Is there a way to turn off the formatting of Root objects? I can't easily tell at a glance if two are the same or merely close....

@MichaelE2 you could use Ctrl+Shift+e possibly?

@CATrevillian Did you mean Cmd+Shift+E (I'm on a Mac)? Crtl+... typed an invisible character (I think). But your suggestion led me to try various Cell > Convert To menu commands. If I first put it in InputForm and then back into StandardForm, the root objects are not reformatted.

So I guess I have a halfway decent workaround. I really was wondering if it could be turned off through a FE option.

@MichaelE2 hah I have the least experience on macs, how curious it changes from Ctrl to Cmd though, it is the shortcut to ShowExpression, IIRC, using it on a notebook reduces it to what I incorrectly call raw form, It has all the boxes n what-nots, I use it to copy and build documentation pages because I refuse to learn new methods and it also is super fun like source code viewing on a website :D ref/menuitem/ShowExpression

What about FullForm?

@CATrevillian I wouldn't want to display the entire expression in FullForm. The roots are part of a bigger answer. I just want to be able to read which root each is and what it's a root of. The numeric value displayed in the summary box has only a couple of digits and isn't that important sometimes. They have tooltips, but I can see only one at a time, and while I can magnify the notebook to suit my eyes, the tooltips stay the same size.

It's just a niggling irritation. Not that important if the solution requires too much work.

@MichaelE2 I recall a use of Epilog to make tooltips appear in a different manner, it was for geotagging of locations on a globe, you might not be able to automate it as easily, but I feel that if we can have them be affected by a mouse click and drag, then the functionality should be right there to have a hover-like tooltip or even to click and then have it stay showing what you want, then clicking again to make it bigger to help with the visibility? I'm just a rank amateur, though.

Hah right no I agree, lazy is the best science (shortest path, after-all)

@CATrevillian For now, unless there's an option I can set, Cmd-Shift-I-N is pretty short and sweet.

Hah I just figured out why my shortcut efforts for `Superscript[[pl],[Prime]] were so fruitless! Darned iconizing--what exactly will that combo do you referenced @MichaelE2? I could poke around for hours I'm sure, immensely curious, though!

Oops i didn't close off the Superscript[[pl],[Prime]]

Superscript[\[pl],\[Prime]] got it hah

@CATrevillian Cmd-Shift-I converts to InputForm, Cmd-Shift-N converts to StandardForm. I can hold down Cmd-Shift and hit I, N with my right hand.

@MichaelE2 wonderful! Thank you :D Hope we can find something decent, I want to implement tooltips (many of them!) too! One last strange idea: you can output graphics primitives with it, yeah? Should you then be able to scale them for better visibility? Graphics Primitives of, say, your desired Form to view them in? It is entirely possible I missed the point you were after, too!

@CATrevillian I was ignoring the tooltips in the root objects because you can have only one showing at a time. Make it hard to compare two or three roots. The fact that they stay small isn't so bad. Of course you can style your own tooltips to be larger or smaller. The system ones might be set via a stylesheet.

@CATrevillian Yeah, the tooltip is in the "NumericalApproximation" style in the Core.nb stylesheet.

Proving myself that elevator service times indeed follow gamma distribution by measuring morning walk time dominated by them dozens of times...

@MichaelE2 ah okay, this makes sense! So you can change it with ease? Or would that output not how you want it to, not giving enough information as before? And @kirma oh my wow that is a pretty graph and what it represents is statistically hilarious! I think? Hah

@CATrevillian Black line is the empirical probability of travel time, filtered using one-minute moving average window, light blue area is 95% confidence region of corresponding generalized gamma probability distribution fit. They agree sensibly.

I have three elevators on my way from home to metro platform and not very much walking... and those are measurements of that journey...

@kirma so the conclusion is then that--oh wow I see! I wonder if the skew has some dependence on the combination of the 3 elevators. As I read it, you are less likely to be running late, but somehow they will more or less line up in some coherent fashion, is it that the closer ones to the metro run quicker than near your home? And that there is a greater variance near your home, as fewer people would go a larger amount of different places, than would the greater number of people that are all

more than likely travelling to the metro?

@CATrevillian I'm not quite certain what should I conclude - especially since I just measure the combined time - but I've read that many elevator systems run on something close to a "minimum variance algorithm", and that theoretically that results a gamma distribution for customer service times (other algorithms result different distributions). And it seems to be the case, generalized gamma distribution is clearly the best fit of those that Mathematica knows.

I think comparing different elevators is a bit hard since first is just a single elevator, second is a group of four, and third is group of two.

@MichaelE2 I don't see a way to turn off the formatting just for Root objects, but if you set BoxForm`UseTextFormattingQ to True then you get the old root boxes

@CarlWoll @JasonB. Thanks!

Not sure why the title came across that feed for the livestream. The actual topic is probability and statistics features in Wolfram Language.

Also, Stephen will be livestreaming later today at 3:30 Central. We'll be continuing the review of entities in the Wolfram Function Repository.

how can define a function to work on arguments that fail (for example) OddQ? I thought f[x_?! OddQ] would work but it doesn't

@ChrisK f[x_?(Not@*OddQ)] or f[x_?(!OddQ[#]&)] or f[x_?(Function[Null, Not@OddQ[#], HoldFirst])]

@b3m2a1 thanks! I don't know if I'd come up with that myself...

The last of those options is a common pattern when tests need to apply to a held version of the symbol, e.g. when f is also HoldFirst.

Your basic issue was two-fold. 1) ? has crazy high precedence 2) (!OddQ)[x] =!= !(OddQ[x])

If you look at the full form of f[_?! OddQ] you see the issue:

f[PatternTest[Blank[],Not[OddQ]]]

@CATrevillian The keystrokes work fairly easily, but I have to do it every time. BoxForm`UseApproximations = False seems nicer in that way. And I can toggle it on/off as desired.

Interestingly, ! has higher precedence than ?

Ah wait no it's just that _?! is syntactically invalid... which in some sense is ! having higher precedence than ?.

! is a unary operator and ? is binary so the precedence parser will parse the entirety of !OddQ first.

@MichaelE2 that should be something we can tie keybindings to fairly easily using @b3m2a1's implementation simplifier, yeah? And @b3m2a1 these are awesome tricks, I love messing with the precedence (when I know what that means, haha, I forget things way too easily ]; )

Anyone wanna suggest a different keystroke combo for a supered prime/add shift for supered double prime?

damn, -> vs :> got me again

I suppose admitting I have a problem is the first step

Also, please feel my pain, I just spent an hour and a half (a paltry amount of time, I know, but still, I was on a roll!) looking for one comma on line such-and-such, even went so far as to download eclipse and fail to open my ,nb in it for edit, tried to use GitHub, finally I just evaluated it all and somehow got lucky to find the one pink comma!

@ChrisK aha! Hah.

@ChrisK Just use :> then :)

Must say that I am far less likely to submit to the function repo now that I know slightly-harsh code review might be livestreamed

But isn't that the equivalent to going up against a belt-holding boxer, and getting beat up on primetime @CarlLange? It's still pretty cool! Hah! c'monnnn You're on TV!!!

@CATrevillian I dunno, I prefer posting in the chat. I don't have enough emotional bandwidth to put stuff out there only to have it potentially blasted. Same thing as emailing my local politician vs running for election ;)