Using #0 as the second argument to GroupBy is certainly powerful, but it's really easy to accidentally get into an unabortable infinite recursion doing it :|

Not really sure what's going on in this question: mathematica.stackexchange.com/posts/216840/revisions I reverted back to the original for the moment.

@CarlLange wat

I have no idea what happened there, but I like it

@b3m2a1 Two people using the same account?

@CarlLange Or someone left their account logged in somewhere

@b3m2a1 I have seen this topic but this was the first time I've faced this feature. Didn't have time to test it though.

@user21 Have you seen this library? libfive.com

@Szabolcs, no I have not seen this. It's bit hard to see from the web page what the scope of this lib is. Have you used it?

@andre314, that is correct. Maybe you can post a question and I'll answer it there.

Ok, I will post a question

@user21 I've used it very little, but it seems amazing. It comes with the "Studio" application which you can experiment with. You can think of it as supporting most exact regions that Mathematica has in 3D (ImplicitRegion, ParametricRegion, etc.), most primitives, as well as CSG, but working much, much much faster than Mathematica. It also generates nicer looking meshes, and has more 3D manipulation operations.

@Szabolcs, thanks I'll take a closer look then.

@andre314, I got to go now, to get a finer mesh boundary make use of ToElementMesh. For OpenCascade, the behavior you describe is correct and wanted.

@CarlLange Hi. This is Jose, from Wolfram. We are aware of an unacceptable slowdown in some Dataset expressions, due to a bad dynamic interaction with the summary boxes of some objects. Both the TimeSeries objects of the coronavirus datasets and the InterpolatingFunction objects in the example given before typeset via summary boxes. We are working on a solution and will release it as soon as it is available.

@jose Thanks for the update. The example I posted does not have any summary boxes though, and it still makes the FE quite literally unusable. Import["https://covid.ourworldindata.org/data/ecdc/total_cases.csv", "Dataset", HeaderLines -> 1]

Are there any plans to deal with the general slowness of formatted content in the FE? I mean the slowness that was already present before 12.1. Just having the above dataset displayed does not cause problems in 12.0, but scrolling it is barely possible. More and more objects get summary boxes (such as Polygon in 12.0) which makes me more and more cautious about displaying anything. It risks at least a temporary FE freeze.

Meanwhile, web browsers can display many more complex objects and still scroll perfectly smoothly. Brower-based notebooks like Jupyter inherit this. Displaying a table with a few hundred lines? No problem in Jupyter.

If anyone has any feedback or complaint about MaTeX, please let me know ASAP. I want to make a new release soon. ( @EmilioPisanty )

@Szabolcs sorry, I was in a deadline crunch last time and didn't have time to chase loose ends, and then the Situation intervened

(the actual situation, not the person, whoever he actually is)

it's working fine on my main notebook

Congrats on the paper!

but I just tried it on a fresh notebook and got the same

@Szabolcs thx =)

Weird

Doesn't look like it's coming from MaTeX code

Don't close that notebook @EmilioPisanty !

@Szabolcs still open =)

happy to run any diagnostics on it as per your directions =)

On the first message, click the (...) and choose to see the stack trace. ANything useful there?

Interpreter[Number][{11.74997}]

{(Interpreter[Number][StringCases[This is XeTeX, Version 3.14159265-2.6-0.99999 (MiKTeX 2.9.6730 64-bit)

entering extended mode



(C:/Users/episanty/AppData/Local/Temp/MaTeX_rrnxennzevvaajln/oxvvodgqselsxots.t

ex

LaTeX2e <2018-04-01> patch level 5

("C:\Program Files\MiKTeX 2.9\tex\latex\standalone\standalone.cls"

Document Class: standalone 2018/03/26 v1.3a Class to compile TeX sub-files stan

dalone

("C:\Program Files\MiKTeX 2.9\tex\latex\tools\shellesc.sty")

then a long LaTeX log file

@jose Hi Jose, everyone else has already said what I was going to say, so here's a new thing: great work on all the Geo functionality in WL! It's such a pleasure to use.

well, not that long

there's three copies with similar Mathematica headers

@EmilioPisanty WHat happens if you evaluate Interpreter["Number"][{"11.74997"}] in a fresh kernel?

the last one ends with

, RegularExpression["MATEX" <> #1 <> ":(.+?)pt"] -> "$1"]] &)[
  "DEPTH"]}

It looks like the problem is with Interpreter. You might have a broken paclet. Just to be clear, I see no problem with M11.2.

No, wait!

I can reproduce it too.

@EmilioPisanty Thanks! I have enough, I can take it from here. And this is definitely an M bug (almost certainly a bad paclet), but now I have enough to report it to WRI.

O.o

That looks pretty messed up to me

from the documentation of Interpreter, that looks like a perfectly acceptable input

O.o

Interpreter["Number"]["11.74997"]

works fine

O.o O.o O.o

.... as indeed does Interpreter["Number"][{"11.74997"}] upon a further evaluation

that's bonkers

good luck with WRI =)

also bonkers

OK, no fancy numbers required

it looks like all the stuff with Fold comes from having Interpreter["Number"]["1."] fail, and then this makes the threaded evaluation go crazy?

anyways, I'll stop here, @Szabolcs

good luck with this stuff =)

I can't really make a minimal example at this stage, as I have no clue what is changing where. I tried testing the values of all things that I define and use before and after adding the comment and I can't find any differences.

One thing I should probably say is that this behaviour occurs when I am testing my solutions for convergence, so the numbers are around 10^-10 and my calculation involves a lot of numerical operations. Is it possible that there is some randomness in the way Mathematica rounds numbers near MachinePrecision and adding the comment kind of changes the random seed or something?

@LeonidShifrin Good to see you virtually!

@halirutan I don't know how it happened that I was here :)

But, good to see you too!

@Feeds @C.E. I think spammers have found PackageData...

@b3m2a1 They started breaking through reCaptcha some time ago, unfortunately. The posts never get published because the site has a review mechanism but the RSS picks up on them anyway. I should fix that. I apologize of the inconvenience.

I don't really mind. It amuses me.

I also enjoy reading my spam email, though, so clearly there's something wrong with me.

lol

@b3m2a1 correction: spammer

;-)

Does anyone with 12.1 know if HatchFilling works with BarCart?

@MikeHoneychurch I believe there was an update to an answer showing that it does ~a day ago

@Szabolcs I have always found Dataset problematic for anything but small data. I am using lists and finding that more efficient. And for stuff too big for lists using python

@MikeHoneychurch ah it was this: mathematica.stackexchange.com/a/216567/38205

Not exactly a BarChart but presumably same underlying code

@b3m2a1 thanks! I hadn't been able to find it in the chat!

Ah it wasn't here that I saw it. I was just scrolling through the questions one day