Wolfram Mathematica

« first day (1489 days earlier) ← previous day next day → last day (2978 days later) »

7:53 AM

It would nice to have more general += -= operators, like:

data // = f

which would work like:

data = f[data]

Don't you think guys?

halirutan

8:46 AM

Creating a decent looking game board with Mathematica takes a lot of tweaking. And if you still want to rotate it, you have to make serious cuts in the quality.

I guess the only reasonable solution is to import the model of the board, because the main factor is the complexity of the mesh that is created.

Btw, is someone coming to the German Mathematica Day this Thursday the 18.02.?

Kuba

9:02 AM

@halirutan Rolf and me

Yves Klett

9:46 AM

@halirutan would have liked to, but cannot make. Have fun, though!

Szabolcs

10:27 AM

@halirutan What version of Linux are you using currently?

@halirutan Could you (if you have time) test IGraph/M 0.1.4 on your Linux system? github.com/szhorvat/IGraphM/releases Just download, extract, drop in ~/.Mathematica/Applications. No additional setup should be necessary. Then load as <<IGraphM` and test with IGVersion[].

Some people tell me that the library won't load on their system and I can't figure out why. I also don't really understand how to make portable binaries for Linux. I though that if most things are statically linked into it then it should be fine. This is what ldd says about my binary now:

$ ldd IGraphM.so
	linux-vdso.so.1 =>  (0x00007fffecdf0000)
	libgcc_s.so.1 => /lib/x86_64-linux-gnu/libgcc_s.so.1 (0x00007f8e7a398000)
	libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f8e79fce000)
	/lib64/ld-linux-x86-64.so.2 (0x000056390dc55000)

Compiled on Ubuntu 15.10, I can't see what is in there that will prevent this from working on Ubuntu 14.04.

1 hour later…

Szabolcs

11:34 AM

I wrote a question in Community:

community.wolfram.com/groups/-/m/t/793461

@OleksandrR. @halirutan or anyone else with C++ and compiling/linking experience, can you take a look please?

Szabolcs

12:04 PM

Is there anyone around who Mathematica uses Linux? @Kuba? @kirma?

Kuba

12:21 PM

Nope, sorry :-/

kirma

12:48 PM

@Szabolcs I used to use it on Linux, but frankly not for three years now on practice... OSX on MacBook Pro is just much more practical.

kirma

1:03 PM

I either don't understand color handling in Mma, or in general. Getting LCH colors to show up consistently through sensible conversions seems next to impossible, even when they're on display sRGB gamut...

I'm trying to order color 3D prints, and frankly, I can only wish the best.

Kuba

1:14 PM

Q: Reshaping associations

Let's say we have a set a\of associations: dataset = { <|"type" -> "a", "subtype" -> "I", "value" -> 1|>, <|"type" -> "a", "subtype" -> "II", "value" -> 2|>, <|"type" -> "b", "subtype" -> "I", "value" -> 1|>, <|"type" -> "b", "subtype" -> "II", "value" -> 2|> } where every entry is u...

1 hour later…

Szabolcs

2:29 PM

@OleksandrR. @halirutan Update to my questions above: it appears that recompiling it with gcc 4.8.5 (still on Ubuntu 15.10, but not with the default gcc 5.2) allows it to work on 14.04. Unfortunately I do not understand why, and whether it has to do with this.

Wolfram Blog

2:53 PM

Wolfram Blog: Mathematica Online Now in Full Release

posted on February 15, 2016 by Andrew de Laix

If you have recently visited Mathematica Online, the cloud version of our flagship software, you may have noticed something missing. That’s right—we dropped the “BETA” tag, and I am pleased to announce that we have a product we can proudly call release ready. It has been a long road from when we debuted the Wolfram [...]

Kuba

challenge accepted... :)

R. M.

A: Highlighting Mathematica code in $\LaTeX$ document

I too had a need for a better syntax highlighting engine for Mathematica that can be used in different formats (so the javascript plugin is ruled out), so I wrote a better lexer and highlighter for Pygments than the one that ships with pygments. From the README: It can currently lex and highl...

If this gets used/people like it, hopefully it'll be included in the official pygments repo.

Szabolcs

3:24 PM

@R.M. I think I could use that. You should post it on packagedata.net as well.

R. M.

3:34 PM

@Szabolcs submitted

Oleksandr R.

4:08 PM

@Szabolcs I'm sorry, I don't think I can be of any help. I haven't any experience with C++ at all, really

Martin Sleziak

4:23 PM

Are questions of this type on-topic at Mathematica.SE: Memory limit hit: optimize code?

Should I (or somebody else) leave a comment for the OP, that he is more likely to get some help at this website. (Since it is specialized to Mathematica, as opposed to Math.SE which is for questions about mathematics in general.)

J. M.

4:39 PM

@MartinSleziak I flagged it for migration.

Martin Sleziak

Thanks!

I wanted to ask somebody more experienced with this site first. (Since if a migrated question is closed, it is "bounced back". So it would not make much sense to migrate if it is not the type of question which would be accepted on this site.)

J. M.

@MartinSleziak Yeah, you're right. Thanks for asking here!

halirutan

5:02 PM

@R.M. This is nice.

If IntelliJ fixed the core environment so that we can use the plugin without Idea itself (and without using old Idea code like in the Wolfram-Parser), it should be not too complicated to set up something similar with the complete force of highlighting.

I guess one easy way would be to just put all codesnippets into one directory and transform each file to a corresponding colored tex-version that can be used with \input.

halirutan

5:41 PM

@Szabolcs I won't be able to do something this week. If it helps you, that I can probably look at it next week.

Szabolcs

@halirutan Thanks! That's useful too.

3 hours later…

Szabolcs

8:50 PM

Could it be that MathematicaScrip/WolframScipt is none other than the old Mathematica Scripting Hack?

What's very interesting is that if I look into MathematicaScript (an executable binary), I find references to mash.c. That rings a bell! — Szabolcs 43 secs ago

Michael E2

9:35 PM

@J.M. Did you know that InterpolatingFunction can store a piecewise (truncated) Chebyshev series interpolating function?

R. M.

10:06 PM

@halirutan I started this project with the intention of using the java parser but that became too unwieldy and complex to work with, package, distribute, etc.

But more importantly, I realized after doing this that a good parser is only a small part of the work. In order for this to be really useful, you'll also have to create formatters that take the stream of tokens and annotations and then produce output in a specific format.

And now you're working with HTML, TeX, JSON, YAML, XML, etc. and getting your hands dirty in each one of them just so that the parser can be used.

Once you look at it from this pov, you start appreciating how much work has already been put into the pygments ecosystem and how widely it is used, so the optimal solution here was to compromise on the highlighting quality a little but try to push the lexer to its limits so that we can let pigments do the rest of the work and have something that can be used immediately.

Also, for something that is meant for static display (e.g. html/tex documents), a full blown parser feels like a bazooka... sometimes you might not actually have a syntactically valid piece of code in these snippets.

But I do love well formatted/highlighted code so I really hope that we can make a simple stripped down parser that can fit in the pigments framework and can highlight lexically scoped variables. That should cover most of the benefits a parser would bring to a highlighting job.

1 hour later…

J. M.

11:36 PM

@MichaelE2 I must say that I don't... how did you come across that tidbit? Are they now trying to sneak chebfun in? ;)