Wolfram Mathematica

halirutan

10:18 AM

@Matariki Good morning.

Matariki

10:35 AM

Good Morning to you too, @halirutan

Although I'm going to bed soon

halirutan

@Matariki You are on the wrong continent ;-) Here it's sunny and half past twelve.

Matariki

@halirutan Yes, I guess I am. Here it is dark and slightly overcast

Good to hear weather is good. Winter has finally left you and is on its way down here ;)

I have installed your plugin but havn't been able to use it in earnest

halirutan

@Matariki yes, thank god. I'm going now by bike everyday I'm in the office.

Matariki

I have to start a rather large package development on Friday and will start using intellij

Biking works for me still. We had a very sunny day but soon it will get wetter

halirutan

@Matariki Since yesterday there is an update out supporting highlighting of builtin functions and patterns and rm-rf implemented a lot so that you can create Mathematica projects from scratch.

Btw, you don't have to install it manually any more.

Matariki

10:43 AM

@halirutan Ok, I'll download it before I start on Friday

I have seen the plugin in the repository

Is that what you are referring to?

halirutan

@Matariki If you have installed it from there, then you can just click on update

@Matariki yes, I was referring that you can go to "browse repositories" inside IDEA

Matariki

That sound very good

So that means you have made it into the jetbrains repo

halirutan

If you installed it from disk, you should first de-install it and then re-install it from the repo.

@Matariki yes

Matariki

Yes, I installed originally from disk. Thanks for letting me know to nuke it first

halirutan

And I can put the updated package to the Jetbrains repo and you see instantly in IDEA that an update is available.

Matariki

10:47 AM

@halirutan Thats great. I really like intelij.I havn't used it for quite a while but when version 7 and 8 where out I used it every day for java dev work

@halirutan I have to go now. I have to prepare a 40 minute conference talk tomorrow and need to have my brain functioning.

I hate those long talks

halirutan

@Matariki Good luck!

Matariki

Thanks. Have a good day.

halirutan

@Matariki Although in Germany you rather say "Viel Erfolg" than "Viel Glueck".

Matariki

Genau

Yves Klett

11:22 AM

@Matariki @halirutan unless you were talking in a mine. Then "Glück auf" would be appropriate :-)

halirutan

@YvesKlett Btw, are you going to Frankfurt?

Yves Klett

@halirutan I´d really like to but right now there is a date conflict :-(

@halirutan are you going?

halirutan

@YvesKlett That's probably one of the few occasions to talk with @LeonidShifrin and since Rolf is coming too, I'm pretty sure I'm going, yes.

Yves Klett

@halirutan give them my regards in any case (unless they read this - regards!) and have a lot of fun

It´s not really far even for the seasoned travelophobe (meaning you) :-)

halirutan

@YvesKlett Compared to my little village, Frankfurt is an ugly big city. Not worth a look;-)

Yves Klett

11:37 AM

@halirutan not sexy, but rich :-)

halirutan

@YvesKlett capitalists all over the place.. lovely

Yves Klett

@halirutan you gotta spend your travel budget somehow, right? Otherwise let me know, always happy to help

halirutan

@YvesKlett Yes, I have to.. otherwise its probably cut.

Vincent Tjeng

12:38 PM

I have a bit of Dynamic code that allows me to choose points of interest on a ListPlot[]. The chosen points are then coloured on the screen so that the user will remember which points have been chosen. Now, I'd like to use Button[] to allow the user to indicate when he is done choosing points, clearing the Dynamic output and then run the evaluated points through another function. Is there any good way to do so? I can't get my code to run properly at all.

I'm in the process of formatting my code so that it's understandable, but i'm wondering whether anybody has an idea of how to do so for now.

Vincent Tjeng

12:53 PM

Using my favourite sample data:

temptime = Range[200]/10;
tempvalue = Sinc[temptime] + 0.02*RandomReal[{-1, 1}];
tempdata = Transpose@{temptime, tempvalue};

paddingPoint = {0, -10};
mFEPtmax = {paddingPoint};

Dynamic[Show[
ListPlot[Button[Tooltip@#, AppendTo[mFEPtmax, #]] & /@ tempdata],
ListPlot[mFEPtmax, PlotMarkers -> {Automatic, 15},
PlotStyle -> Green], ImageSize -> 700]]

I would like to have a button that allows you to press "Done" to quit the Dynamic and collect the points in mFEPtmax

P. Fonseca

1:07 PM

@halirutan I'm also going to Frankfurt!. It's not far from where I am :)

halirutan

@P.Fonseca I feel the pressure to join is growing ;-)

P. Fonseca

@halirutan So lets just hope that no one loses his head :) I think there can be more than one!

Yves Klett

@halirutan peer pressure, eh?

P. Fonseca

@YvesKlett Immortal pressure... (a thing of the 90's...)

halirutan

@YvesKlett Yes, this is how it is called.. at least that is what they told us in school: "Sozialkunde"

Yves Klett

1:20 PM

@P.Fonseca ahhh, fond memories - but Highlander was from 1986...

P. Fonseca

@YvesKlett I'm I that old?!

Yves Klett

@P.Fonseca look at us, reminiscing away :-)

but to be fair, I only got to see it in the 90s

just like Top Gun :-P

P. Fonseca

@YvesKlett talking about age... what is going to be the big suprise for the 25 years celebration? the online interface? the iPad version?

a minority report interface associated with google glass?

Yves Klett

@P.Fonseca I´d be all for an Android port, though. I guess @halirutan would also be hugely in favor of anything linux-based.

kirma

I would certainly welcome reasonable interface for Android. quad-core phone with 2GB of RAM, full HD display and LTE would be well served at least with cloud kernel backend, and phone frontend

too lazy to attempt setting up a remote desktop for this purpose at home, and a bit awkward to use I bet...

in comparison to Alphas that were quite sufficient for Mma use in mid-nineties, these phones are far superior in every way.

halirutan

1:35 PM

Would you really sit anywhere and program Mathematica on a phone? I'd rather watch out for girls when I'm out.

2

kirma

not program, but do stuff that one can't do properly on Wolfram Alpha.

even somewhat complicated oneliners are a problem for Alpha, after all

P. Fonseca

I'm for an online interface that allows a flexible licensing, and easy colaboration

Pay per use would be of interest in some of my project

What will we do if WR presents us with a stackexchange type of site that allows direct editing and manipulation of the Mathematica content?

They seem to already have all the ingredients...

kirma

@P.Fonseca I think you're overestimating their capabilities to deliver something that doesn't necessary have good potential to be sold or to be used as a long-term building block of further development... considering 9.0 was released just half a year ago. :-/

capabilities, or priorities

Mr.Wizard

It seems to me this question is worthy of attention, though perhaps an answer would take too much work? I am not competent to answer it myself.

2

Q: Solving recursion relations using Mathematica

I want to solve the recursion relation given in equation 2.7(a/b) on page $6$ of this paper. (..the initial seed is $F_1 = G_1 = 1$ and the functions $\alpha$ and $\beta$ are defined on page $5$ in equation $2.4$ on page 5..) Eventually one defines the function $F_n^{(s)}$ and $G_n ^{(s)}$ as giv...

kirma

I'm interested to see something like a mobile-OS frontend and could kernel stuff, and some web/cloud services, but I have some doubts if they would have suddenly developed this stuff on HTML5/WebGL instead of awkward plugins

P. Fonseca

1:52 PM

@kirma I think that the potential of an online interface is huge! Present the capacities of the tool, without people needing to install it. Play in any device that has a browser. Eventually easily access the processing capacity of a super computer, just by running on an online interface. Easier distribution of access to schools. Direct simultaneous collaboration, etc, etc.

Yves Klett

@halirutan you got me there. As for girls, one ring... :-)

@halirutan but just think of a cross between Whatsapp and Mathematica.

halirutan

@YvesKlett hehe

kirma

@P.Fonseca I'm a bit skeptical of fluidity of the user interface, even if it actually has full capabilities. Last time Mr. Wolfram demonstrated the web interface, it clearly didn't have native graphics functionality in the browser, for instance.

P. Fonseca

@kirma I never saw it, besides some print screens made publicly available

Nevertheless, the potential is huge...

Yves Klett

@kirma the demo in Champaign was fun, mostly because it really did not work at all

kirma

1:58 PM

@P.Fonseca on some SXSW presentation, which should be pretty googlable.

couple months ago

P. Fonseca

@kirma I think I saw that one, but it was a very short thing, and I don't know how it relates to what they intended to show...

kirma

@YvesKlett Not very promising if that service would be launched on anniversary ;)

J. M.

@Mr.Wizard "this question is worthy of attention, though perhaps an answer would take too much work?" - yes and yes.

Yves Klett

@kirma not really, but you know - demos.

kirma

@P.Fonseca the impression it gave to me is that usability has ways to go. Javascript canvas/WebGL driven graphics frontend is mandatory before such an UI can be compared with native UIs. The fortunate thing is that Wolfram Research really should have solid basis to implement that stuff with sufficient time.

P. Fonseca

2:04 PM

@kirma so we are left with tablets for this 25th anniversary?

kirma

@P.Fonseca don't ask me. :) I'm just the pessimist in making sudden great leaps as I've always been. ;)

P. Fonseca

@kirma, actually, we all know that the ones who know about it all have NDA, and so, this is just plain guessing :)

kirma

@P.Fonseca Yep. Still: look at the web at the moment. How many (ported) technical applications that work truly nicely on plain modern browser have you seen this far? I think I haven't heard of any beyond some editor/office/development tools that are mostly text-driven or very limited in their extent of graphics.

J. M.

"Look at that schmuck. Thinks he's so fancy making them colorful plots on his tablet..."

kirma

I do wish things happen quick and become useful in a similar manner, but wishes and expectations are two different things.

J. M.

2:11 PM

@kirma Most of the online flowchart makers I've tried sucked in one way or another.

kirma

@J.M. Unless Mma FE is mostly made of horrible C/C++ spaghetti, it shouldn't be impossible to transport it into different runtime environment. Well, buggy runtimes (all that javascript and API implementation mess) may be a considerable roadblock.

J. M.

@kirma "Mma FE is mostly made of horrible C/C++ spaghetti" - That is probably one thing we can ask Fultz, should he come over...

kirma

@J.M. I bet such implementation details are company confidential trade secrets! At least if I'd work there and know it is horrible spaghetti... :-P

J. M.

@kirma Fair point. :D "We built our application on top of spaghetti, all covered with cheese..."

kirma

@J.M. Thankfully that (possible) spaghetti is on the top, not on the bottom

:)

... off for a while. Or maybe even longer than that. ->

J. M.

2:21 PM

@kirma For that purpose, I use "off for a few" and leave the guessing of the units to the reader. See you!

Szabolcs

3:10 PM

@Oleksandr, are you around?

Oleksandr R.

3:46 PM

@Szabolcs hi. It seems to work only if you do it inside the front end. This shouldn't be necessary in principle but IIRC MathLink`AddSharingLink used to be dependent on JLink but now doesn't seem to be; could be the functionality was moved into the FE I suppose. Let me see.

Szabolcs

@OleksandrR. Sorry, I wanted to write a few more details but I was interrupted.

Oleksandr R.

@Szabolcs oh, of course... MathLink`AddSharingLink alternates $ParentLink between the alternatives given, but in the terminal $ParentLink has no value and the kernel does not pay attention to this. The function has some other options which may do the job, but I don't know off hand how to use them.

Szabolcs

@OleksandrR. Is it normal that LinkActivate in kernel B (green) blocks until LinkActivate in kernel A? You suggested to do the evaluations in a strict order, so I did. Also, is LinkActivate actually necessary? There seems to be little documentation in it except for "Network Programming with Mathematica" (google books)

Or is LinkActivate maybe deprecated as of ML level 3?

@OleksandrR. got it

Oleksandr R.

@Szabolcs yes, that is normal. You must use LinkActivate. (MLINTERFACE 3 was actually introduced in 6, I think)

@Szabolcs Network Programming is where I learned about this, too. It is undocumented otherwise, I don't know why.

Szabolcs

@OleksandrR. Got it, so I have to make sure $ParentLink has a setting. That means I should be able to launch kernel B from kernel A using LinkLaunch, and also connect to kernel B using kernel C. Then both A and C can request evaluations from B.

Oleksandr R.

3:59 PM

Yes, I think that should work. Probably there is some other way to make the kernel pay attention to this but I don't know what it is.

Rojo

6:50 PM

@Szabolcs You around?

Szabolcs

8:11 PM

@Rojo Hello

Stack Exchange

8:22 PM

Convert an image to one of the newly supported CIE color spaces: ColorConvert[ExampleData[{"TestImage","JellyBeans"}],"LAB"]

Tweeted by MathematicaTip on May 15, 2013 at 8:14 PM

hhh

8:43 PM

Does Mathematica have some settings to control the environment variables?

(I cannot understand why I see some output in a very-very small size so very hard to read while @Mr.Wizard Mathematica apparently not)

I am trying to debug this, mathematica.stackexchange.com/a/23562/2000, I cannot understand why my Mathematica 8 and Mathematica 9 result into different results as with Mathematica 7.

(I don't have Mathematica 7 to test it, sadly ... but it looks like the results between 8 and 9 are about the same except the output color different)

@Mr.Wizard You can see the output of your code with Mathematica 8 and Mathematica 9. (I am trying to get them larger the images but doing it over SSH is pretty slow...)

Guillochon

Hi guys, brief question...I have a list plot with InterpolationOrder->0, but it's not plotting all descending lines:

hhh

Anyway the third from top-left is an error at least in both implementations, @Mr.Wizard.

Guillochon

you can see it in the blue dataset

hhh

@Guillochon can you provide the data? Is it corrupted?

Guillochon

No it's not corrupted

However I think it's actually because it's a logarithm of bincounts, and some bins are zero...

so the value is indeterminate

I guess I just need to replace the indeterminates

hhh

8:59 PM

@Guillochon small epsilon

Guillochon

Yeah that was it, fixed

hhh

Good :)

Szabolcs

9:48 PM

Something very weird going on here:

0

A: Does SetSharedVariable allow simultaneous reads?

To answer your actual question, SetSharedVariable does not allow simultaneous reads. It forces the variable to be evaluated on the main kernel, effectively disabling the parallelization. A more interesting question for me is: why is ParallelMap so slow when not using SetSharedVariable? This o...

Szabolcs

10:23 PM

@OleksandrR. Can you take a look at this? You know a lot about the parallel tools. Have you seen this problem before? --> mathematica.stackexchange.com/a/25268/12

Transcript for