing the interface.
 
12:35 AM
Not true, of course. Something isn't working
 
acl
mine's just fine
(0, 10, 30 :) )
 
@acl I guess something is preventing the script to recognize the date change
 
acl
@belisarius do the two last days's votes add up to the right one? maybe eg the script has changed to consider date in a different timezone than yesterday and it's lagging?
wonder how we could find out
 
@acl Too much trouble .) I'm sure tomorrow I'll not get to 480. A pity :)
Well, it isn't the browser. I restarted it and got the same
 
1:07 AM
@belisarius The "today" in that is your today, not UTC, if I remember correctly
 
acl
1:19 AM
@belisarius time is relative anyway
 
@rm-rf "Puestos a elegir entre la verdad y el placer de buscarla, elegiríamos lo segundo." A.Machado
@acl Bullshit. tell that to Harrison en.wikipedia.org/wiki/John_Harrison
 
acl
@belisarius interesting association you came up with there
also led me immediately to this
100 Greatest Britons was broadcast in 2002 by the BBC. The programme was based on a television poll conducted to determine whom the United Kingdom public considered the greatest British people in history. The series, Great Britons, included individual programmes featuring the individuals who featured in the top ten, with viewers having further opportunities to vote after each programme. It concluded with a debate. All of the top 10 were dead by the year of broadcast. The poll resulted in nominees including Guy Fawkes, who was executed for trying to blow up the Parliament of England; Oliv...
 
@acl Which in turn, led me to this
Connections is a ten-episode documentary television series created, written and presented by science historian James Burke. The series was produced and directed by Mick Jackson of the BBC Science & Features Department and first aired in 1978 (UK) and 1979 (USA). It took an interdisciplinary approach to the history of science and invention and demonstrated how various discoveries, scientific achievements, and historical world events were built from one another successively in an interconnected way to bring about particular aspects of modern technology. The series was noted for Burke's cris...
 
acl
actually I remember the poll as I was living in London at the time. didn't remember the results
 
@acl Mr. Big Ben?
 
acl
1:28 AM
@belisarius this must have been nice
The Ascent of Man is a thirteen-part documentary television series produced by the BBC and Time-Life Films first transmitted in 1973, written and presented by Jacob Bronowski. Intended as a series of "personal view" documentaries in the manner of Kenneth Clark's 1969 series Civilisation, the series received acclaim for Bronowski's highly informed but eloquently simple analysis, his long unscripted monologues and its extensive location shoots. Overview The 13-part series was shot on 16mm film. Executive Producer was Adrian Malone, film directors were Dick Gilling, Mick Jackson and David...
but I have never seen them. the book was fantastic though
 
@acl I think Harold II deserved the first place. You know: "Seven feet of English ground, or as much more as he may be taller than other men. "
 
acl
@belisarius ha
 
@acl I've really enjoyed BBC's docs when I was young. They are masterpieces (with some exceptions, of course)
 
acl
yes the BBC know what they're doing
 
Are there any special mechanisms for storing application data, or should I just write into $UserBaseDirectory/ApplicationData/MyApp directly?
 
1:32 AM
@acl Sometimes
In September and October 2012, almost a year after his death, allegations that the English DJ and BBC television presenter Sir Jimmy Savile (1926–2011) had sexually abused under-age adolescent and prepubescent girls and boys, and adults, became widely publicised. By 11 October 2012 allegations had been made to 13 British police forces, and led to the setting-up of inquiries into practices at the BBC and within the National Health Service. On 19 October the Metropolitan Police Service launched a formal criminal investigation, Operation Yewtree, into historic allegations of child s...
 
Hi @rm-rf
 
acl
@belisarius right.
princess diana ahead of newton
guess I have a different worldview
 
@acl Newton was an acid old man. Never heard about savoir faire
 
acl
@belisarius I'd say that he did, otherwise we wouldn't have heard of him
although maybe in a different way than her
"maybe" :)
 
@Szabolcs Hi
 
1:36 AM
@acl All that buzz was because Europe was against Germans. Leibnitz did it all :)
 
acl
@belisarius neal stephenson has written some nice books using that dispute as a plot line
 
@rm-rf Could you please push the error reporting fix? It would be quite useful for me for testing.
 
@acl Yup. The Mozart-Salieri relationship type sells many books
 
 
2 hours later…
acl
3:28 AM
9.0.1 seems to have the kinks worked out
for me at least
 
@acl That only means you haven't spent enough time with it/her
 
acl
@belisarius always the optimist
 
@acl The name is experience
 
acl
(I think of mma as her, for obvious reasons)
 
ups
 
acl
3:33 AM
:8219599 a bit more nuanced than that :)
ha
 
@acl erasing the past leads to one's destiny
fate, you know. It's all about fate
 
acl
@belisarius I go for "our future is bright, the past, uncertain"
better this way
 
You can't measure yourself and the future at once. So, when you are able to see the the future, you don't know if it's yours
Heisemberg dixit
Causality is but a cloudy homage to past victims of fate
2
 
acl
@belisarius that sounds deep but I can't work out what it means
starred though
 
@acl Me neither. I should try to write some bestsellers
Interesting ... it seems this very day never ends for SE. I've 310 rep accumulated in one day. Is groundhog day today?
 
3:57 AM
@belisarius Don't complain... this has been a slow month here.
Actually, slow year.
Mr.W has increased the gap between the two of us by almost 5k
 
@rm-rf Do you think killing him is enough to alleviate your pain?
 
@belisarius The ghost of his 800 answers will continue to accrue rep, so no... at least, not until I post 400 more will we be even
 
@rm-rf You may always CW all of them
@rm-rf Some of these days you'll come up with a proposition he wont be able to refuse
(I've read that book -first edition- :) )
 
@belisarius Well, I have one absolutely foolproof way — I just need to join PE and overtake @mr.wizard's tally so that he'll be forced to spend more time there =)
 
@rm-rf Hmmm ... and what's PE?
 
4:11 AM
Project Euler
 
@rm-rf I think it's better to keep your current job and just kill him.
 
 
2 hours later…
6:13 AM
@belisarius Now you're plotting my destruction? :-O
 
@Mr.Wizard It was @rm-rf! He tempted me to help. When I told him that you're my favorite mod and I could never kill you he trashed all his messages using his toad.fu and leaved me appearing as the culprit. I swear!
 
@rm-rf I have no delusions of any great accomplishment on Project Euler. Sasha solved every problem I did (he had a list by number) and nearly a hundred more in short order once he started.
@belisarius "This woman you gave me handed me some fruit and I ate." -- sound familiar? :-P
 
@Mr.Wizard yup. heard it somewhere
 
7:04 AM
How do I turn an expression in matrix form back to it's list form, i.e. m=( a11 a12 a21 a22 ) into { { a11, a12 }, { a21, a22} }
 
@AdamDreaver You mean m ={a11 ,a12, a21, a22} as a starting point?
 
Take a look at Partition[]
 
err, no I mean I have a matrix, i.e. a = MatrixForm[a]; how can I remove the matrix form from it?
 
@Mr.Wizard Well, there's a huge difference in ability between Sasha and me... I don't think you'd be able to stomach me overtaking you in PE :D
@AdamDreaver Identity @@ a
 
7:12 AM
Ahhh, the form I desire
ty
 
@AdamDreaver btw, don't use MatrixForm to do calculations... use only for display/pretty printing. Read this:
20
A: Why does MatrixForm affect calculations?

rm -rfMatrixForm is a wrapper that pretty-prints your matrices. When you do the following: cov = {{0.02, -0.01}, {-0.01, 0.04}} // MatrixForm you're assigning the prettified matrix to cov (i.e., wrapped inside a MatrixForm). This is not accepted as an input by most functions (perhaps all) that take ...

 
@rm-rf Ok i'll read it :D, ty
 
 
5 hours later…
12:34 PM
Welp, I've found my new favorite thing to do:
Rotate[Manipulate[Plot[Cos[x /a], {x, -5, 5}], {{a, 1}, 0, 5}], 1]
Hmm...
Rotate[Slider[Dynamic[x], {0, 2 Pi}], Dynamic[x]]
 
Perhaps I'm missing something, but shouldn't
CompoundExpression[Print[x], Return["A"], Print[y], Null]
 
Hello people
Are you able to change the default stylesheet without the FE hanging?
 
Return "A" insted of Return["A"]
@Rojo You probably mean in mma 9.0.1?
@NeuroFuzzy Hehe, good one:)
 
@Ajasja Perhaps there's another way to do it because I seem to recall having done it successfully, but I also recall having failed with v8.0.0 too. I mean, setting the global option DefaultStyleDefinitions to some "blablacustom.nb"
 
@Rojo Just started using style sheets recently, so probably not much help here.
 
12:49 PM
@Ajasja, maybe it's just odd because it's not used in a function?

f := CompoundExpression[Print[x], Return["A"], Print[y], Null]; f

works fine
 
Wll, the help file says
The returned value can be the result of Return[expr].
The evaluation of the Subscript[expr, i] can be affected by Return, Throw and Goto.
 
@Ajasja see @Rojo's excellent description here. Basically, because CompoundExpression doesn't effect any sort of transformation, when Return walks back up the stack, it doesn't find anything. Hence you get Return["A"] sent back to top level.
 
@OleksandrR. Ahh, thanks
@Acl @OleksanderR @all Any recommendation for a C++ numeric library? I'll mostly just need one huge list of 1D vectors (that should be able to resize dynamically), some Gaussian random numbers and some eigen value calculations.
I think this is coved by most (eigen, it++, Aramadilo...). Actually the most important thing is **friendly syntax**.
 
@Ajasja blaze? I haven't used it myself (don't know C++) but seems to be well regarded
 
@OleksandrR. Thanks. Looks interesting
 
1:45 PM
@OleksandrR. Hey, do you know about changing the default stylesheet
 
@Rojo Hi! Wanna die from envy?
Hi all!
 
@belisarius 2 favourites this week only? Lame
 
@Rojo I don't use them too much
 
@belisarius I'm kidding. I didn't use them either, but I also didn't know that you see highlighted those that have updates
I have one of those screens that you have to look sideways to realise those color subtleties
 
@Rojo I've one of those eyes that you have to guess what is displayed at the screen
 
1:57 PM
@belisarius What about the other eye?
 
@Rojo I sold it
thirld world, you know
 
2:08 PM
@belisarius Believe me, I know
 
acl
2:40 PM
0
Q: "open" boundary condition in NDSolve

wdgThe original equation comes from here. The central equation to solve is equation (16), which is of the form: $\dfrac{\partial f_n(x;\tau)}{\partial \tau} =\left[ [-(n-6)+\dot{\bar{A}}x]\dfrac{\partial}{\partial x}+2\dot{\bar{A}}-(c_++c_-)n \right]f_n(x;\tau) +c_+(n-1)f_{n-1}(x;\tau)+c_-(n+1)f_...

"please read this paper, work out what the correct boundary condition is and give me the code to do it".
@Ajasja does it have to be C++?
(also eigenvalues of matrices which are how large? are they sparse and does it matter?)
do you need all the eigenvals or just a few of the lowest?
 
3:19 PM
Anyone have experience with keeping contexts clean of valueless symbols?
For instance when defining f[x_]:=x^2 in a package with context pkg', it'll export both f and x, which means that future definitions to x set pkg'x rather than Global'x. But removing x in the package leads to ugly definitions for x eg: f[Removed[x]:_]:=Removed[x]^2.
Are there nicer ways around this?
 
@jVincent You put those definitions in a private context within your package and expose publicly only the functions that you will use
To "expose" a symbol, just use it somehow after BeginPackage and before Begin
 
So I basically need to put all of the actual implimentation outside of the BeginPackage, and then just write a list of function symbols that make up the actual package?
 
No, the implementation will be inside a private context (with Begin and End) inside of BeginPackage and EndPackage. Hold on...
By defining a usage message for f outside the private context, I make it public or in other words, in TestPackage`f, so that when you load the package, TestPackage` is on the context path and f is accessible. x however will reside in TestPackage`Private` , which is not on the path
 
3:34 PM
@rm-rf Hey
Can you change the DefaultStyleDefinitions without your FE crashing?
 
Oh to use a new default stylesheet?
 
@rm-rf But this completely removes the definition from the symbol right? At least ?? won't give it anymore.
 
@rm-rf Right
@jVincent Hello
 
ohh wait. Need to clean up this mess.
 
?? won't find it the context isn't in hte context path
 
3:36 PM
@Rojo I haven't done it in 9 yet... I did it in 8 and it didn't crash. However I had other issues with output cells not respecting page width
 
@rm-rf Could you confirm how exactly you did it? Because I recall it crashing too in v8.0.0
and I'm about to test now in 8.0.4
 
Got f being snapped up by the wrong context. So Now I'm left with definitions referencing private contexts, which I suppose is nicer than the Removed symbols.
 
@Rojo SetOptions[$FrontEnd, DefaultStyleDefinitions -> "path to stylesheet"] or via the options inspector
 
aaand, it crashed. Tried it from the options inspector, and with CurrentValue. I'll try with SetOptions
 
@Rojo Did you use full path?
 
3:40 PM
@rm-rf No
Adam Dreaver is a massive starrer
 
@Rojo heh, I noticed.
@jVincent Yes, it will reference the private context because that is where the symbol x resides. Typically, if you're writing a package for others, the idea is that users won't need to worry about internal implementations, so you just expose the public ones and hide the rest (via read protected). Those who know how to look will look anyway. As the developer, you have access to the actual file which is free of all the foo`private` stuff
 
Does Rasterize have some time constraint? I'm Rastetrizing a manipulate control and the inner Graphics just becomes $Aborted with a few locators on it
 
@jVincent If you still want to read definitions when working, you could either use something like my simple redefinition of Definition here or use Simon Woods' Spelunk
 
@ssch No error msg?
 
@rm-rf Sure, and if someone want's the symbols to look nicer I suppose he can put the priave context on the path and get the definition and remove it again.
 
3:46 PM
@ssch Probably Manipulate is the culprit rather than Rasterize... I had this happen on v9 with some other function (inside manipulate). If I remember right, v8 would've waited till the computation finished
I didn't care enough to investigate...
 
@belisarius Nothing
@rm-rf I'm on v8, but I'll dig around a bit in Manipulate options and see
 
@ssch Have you tried a simpler image?
 
Huh... TouchscreenAutoZoom
> Settings for TouchscreenAutoZoom will have no effect unless Mathematica or Player is running on a touchscreen device supported by Wolfram Research, such as the iPad.
I didn't know you could run CDF on iPads... or is that meant to be future proof?
 
@belisarius Yes, It's this code: mathematica.stackexchange.com/a/20015/1517 putting divisions to 8, clicking Star, and then Rasterize[<that manipulate>] for lower values of divisions like 3 it works
(I cheated with the screenshot there and put a rasterized version of the graphics inside a manipulate with same controls)
 
@ssch Let me try it here
Gimme the values
 
3:50 PM
Divisions: 8
Multiplier: 2
Click Star
 
ok
@ssch How are you Rasterizing it?
 
@belisarius I type Rasterize[ in front of the object and ] on the other side
 
@rm-rf Hangs also with the full path :S, perhaps it needs a wrapper
 
@belisarius tutorial/AdvancedManipulateFunctionality mentions there is a time limit on manipulate evaluations, but not where it is set: And beyond five seconds you will start seeing $Aborted instead of the number, because the system is protecting itself from unreasonably long evaluations, which block other activity in the front end in this situation.
I'll make a question about this
 
4:04 PM
@ssch Same result
 
SynchronousUpdating seems to be key
 
4:17 PM
@ssch but it only works for FE functions
@ssch you may use "save selected cell" to get a png image
 
@belisarius That was more convenient than using Export in the first place :)
 
@ssch Export aborts here
 
Yea, that' s how I noticed $Aborted in the first place
 
@ssch But using FE functions should work
 
5:03 PM
@Rojo Could the problem be that it is not a "proper" style definitions file?
 
@rm-rf It works if I set it manually
So, a notebook with StyleDefinitions -> "NewStylesheet.nb" works
OH
Nope, hang again
 
5:47 PM
@Rojo nope, sorry... never use stylesheets (but I guess you got your answer from R. M. now anyway)
 
5:58 PM
@OleksandrR. Well, I still couldn't make it work
 
@Rojo sorry to hear it. But I'm going to the pub :)
Best of luck!
 
@OleksandrR. Go do what you must
Thanks
 
6:52 PM
Is there an easy, built-in way to retrieve the text of a message?
The documentation of Message describes how messages are searched for, but this is a complicated procedure.
E.g. if I want Power::infy, I can't just evaluate it because it actually maps to General::infy
Then there's the whole $NewMessage mechanism
 
 
2 hours later…
9:10 PM
May be it is just me, but I find it strange that on wolfram.com/support/community WRI does not list this Mathematica site as part of the community, but they list the internet Mathematica newsgroup there! Yet, Mathematica on stackexchange is much much larger now than the newsgroup and has much more useful Mathematica information for the community.
Is it possible that WRI does not know about this site?
 
Hmm, I'm 99% certain they know about it, we probably are a competitor for their training programs. Newsgroups are ok, but SE blows them out of the water.
Hell I emailed a rep last night saying I don't need any of their training, SE + docs + google gives me all the training I need.
They were hunting me after I watched one of their training vids on "functional mma programming".
 
9:46 PM
@Nasser When you say that stackexchange (which is wonderfull) is much larger that the newsgroup, do you take the archives into account ?
 
10:01 PM
I must leave. Bye
 
acl
10:37 PM
@Nasser it's not their forum, perhaps
 
@Nasser oh they definitely know. I think they want to have something without the strict Q&A format.
 
@andre, sorry was away. Yes archives for Mathforum are very useful ofcourse. @others, well, for me, WRI should be interested in making Mathematica more popular I would think. Hence not letting people know about other sites such as this with lots of useful information about Mathematica and where someone can get fast help on Mathematica does not seem to make much sense to me specially if done by the same company that makes Mathematica. But what do I know, I am just an engineering student.
 
Just dropping in. See you all later.
 
10:54 PM
I'm afraid so. I'm learning C++, since this is what most people use and what a lot of the programs I use are written in (NAMD, VMD, ...)
For the project I have in mind I don't need much:
a list capable of holding ~4*10^7 elements of n sized vectors (where n is the same in the whole list and from 1 to 4).
Then I have to make a histogram list of these elements and calculate the eigen values and vectors of the covariance matrix obtained from the elements. In short I only need to get the eignevalues of at most 4x4 matrix.
 
acl
I asked about C++ because I was going to suggest using the GSL, which is written in C but can be used from C++ (this is how I use it). I am not sure how well it performs with large-scale stuff as I've never used it for that. certainly the eigenvalue part is doable there
 
@Nasser Sure they do. A few regulars here work at WRI.
 
acl
however when you put it like that, why do you need any libraries? you could use STL containers to build up whatever you need; binning things is not that hard, I think. and solving a 4d eigensystem can be done as inefficiently as you want :)
anyway the problem with GSL is that you need to think in both C and C++ to use it (or anyway I do, but I'm pretty bad at both; others will probably find it easy)
 
@acl Yes, that's what I'm considering now. But from experience the project requirements always grow... :) Also I don't have any particular desire to write basic containers. Butt std::Vector is probably sufficient for my needs.
About GSL: I'd rather write J(1,2)=343 than gsl_matrix_set (J, 1, 2, 343);
 
acl
@Ajasja you're not alone.
 
11:09 PM
@acl OleksanderR recommended blaze, which has a pretty nice syntax
 
acl
@Ajasja I've never used it
the syntax certainly looks cleaner than what you end up using with gsl
 
Eigen is another interesting choice.
But GSL is certainly the largest and most well used. I guess at this point I'll just choose the one whit the best syntax. Got to go -- bed time :)
 
Damn! I can't find the "obliterate" button mathematica.stackexchange.com/q/20037/193
 
@belisarius lol
 
acl
@belisarius it's not that bad
 
11:22 PM
@rm-rf Sometimes I oscillate between feeling too old or/and socially handicapped
 
acl
(off topic though)
 
@belisarius How old are you?
 
@acl C'mon! What is the meaning of the second image?
 
acl
@belisarius eg, why is this better?
 
@rm-rf 14
@acl It isn't. I downvoted it
 
11:24 PM
@belisarius "I have no intention to use any value that has any negative value."
I think he plotted something and truncated at 0
(or Max[x, 0])
 
@rm-rf I think he doesn't understand his own question
 
acl
@belisarius I don't know. but he/she at least tried. the other question just says: "hi! read this PRE, understand what the boundary condition is, and tell me how to do it in mathematica".
 
@belisarius Probably should've thrown a close vote at it too...
 
acl
I wonder if the people who do this here would really go into their supervisor's office (I am guessing they're students) and ask it like I just did
 
@acl You're right. But it's also like "find Wally for me"
 
acl
11:26 PM
@belisarius well yes, in the sense that stealing is stealing whether you steal a bottle of water or the entire life savings of a poor mother on her way to buy bread with the last of her money for her starving child
 
@acl Oh, yeah! They call themselves physicists, engineers, chemists, you know. One week ago a colleague said that the algorithm we were using isn't the robust one because that paper was too long
 
acl
@belisarius concluded it's not robust because the paper was too long?
 
@acl No. He rolled up his own (very simplified) one. It has problems only in some corner cases. Regrettably, that corner cases are the only ones that matter to us.
 
acl
@belisarius that's what I tend to do too, write up some random half-thought-out algorithm I just came up with to do things. very slowly, not robustly (so I need to do it in two different ways to check), takes longer than actually looking up how it should be done. but no.
I have managed to avoid publishing anything wrong by a continuing series of miracles
 
11:42 PM
@acl The first needed miracle is that nobody checks deep enough what one writes
 
acl
@belisarius that's one approach. publish obscure stuff in obscure journals so that nobody notices it and checks.
sadly you can't do that for very long as you run out of job
some people do get away with things though
(for some time anyway)
 
@acl I've seen the word evidently used in the most curious ways
@acl I love his question Remind me again, why is it that universities are paying large sums to get these journals?
nobody really knows
 
@belisarius back issues; you can't do research if you can't follow a reference trail back into at least the 1950s
 
@Xerxes Ok. Now tell me about how much a good pub pays to its referees
or its authors
are the costs "printing costs"? paper? ha!
 
@belisarius hey, I'm not defending the publishers; just saying why the universities are effectively held hostage to them
we really do need access to that stuff
 
11:52 PM
@Xerxes Just because being a referee and/or author in those journals gives you prestige. How much does it cost to provide the same service in a site like this one?
The ivory towers should fall. And will
 
acl
@belisarius certainly not to ensure quality, as that is (meant to be) done by the referees. who are volunteers.
@Xerxes he's not disputing that though; just pointing out that it is a very strange situation, if you think about it
 
@rm-rf My daily rep are still continuously increasing. I'll beat my record :D
@acl Exactly. Journals are badly needed. Referees too. Authors, of course. Nobody gets paid. Why unis pay?
 

« first day (401 days earlier)      last day (4097 days later) »