Wolfram Mathematica

rm -rf

12:03 AM

@acl There's one way to find who it was and force their hand :P

belisarius

12:13 AM

@rm-rf Is it our dear random downvoter?

acl

12:49 AM

@rm-rf pray tell!

acl

1:00 AM

(unless your comment there was The Way To Find Out Who It Was, in which case it flew well over my head)

Oleksandr R.

@rm-rf JMHO but I'm not enthusiastic about cutesy chatroom names. I like it well enough as it is.

acl

@OleksandrR. I thought you were off to write the thesis...

Oleksandr R.

@acl it's 2am... I came home and am about to go to bed.

Bye!

acl

@OleksandrR. excuses are cheap :)

night

rm -rf

@acl No, I'd have to close and reopen which will then show who it was :)

acl

1:13 AM

@rm-rf show how? is this (viewing those who voted to close) some kind of Yodaesque power that I don't have due to my puny rep and pathetic non-mod status?

rm -rf

@acl No, if a question is closed, then the names are made public

You can always find who the closers were in the history (after it has been closed)

acl

@rm-rf I see, much simpler than I thought. But what would be the reasoning behind closing it?

or voting to close

rm -rf

I don't know, and I wouldn't have voted to close. I guess whoever it was, probably thought it was an excel question even though the very first line says otherwise

Meh, not worth getting hung up over stray votes... if 5 of them voted to close, then we have a problem :)

acl

@rm-rf maybe it's an Eloi vs Morlock thing

@rm-rf right, I was just wondering whether I missed something

I fairly recently discovered two things: 1) I am not all that unique in enjoying Neal Stephenson's books (I thought he was a fringe writer and I the discoverer of some hidden treasure, until I googled him...) 2) he's a Mathematica programmer

J. M.

2:02 AM

@rm-rf Maybe the FullForm[] for verbosity's sake: Get["Chat`"]

@acl I read his stuff, but I still think he's relatively fringe. He's not Dan Brown or even Neil Gaiman-level of being famous, I'd say.

Mathematica StackExchange (meta)

4:20 AM

0

Q: Safe-keeping and long term storage of MMA SE Questions/Answers

I wanted to start out by saying that I love the new MMA SE and I hope the meta is the correct place for this post. The MMA SE is turning out to be great and the community is really helpful and I much prefer the interface to the older listservs and usenet newsgroups. However, listservs are easily ...

discussion

Mr.Wizard

5:47 AM

@sjoerdc.devries I don't think this question is good enough to merit migration.

chris

11:28 AM

hello; quick question: can mathematica present a set of {theta,phi} points as a Mollweide representation? Can we do the reverse of mathematica.stackexchange.com/questions/11834/… with a set of coordinates?

it seems the function Mollweide exists and is being used by the WorldPlot/ref/Mollweide package but I did not understand how... May be I should look at the code?

chris

11:56 AM

I add coded something similar in 2001 (!) which extracted level curves from a given contour and then built a country "astro". But it seems my code is not mm8 friendly...

halirutan

@chris, what exactly do you want to do?

I mean if you only want to transform from {theta,phi} to {x,y} you could just use the formula from wiki.

chris

@halirutan I would like to do a 2D plot which shows the whole sphere at once with a line corresponding to the trajectory of a point on the surface of the sphere.

ie I would like to do !Mathematica graphics

but in a Mollweide representation

as I mentionned I wrote a code which used to convert

!Mathematica graphics

into !Mathematica graphics

but i) this code is for contours and ii) it doesn't work anymore ;-(

Do you think I should set this up as a question?

my code was doing something like 'Do[WorldData /: WorldData["astro" <> ToString[i]] =
MapThread[
Join[#1, #2] &, {LevelCurve[cp, cts[[i]]][[2]] /. Line[a_] -> a,
LevelCurve[cp, cts[[i + 1]]][[2]] /. Line[a_] -> a}]*10000/
Pi /. {a_, b_} -> {a/4, b};, {i, 1, Length[cts] - 2}];
WorldPlot[{Table["astro" <> ToString[i], {i, 1, Length[cts] - 2}],
RandomGrays}, WorldProjection -> Mollweide]'

where LevelCurve extracts from a given contour plot cp the contours cts[[i]]

halirutan

@chris Yes, this seems worth a question.

halirutan

12:52 PM

@chris Still here?

chris

1:07 PM

@halirutan yes... (had to attend a phd defense)

halirutan

Want to have a head start for your problem?

(Maybe head start is the wrong word here..)

chris

@sure ! Well I am not sure what you mean but...

halirutan

Starting with the first equations from the wiki page of the Mollweide projection:

cart[{lambda_, phi_}] := With[{theta = fc[phi]},
  {2 Sqrt[2]/Pi*lambda Cos[theta], Sqrt[2]*Sin[theta]}
  ]

Now we need an solver for theta which is used inside and has no analytical solution:

fc[phi_] := Block[{theta},
   If[Abs[phi] == Pi/2,
    phi,
    theta /.
     FindRoot[2 theta + Sin[2 theta] == Pi Sin[phi], {theta, phi}]
    ]
   ];

Create a grid to test the stuff:

grid = With[{delta = .1},
   Table[{lambda, phi}, {phi, -Pi/2, Pi/2, delta}, {lambda, -Pi/2,
     Pi/2, delta}]];
gr = Graphics[{Line /@ grid, Line /@ Transpose[grid]},
   AspectRatio -> Automatic];

And now we transform all points inside the graphic with our transformation cart

gr /. Line[pts_] :> Line[cart /@ pts]

Maybe this helps, but remember, I have no idea what I'm doing here. First time I looked at the formulas of the Mollweide transformation.

With your graphics, which is a set of points or lines too, the approach should be similar.

chris

@halirutan thanks... I ll give it a shot... I was just thinking of translating my problem to the existing WorldPlot formulation as it has all sorts of projections already built-it. But if it fails I ll try your method.

J. M.

2:10 PM

@halirutan The equation being solved there looks awfully like the Kepler equation...

belisarius

3:22 PM

@J.M. hi! Do you remember if there was a q about using the conditions resultant from Resolve[] to define Piecewise functions?

J. M.

4:12 PM

@belisarius If there was, I do not remember it...

belisarius

@J.M. mmm... perhaps it's time to post it

rm -rf

5

Q: Converting ConditionalExpression to Piecewise

I have Solve[] return a list like {{F->ConditionalExpression[EXPR1, 0<z<a1]},{F->ConditionalExpression[EXPR2, a1<z<a2]},...} and so on -- EXPR* and a1,a2,... are somewhat complicated expressions, containing things like Root[]: EQ = F^4 (138880 z + 318688 F z + 98415 F^10 z + 6...

functions numerical-integration function-construction

belisarius

Hehe ... but that one doesn't have generated parameters ...`C[1] \[Element]
Integers && ((\[Pi] + x > 2 \[Pi] C[1] && \[Pi] + 2 x <
4 \[Pi] C[1]) || \[Pi]/2 < x - 2 \[Pi] C[1] < \[Pi])`

rm -rf

Dump alert!

belisarius

$c_1\in \mathbb{Z}\land \left(\left(x+\pi >2 \pi c_1\land 2 x+\pi <4 \pi c_1\right)\lor \frac{\pi }{2}<x-2 \pi c_1<\pi \right)$

@rm-rf O my

@rm-rf Example seems a forbidden word

@rm-rf @rm-rf Using pastebin as an external repository doesn't ensure data survival

rm -rf

4:35 PM

@belisarius No, but in case it goes down (even though I set it to "never expire") or goes out of business, the original data is available in the history here. So as long as paste bin is alive, we can enjoy a short code block. — rm -rf 1 min ago

belisarius

@rm-rf @rm-rf You're right. I guess in these cases it's better to post a warning notice about the data being in the posts' history upon modificat :DDD

rm -rf

@belisarius Warning for the OP?

belisarius

@rm-rf No, for future viewers

If pastebin isn't accessible ... etc

Anyway, the question in question seems nonsensical

rm -rf

@belisarius Ok, see the edit now

belisarius

@rm-rf Yep. Looks very nice

@rm-rf Congrats! ;P

rm -rf

4:42 PM

I doubt if we'd really want to archive the data in this one :P Looks like it's too localized :D

belisarius

@rm-rf I'm trying to understand the integral. Soooo messed up

rm -rf

@belisarius Maybe Artes' answer here might help: mathematica.stackexchange.com/a/1752/5

Might even close it as a dupe...

belisarius

Calculating now

@rm-rf he used Artes' code

acl

I don't see what the question is actually

why are they irrelevant, for instance?

It looks like NARQ to me

belisarius

{epsilon1, w} =
ToExpression@Import["http://pastebin.com/raw.php?i=Z08GwMa8"];
f = Interpolation[Transpose[{Flatten[w], Flatten[epsilon1]}]]
epsilon2[w_] :=
epsilon2[w] =
2/Pi NIntegrate[a f[a]/(a^2 - w^2), {a, 0, 5.5793},
Method -> "PrincipalValue", Exclusions -> {(a^2 - w^2) == 0}] //
Quiet
ListLinePlot@Flatten@epsilon1
Plot[epsilon2[w], {w, 0, 5.57932}, AxesOrigin -> {0, 0}]

acl

4:59 PM

@belisarius it's taking too long here

belisarius

@acl here too

I just wanted to show you "working" code

not the one the OP posted

acl

But I guess my problem with this sort of question is that it looks as if they wrote down literally the definition, it didn't work, and they hurled this at us

belisarius

@acl "it looks as if" == "We're all certain that"

acl

@belisarius right, I'd be surprised if they did try anything else

belisarius

@acl He copied Artes' code for a different problem, and it (of course) didn't work

so ...

here we are

acl

5:04 PM

aha, yes I just saw it!

ha

Rojo

5:48 PM

Sup

belisarius

5:59 PM

@Rojo Inf

Rojo

@belisarius Always contradicting me

belisarius

@Rojo At least I'm frogs now, and not a RGB component

Rojo

Hehe

How's everything going

belisarius

6:16 PM

@Rojo Nice. Still alive.

Stack Exchange

6:30 PM

Blog - Stack Exchange: SE Podcast #33 – It’s Back!

Alex Miller on October 15, 2012

It’s Back! Welcome to episode #33 of the Stack Exchange podcast. We’ve got a brand new co-host (Jay Hanlon, our new VP of Community Growth) plus our guest this week is David Fullerton, VP of Engineering at Stack Exchange.

So what’s new in the seven months since our last podcast? Check out the new and improved review queue! If you’ve got enough reputation, you can see the review button at the top of any Stack Exchange site. The new system is clearer to use and it’s fast thanks to a ton of AJAX goodness. …

Sjoerd C. de Vries

8:20 PM

@Mr.Wizard Here is the version 7-save version (I hope) of the white space problem demonstrator.

Show[
GraphPlot[GraphData["DodecahedralGraph", "EdgeRules"],
EdgeRenderingFunction -> (Function[{e, v},
Inset[Graphics[Rectangle[{0, 0}, {10, 1}]], e[[1]], {0, 0},
EuclideanDistance[e[[1]], e[[2]]], e[[2]] - e[[1]]]]),
VertexCoordinateRules ->
MapIndexed[First[#2] -> #1 &,
GraphData["DodecahedralGraph", "VertexCoordinates"]]],
Method -> {"ShrinkWrap" -> #}
] & /@ {False, True}

acl

8:47 PM

afterliff

(I thought some here would appreciate that)

acl

9:10 PM

this gentleman found the default mma IDE lacking, and asked for help:

-6

Q: Custom File Palette

I find the default Mathematica IDE lacking. I am missing a simple project file tree on the left (as in eclipse, visual studio etc.) The following would be very cool: All your files (in a given directory) would be displayed in a Mathematica Palette. Clicking them changes to the clicked file or ...

palettes

I answered and managed to get 15 upvotes before the question got buried under downvotes

is any kind soul going to help me get a gold badge by reaching +20 votes?

2

(yes I am soliciting upvotes :-| )

halirutan

9:25 PM

@acl you're right behind me in rep. After the good old rules of tactical downvoting I should not have upvoted.

acl

@halirutan but we're not doing this for rep, but for altruistic reasons, right? :)

thanks though, I'd like to have a gold badge for something other than coming here every day

halirutan

@acl yep. That's why you need only 4 more upvotes now ;-)

Apropos we doing it for...

My stackexchange t-shirt, sticker, .. arrived today ;-)))

acl

@halirutan and mine a few days ago. although come to think of it I forgot to even open the package

way too much work, leaving me with a half-asleep brain every night

halirutan

And the best part is, when I bought "Joel on software" I didn't really know who this "Joel Spolsky" was and now I have a letter almost directly from him.

@acl I use the nice trick of sleeping often, not long..

acl

@halirutan doesn't work for me. I go for "work 12h, interact with people and drink beer for 4, sleep for 8"

this in times of high workload of course, I don't always work 12h a day!

nice t-shirt, I just opened the package

Mr.Wizard

9:50 PM

@SjoerdC.deVries Thanks, that clearly shows the problem.

Sjoerd C. de Vries

@Mr.Wizard Reminds me of what SphericalRegion does

acl

I don't think the tags here are the right ones

any ideas?

I'm guessing that complex and system ended up there because he/she thinks of their problem as a "complex system", but here they mean something else

Mr.Wizard

@acl let me look

What tag(s) do you propose? There is no "system-modeling" but there is rather confusing hyphenated "system-modeler" tag.

acl

10:05 PM

@Mr.Wizard I don't know. I thought the system-modeler one was for the system modeler, while this question isn't about that

Mr.Wizard

@acl Yes, exactly. I was about to creat a "system-modeling" tag but I thought it would be confusing alongside that one. I guess I see why system-modeler was hyphenated but maybe it would be better if it wasn't, for distinctiveness?

acl

dynamic (already there) and differential-equations seem appropriate

@Mr.Wizard my reasoning would be: suppose I was looking for an answer to this. what tag would I be looking for?

Mr.Wizard

@acl Not inappropriate, but sufficient? Clearly the OP was focusing on "complex" and "system" in his own mind...

acl

and I guess I'd be looking for the two I mentioned, and I can't think of a tag which would refer to the perturbation

@Mr.Wizard and we don't have that many questions on delay equations so I don't know if it makes sense to have a tag about them

@Mr.Wizard that's too domain specific to be useful here, I think. what's complex to me may be simple to you and vice versa

anyway I don't know. that's why I asked here

Mr.Wizard

@acl Right, I'm not for that. ;-)

Do you not see value in a new tag "system-modeling" ?

acl

10:12 PM

@Mr.Wizard ouch

@Mr.Wizard well, what does it mean? or rather what does it exclude? it seems to me to be describing the motivation behind solving a problem (namely, to model some system), not any property of the set of equations you're solving

Mr.Wizard

I see your point. It might be informative to look for similar tags on other SE sites to see how well they appear to be working. Thanks for raising this question, BTW. Perhaps it is worth a Meta post.

acl

@Mr.Wizard probably. we can't decide what to do, so let's ask someone :)

Mr.Wizard

10:57 PM

@acl are you still here?

acl

yep

Mr.Wizard

I'm not understanding this question:

3

Q: Mathematica: summing over n abstract indices

This probably has a one line answer, but I'm totally stuck. I have two tensors, i.e. two objects that depend on NN abstract indices which i've labelled i[m] (m=1,...,NN). I want to keep NN general for now. Each index i[m] ranges over the values {1,2}. I want to sum over all the i[m]s, i.e. Sum[...

list-manipulation

Could you try explaining it to me, please? :)

acl

I think he literally wants to do Sum[a[i[1],...,i[NN]] * b[i[1],...,i[NN]], {i[1],1,2},{i[2],1,2},...,{i[NN],1,2}]

so, he has two N-dimensional tensors a and b

that is, $a$ and $b$ are like matrices but, instead of having 2 indices, they have $N$; so the elements of $a$ may be specified as $a_{i_1,i_2,...,i_N}$

so I think he wants to do this, with the sums ranging from 1 to 2

is that clearer?

Mr.Wizard

Better, yes.

acl

this looks like it should be solvable by some sort of awe-inspiring one-liner, fit to scare small children with, but I've always used the same approach jVincent describes

(by the way did I explain whatever it was that was puzzling you? or did I miss the problem?)

Mr.Wizard

11:11 PM

I think so, but I'm still looking at at. If SeedRandom[1]; {a, b} = RandomInteger[9, {2, 2, 3, 4}] then ...?

acl

No I think Dimensions[a] should be {2,2,2,2,...,2} (and there should be NN of them, ie, Length@Dimensions[a] should be NN

Mr.Wizard

Damn, I'm just dumb as a stump today. <:-/

acl

so for NN=3 it should be something like

SeedRandom[1];
{a, b} = RandomInteger[9, {2, 2, 2, 2}]
Total@Flatten@
  Table[a[[i, j, k]] b[[i, j, k]], {i, 1, 2}, {j, 1, 2}, {k, 1, 2}]

or anyway I think so, it's pretty late here and I'm tired

Mr.Wizard

Thanks for humoring my stupidity.

acl

(and I'm also not good at thinking in front of computers so take all this with a grain of salt)

no problem. someday, and that day may never come, I will ask a favor of you.

Mr.Wizard

11:17 PM

lol

acl

ah found the actual quote: "Some day, and that day may never come, I will call upon you to do a service for me." better :)

Mr.Wizard

11:31 PM

@acl After staring at this for a while trying to see the hidden complexity, it looks like nothing more than Total[a b, -1] -- am I still missing something?

acl

@Mr.Wizard awesome

post it as an answer

Mr.Wizard

Seriously? That's all?

acl

apparently

Mr.Wizard

LOL -- damn, talk about making a mountain out of a mole hill. I was sure it wasn't that simple.

I'm gonna get some food. See ya later.

acl

bon appetit

I guess this would be more involved if it was something like this

well here you could transpose and do what you did, but with more indices, contracted with non-corresponding indices, it would probably be too much to keep in one's head at once

drN

11:36 PM

This may seem kinda sophomoric but how do you find the Min value of an interpolating function? I have an interpolating function in x y and t. With FindMinValue what values t must I use (the inter. func. doesnt just have 3 rows so I am dumbfounded)?

Transcript for