@b3m2a1 I was trying to fool myself into making a packlet again, but I wondered if the basic format of the init file was included in the docs or not, I’ll find the exact page again when I get back to my pc, I was just clicking my way through the package pages and what not, I think I was on a guide and that was listed as a related workflow page. It was the only one listed, though.

Something related: community.wolfram.com/groups/-/m/t/935450

Can someone suggest a short snippet of M-code which is hard to understand. Something like this useless thing

f @@@ g @@ h /@ RandomInteger[{-10, 10}, 10]

The Y-Combinator ?

So no "true" or "simple" method to something like a ListLinePlot3D, hey? Just gotta choose something to be a line? Or does ListPointPlot3D with the /.Point->Line seem to be a good one to do?

@andre314 I was not precise enough: What I want is an expression that is hard to understand because it uses a combination of operators where you need to know the precedences in order to the the evaluation order right.

For me the most unobvious precedence is with : a // f /. a -> b .

@halirutan I'd expect not, but I would benefit from seeing a discrete cross-section-like profile change across a parameter iteration, and you can be given false information with a surface plot, whereas with a bunch of points, the obviousness can be lost in translation during presentation

@andre314 The reason I'm asking is that I have implemented a "Parenthesonator" which displays in which order an expression is evaluated

and I need a good showcase.

@CATrevillian Hmm, I'm not sure I understand how the result should look like.

@CATrevillian Let's say you have something like this

How do you want to connect the points with lines?

One obvious way is

Graphics3D[{Line[#], Line[Transpose[#]]} &@
  Table[{x, y, Sin[x + y]}, {x, 0, 2 Pi, Pi/5.}, {y, 0, 2 Pi, Pi/5.}]]

But this assumes you have the points in a certain fashion. ListPointPlot3D can handle unstructured points as well.

Ahhh ok I see what you mean @halirutan, my hope is that I would be able to connect say like, one direction or the other if one way is one parameter and the other way is the other parameter, but this gives me precisely what I'd hope for. I'd just pop that in a frame with ticks and I

'm good to go!

@halirutan Length@{1,2,3,4}/2 is not clear too

The presumption is that one way is one drag of parameters, and the other way is another drag of parameters so that the axes are corresponding to each parameter set. @halirutan the parenthesonator looks like it will be great!

@andre314 Yep, that's a good and simple one. Thanks.

@halirutan If you don't care about how realistic the example is, how about something like f[a -> b & // f@*g[x]~h~# @@ # &]]?

@LukasLang A real example would be nice, because my plugin already catches a lot of the suspicious stuff and gives warning messages.

@halirutan Here's one that's pretty close to what I sometimes use: Prepend[#]@*List /@ {##2} & @@ Range@10~Join~Range[100, 110]// MatrixForm

@LukasLang That is an excellent example!!

@halirutan How about 2x / 3y - 1 /. {x -> 9, y -> 2}, classic bane of beginning algebra students everywhere.

@WReach Thanks. I guess I need about 3 examples and you were extremely helpful :)

Does the combo of Scan and Table save time in addition to saving RAM?

One can only imagine that it does, but if it doesn't save time, I won't waste mine figuring out how to apply it!

This question brought to you by the apparently slightly faster application of Table over odd or even indices versus just choosing them as such within Part...?

Has anyone been bitten in v12 by too much data in a message crashing the kernel...? I've had like 3 kernel crashes in the past hour because Part emits a message...