12:00 AM
is a tensor of rank 2 (aka a matrix). It perfectly fine to use it as input.
> given that you always specify all the variables as certain types
Usually, yes. I'm not sure I ever used N inside a compiled function because types. Types are derived in most calculations.

Isn't
{{1}, {1,2}, {1,2,3}}
a tensor of rank 3 - it is just that its indices run over different numbers.

In[10]:= TensorQ[{{1}, {1, 2}, {1, 2, 3}}]

Out[10]= False

I see. Ok, maybe I have to revisit the mathematical definition of a tensor then, before proceeding

12:18 AM
Basically it's not that hard. You can ask yourself one of two (hopefully) equivalent questions:

1. Can you specify the Dimensions of your tensor down to the last level
2. Do all elements on the same level have the same length

@halirutan Thank you!

7 hours later…
7:25 AM
I wonder if WL should even have For. I see many questions where users are obviously coming from a different language and are confused that For isn't the standard looping construct
(And I feel like there are many questions that are like "Do some specific thing within a For loop")
I For one don't think I've ever used For...

@CarlLange Yeah, that's the main argument I was trying to make here: mathematica.stackexchange.com/a/134610/12
> The problem with For is that it hinders learning and makes it very easy for beginners to introduce mistakes into their code.
(Not that it is procedural / not functional.)

2 hours later…
9:03 AM
@ΑλέξανδροςΖεγγ has edited the word Mathematica in many question titles to $\textit{Mathematica}$. Recent example. This is clearly done for cosmetic reasons. Unfortunately, it hinders searching. Example search that no longer gives the most relevant result. I think we should not do cosmetic edits if they cause trouble elsewhere (searching).
I cannot ping him in chat to bring this up though. Can @halirutan or another moderator help me and ping him here so we can discuss the issue?
If anyone else has any input on the topic, please comment. Maybe people don't agree with me on this.

9:31 AM
@halirutan OK, thanks!
@Szabolcs Understood.

9:50 AM
@ΑλέξανδροςΖεγγ Thanks so much for all the work you put into editing! :-) This is just a special situation where it seems to me that avoiding a specific type of edit is better.

3 hours later…
12:39 PM
Gee, don't remember ever realizing this: Cos[Sqrt[x]] is holomorphic.
Plot[Cos[Sqrt[x]], {x, -1, 1}]

2 hours later…
2:56 PM
Woah that’s awesome @MichaelE2!!! Hey general Q, does anyone know or can provide reference, what is better for assumptions? _Integer Or _?IntegerQ ? Is it system dependent? Or is one definitively better than the others? Integer can be any assumption type.

3:30 PM
sounds like an exciting DSP discovery - fast inverse chirp z-transform.
https://www.nature.com/articles/s41598-019-50234-9

3 hours later…

1 hour later…
7:30 PM
@Szabolcs Re: our discussion yesterday - there's a footer that shows up when you hover over the column headers in a dataset that explains the level that it's at.
I guess it could be clearer since I use dataset all the time and never noticed it, but there is something that exists.

3 hours later…
10:14 PM