Mathematics - 2021-06-20

12:21 AM

should have invested in lesliecoin. it only goes up.

12:41 AM

the sub title from last week's economist article on said topic: cryptocoins are proliferating wildly. what do they all do?

Yorch

can I buy lesliecoin with monero

and more importantly, can I buy monero with lesliecoin?

leslie townes

once obtained, lesliecoin is not transferrable. we're working on that

we've worked out how you buy it. selling it is phase 2

Yorch

that's unacceptable

leslie townes

you'll want to get in on the ground floor. it's like being a pre-IPO investor

Yorch

I'll stick to my prefered currency

mse reputation

leslie townes

12:54 AM

i'll post a suggestion on meta that reputation be transferrable. let the market sort it out

Yorch

mse reputation is mined via crossvoting, but sufficiently complcated methods must be used to avoid deletion of the mined asset.

leslie townes

haha

copper.hat

i have a spac that you might leverage for stage 2 of lesliecoin. some terms of course, only slightly onerous

Ted Shifrin

@Yorch Far too complex for me.

shai horowitz

1:13 AM

Can anyone tell me what the 3 adic representation of 1/3 is? Really naively its .1 but decimal points in padic seemed wrong to me

Yorch

P-adic number

In mathematics, the p-adic number system for any prime number p extends the ordinary arithmetic of the rational numbers in a different way from the extension of the rational number system to the real and complex number systems. The extension is achieved by an alternative interpretation of the concept of "closeness" or absolute value. In particular, two p-adic numbers are considered to be close when their difference is divisible by a high power of p: the higher the power, the closer they are. This property enables p-adic numbers to encode congruence information in a way that turns out to ha...

the link is to the positional notation part, not the general article

shai horowitz

Awesome some how I missed that part despite reading the article multiple times shrug thanks

The k<0 part that is

Yorch

no need to thank me, I am a bot.

shai horowitz

Bonus question. Does anyone know of an online calculator that I can input a rational number and prime p and get the padic representation?

leslie townes

1:30 AM

i was googling 'does wolfram alpha support p-adic arithmetic' and the google autocomplete was 'does wolfram alpha snitch'

i don't know what to make of that

maybe that's my default autocomplete for 'does X ___' because of my considerable personal use of google to determine who's a snitch

shai horowitz

Everyone should know wolfram very sus

leslie townes

anyway wolfram alpha seems to not know p-adic expansions

shai horowitz

Yeah nor mathematica from what I've read

I may end up just trying to code one which will without doubt result in more questions here

leslie townes

pari.math.u-bordeaux.fr/gp.html appears to be an online pari/gp interpreter although who knows what packages it has loaded or whether it is full featured. i do think vanilla pari supports p-adic stuff

shai horowitz

Thanks for the lead, I'll give it a shot

Ted Shifrin

1:55 AM

The story of the moral is: Law school taught Leslie to google.

robjohn

2:11 AM

@shaihorowitz here is a Mathematica notebook from the Wayback Machine for p-Adic stuff.

PM 2Ring

That discussion on the possible danger of comparing truncated series reminded me of Borwein integrals: johncarlosbaez.wordpress.com/2018/09/20/… Also see math.stackexchange.com/q/111440/207316

leslie townes

2:41 AM

the comment thread to mathoverflow.net/a/11607 is pretty funny. apparently john borwein reported the sinc integral pattern as a 'curiosity' to maple and someone there (who participates in the thread) had to do the work of figuring it out.

PM 2Ring

Ah, right. I'd forgotten about that little prank.

On a related note,

Also, I hear the 4th root of (9^2 + 19^2/22) is pi.

I think I first heard about Borwein's prank in the xkcd forum thread for that comic. But maybe it was in that MathOverflow thread. ;)

Ted Shifrin

3:10 AM

This prank is over my head.

leslie townes

3:29 AM

it's a mean prank because he played it on his own team. that does seem to defeat the purpose of being on a team. that hat guy is no good.

play it on a friend who is doing homework, or something.

ted: roughly, hat guy wasted a lot of his team's limited time in a programming competition by causing people to believe that some aspect of their solution-in-progress, when tested, was generating errors when it was not.

by telling them, oh, test it on e^pi - pi and see what you get. oops that's wrong.

Ted Shifrin

Oh, basically a known bug?

leslie townes

if you'd been told that e^pi - pi was 20, you'd think that getting 19.9990... was a bug.

it's hard to imagine an ACM problem that would actually turn on floating point arithmetic. the problems i have seen tend to turn on other stuff. but the premise is kind of funny.

Ted Shifrin

This is just so far removed from my sphere of experience or knowledge ...

But now I'm putting this into Mathematica to see what happens :D

leslie townes

[smoke begins pouring out of ted's computer]

Ted Shifrin

OK, so I got 19.999099979...

Interesting nerdliness.

Yorch

3:41 AM

@leslietownes unfortunately floating point precision sometimes appears in some regionals

PM 2Ring

4:04 AM

You get numbers that are very close to integers from $exp(\pi\sqrt{n})$, where $n$ is a Heegner number. As Wikipedia mentions, Martin Gardner did an April Fool's prank using the value of the above expression for $n=163$. You can't calculate it properly using simple 64 bit floats, you need higher precision. It's ~262537412640768743.9999999999992500725972078

If you want more digits, you can edit this Sage script

Some non-prime values of $n$ also give near-integers, eg $n=25$ gives ~6635623.999146

copper.hat

6:11 AM

time to repeat my xkcd cunnilingus story ...

what, nobody here on a saturday night?

hyper-neutrino

i'm here technically o/

copper.hat

:-). just got home from a dinner party, the early hour reflecting the average age :-)

pleasantly buzzed, hopefully not in an impactful way with regards to the morrow.

hyper-neutrino

well, i have no idea what timezone you're in, but i'll assume in one where it's an early hour, from what you said :P

it is 2:21 where i am

copper.hat

albany ca, its 23:22 here :-)

leslie townes

it's hardly early, copper. this isn't europe.

copper.hat

6:22 AM

i can pretend :-)

can start have convos with my daughter in london and my greek friend in edinburgh

the lynch-moussas 2016 that took 4 mos to arrive was well worth it :-)

the bigboy investment gulfstream flying over our house right now. so many things wrong with my knowing that.

of course, a large dose of irish guilt at the silly $66/bottle cost.

where is my ip addr?

Adil Mohammed

8:23 AM

Hii

No wait europe is currently sleeping, ill ask my doubt later lol

robjohn

Like Europe is the only continent that can answer your question?

6

napstablook

8:55 AM

@leslietownes I searched it up. Apparently people doubt that wolfram alpha snitches you to your school :p

Andrew Micallef

We all know there is no intelligent life elsewhere (jokes)

napstablook

Does wolfram alpha tell your school? is on faq section of google.

schn

11:12 AM

If $f\in C^m$ for some $m$ fixed, then $$f(t-hu)=\sum_{l=0}^m \frac{f^{(l)}(t)}{l!} (-hu)^l+o((hu)^m).$$

How is this formula modified if using $O$ instead of $o$?

Is it simply $$f(t-hu)=\sum_{l=0}^{m-1} \frac{f^{(l)}(t)}{l!} (-hu)^l+O((hu)^m) ?$$

The expansion is centered at $t$ and $hu$ is the variable.

Andrew Micallef

11:28 AM

Whats the little $o$ refer to?

Is $O$ the growth rate?

@schn

schn

@AndrewMicallef This may help.

Andrew Micallef

Actually wikipedia and mathworld were more informative

Sorry cant help, except it seams that from math world $o(\psi) \implies O(\psi)$ so doesnt that mean you would be right about just replacing the little o with the big one?

satan 29

11:44 AM

can anyone help me with a z transform question?

A-Level Student

11:54 AM

@satan29 It depends how hard it is for me :)

satan 29

0

Q: Determining the inverse Z transform

We are required to find $u_{2}$ and $u_{3}$ of the sequence $u_{n}$ whose z-transform is the function F(z): $$F(z)= \dfrac{2z^2+5z+4}{(z-1) ^{4}}$$ We are required to find Approach: I managed to split it into partial fractions: $$\dfrac{2}{(z-1)^2} + \dfrac{9}{(z-1)^3} + \dfrac{11}{(z-1)^4}$$ and...

z-transform

Adil Mohammed

@robjohn haha interestingly i remember seeing a meta-discussion somewhere (i think around 2012-13) in physics/maths about what time activity was highest and lowest, something along the lines "what is the best time to ask a question"

i may be wrong, cant be sure

Pherdindy

1:13 PM

Should we use the concept of dividing by n-1 for calculating the Mean Absolute Deviation of a Sample Data too? Why in the formula in MAD dividing the sum of differences of individual data points from the sample mean by n, not n-1?

I know typically n-1 is done because n underestimates the standard deviation isn't it also true for Mean Absolute Deviation?

robjohn

@Pherdindy does this help?

some correction is probably needed

probably need to multiply what you compute by $\sqrt{\frac{n}{n-1}}$

Perhaps$$\frac1{\sqrt{n(n-1)}}\sum_{k=1}^n|x_k-\bar{x}|$$

Pherdindy

Thanks, btw is there a plug-in to see latex formulas in chat here

I only see it as text format

Oh nevermind found it in the room info section

@robjohn thanks a lot. Will take a look at the stuff there including your answer. I saw a simulation before by khan academy explaining a graphical approach to the idea of n-1 in a sample standard deviation setting but never really looked into the math into it. Just wondering if it is the same to mean absolute deviation as well

robjohn

1:35 PM

@Pherdindy yes look here

The average deviation is less than the standard deviation, but of a similar order, and the standard deviation is multiplied by $\sqrt{\frac{n}{n-1}}$, so it would seem reasonable to think the average deviation would need to be increased as well.

@Pherdindy: were you able to install ChatJax?

Pherdindy

Yes I see the formulas already after bookmarking them

robjohn

There is a lot of latex flung around this room. It looks like gibberish if it is not rendered.

Pherdindy

Right I don't often go into math forums but realized it was probably a script/plugin issue but will try to apply the formula you suggested

I am looking into optimization techniques to apply to exponential smoothing for my forecast but i'm trying to understand why mean absolute deviation and/or mean squared error is being minimized

robjohn

I haven't done a proof of that for average deviation, only the standard deviation.

Pherdindy

I suppose standard deviation balloons the values too much compared to the average deviation

robjohn

1:47 PM

If you have the whole distribution, you don't use Bessel's correction.

That is only useful when the sample sizes are small

Pherdindy

Right it's just a sample dataset though only around 18 samples

Just trying to create a line that best predicts the dataset I used weighted moving averages only and I simulated the period lengths to see which minimizes the error

Was thinking of trying a different forecast method like exponential smoothing and compare the models

But to determine the best curve i'd need to use optimization techniques on exponential smoothing

And just realized there was no n-1 in average deviation so was curious

robjohn

If you read that post about the standard deviation, the same considerations are active for the average deviation; that is, the mean computed from the sample data will be closer to the sample data than will the actual distribution mean, so a correction will be needed.

The mean average distribution of the normal distribution is $\sqrt{\frac2\pi}=0.79788456$

Of course to minimize the average deviation, the median rather than the mean is better.

schn

2:08 PM

@schn I'd be grateful if anyone can confirm if my Taylor expansion with $O$ and $o$ respectively was correct. Click the arrow to get to the post.

Pherdindy

Thanks a lot very nice to know. There's a lot of things I still have to look for in building my model quite difficult for me to approach the problem for now given my lacking math foundations. I'll have to spend time reading some papers and other references including what you sent

robjohn

@Pherdindy for example if you measure the absolute deviation of $\{1,2,4\}$ from the mean, $\frac73$, you get $\frac{10}3$, where as measured from the median, $2$, you get $3$.

Pherdindy

I see that is interesting the references I look at minimize the deviation from the mean

It must take a lot of experience to see these easily like you do lol

robjohn

@Pherdindy It is easier to compute the mean from totaled data. The median requires holding all the data around.

Pherdindy

Thanks a lot i'll take these into consideration while I look at my data and try to make sense of how to approach it hah

robjohn

2:19 PM

the same goes for the absolute deviation as well. Thus, all one needs to keep for the standard deviation is the sum of the quantities and the sum of the squares of the quantities.

for the absolute deviation, from the mean or median, one needs to keep all the data around, as far as I know

EarthIsNotFlat

2:39 PM

in calculus we differentiate a function and equate it to zero to find minima and maxima. Can we do something similar for functional?

Rover

4:35 PM

Can't we say $Tan(3\pi/2)=- infinity $ ?

leslie townes

if that's what makes you happy

i don't think it's a convention because there's a kind of tie, right? you approach 3pi/2 with x's from one side, tan(x) gets very positive, you approach it with x's from the other side, it gets very negative

Rover

@leslietownes Ya, Tan(3pi/2)=+-infinity, so we can take any...value which we need ?

Like consider g(x)= 1+Tan|x| , then sgn(g(x))=+1 or -1 ?

leslie townes

4:54 PM

when i look deep into my own soul, i find that i don't need to define any value of tan(3pi/2). if you adopt a value in something written it would be helpful to say "i define tan(3pi/2) to be ___" so the person following along doesn't get lost

the nearest people often come to wanting to do that is with limits, but there, the definition of any limit at 3pi/2 (left, right, or two sided) doesn't require a value at 3pi/2

Rover

5:24 PM

Okay, thanks :-)

robjohn

@EarthIsNotFlat are you talking about things like the calculus of variations?

@Rover it depends on which side of $\frac{3\pi}2$ one approaches.

The circular functions are best viewed on a one point compactification of the line rather than a two point compactification.

that way $+\infty$ is right near $-\infty$ (actually, there is only $\infty$).

leslie townes

6:18 PM

a circle function unitarian.

i am a dualitarian. i wonder if there are any trinitarians in the room

robjohn

6:34 PM

@leslietownes I had a trinitarian monitor once.

Koro

I saw name of a vector calculus book either in Bromvich infinite series book or in Hardy's book on a course of pure mathematics. Unfortunately I forgot the name of the book and also where exactly I saw it. The book was published in 20th century or before. If you know about it, please tell me the name. Thanks.

It had a chapter which dealt with divergence and curl

I don't remember more :'(

I want to read divergence/ curl so I want that book but I forgot its name :'(

Ted Shifrin

@robjohn But limits/continuity are hopeless, so I dunno.

@Koro you mean physics-y? As opposed to mathy?

Koro

Hey Ted! mathy not physics-y

It had the irrotational vector , solenoid vectors also

schn

Suppose $f\in C^m$ for some $m\geq 1$, then $$f(t-x)=\sum_{l=0}^m \frac{f^{(l)}(t)}{l!} (-x)^l+o((x)^m).$$
How is this formula modified when using $O$ instead of $o$?
Is it simply $$f(t-x)=\sum_{l=0}^{m-1} \frac{f^{(l)}(t)}{l!} (-x)^l+O((x)^m) ?$$
The expansion is centered at $t$ and $x$ is the variable.

Koro

@schn: $o(x^n)$ is replaced by $O(x^{n+1}), n\ge 1$.

schn

6:51 PM

@Koro Right, because $O(x^{n+1})\implies o(x^n)$. However, despite that implication, are the two formulas I wrote for $f(t-x)$ still correct given $f\in C^m$?

Because in order to replace $o(x^m)$ by $O(x^{m+1})$ one would require $f\in C^{m+1}$, right?

Koro

I think that it should $o((-x)^m)$ instead of $o(x^m)$, from there you should clarify how you got $o(x^m)$. That is one. Another thing is: Why do you think one would require $f\in C^{m+1}$ to write $O(x^{m+1})$?

Ted Shifrin

7:07 PM

The point is that these are physical interpretations. The classic book is Schey, Div, Grad, Curl and All That. The right mathematics ties it all in with differential forms. This is in my book and videos,

Koro

I was gonna ask you about your book. Glad you told. Thanks. I'll take a look.

@schn I think it's a very good question. I have recently read this concept almost two days back. I know it as Peanno's form of Taylor series except that in that form all derivatives at centre are 0. Though I never thought about replacing small o by Big -O

leslie townes

"the book was published in 20th century or before" makes me feel old.

Koro

Hi Leslie, I was wondering what happened. Leslie didn't say anything and now you did :-)

leslie townes

i was doing some housework and preventing my daughter from getting her face clawed off.

Koro

The cat would dare to do that?

Ted Shifrin

7:21 PM

My kitten tried to bite my face when I discovered his latest favorite "hiding" place sitting in the bedroom window behind the blinds.

So a relatively low-rep user is now posting "As this question violates the Enforcement of Quality Standards, you should not answer this question, as it might lead to a bann [sic] of yours." I suppose this is better than my being hassled for commenting with hints for the OP, but I think things are getting a bit out of control. 1984 and the thought police.

copper.hat

I am curious as to the effectiveness of this whole crusade. Not that it matters, but it certainly has curtailed my involvement significantly.

It means that I spend more time in this chat room bugging you guys :-).

Ted Shifrin

7:37 PM

I lik the intent, but it's ironic that I was told numerous times that engaging in hinting conversations with OPs was not in compliance with the site principles. Make up your minds.

leslie townes

the impression i get is that hinting comments directing the asker of a PSQ toward an answer are discouraged, but comments directed toward the asker of the same question toward improving the question are not. seems like a weird place to draw the line, if it even is a line.

maybe the concern is that hinting comments may not lead to an answered question, just a question with a chain of comments containing an answer spread across multiple comments? beats me

Axel

Could somebody tell me what PSQ means please?

leslie townes

and a general concern that helping askers of PSQs without pushing them away from PSQs is like feeding seagulls at the beach

copper.hat

problem statement question, i believe

Axel

Thank you!

copper.hat

7:45 PM

i think MSE is (was) different in character from other SE/SO sites. partly because of the nature of mathematics, i imagine, but also i think because of less enforcement of the "avoid interaction" goal that seems to underlie SE/SO operations.

Yorch

It should be noted that psq's are a gateway drug

copper.hat

One encounters some really obnoxious folks on SO, along the lines of "I know it, you don't therefore you are stupid and deserve some denigrating remark".

Yorch

On payday you can frequently find all of the psq askers wasting their pay on bounties trying to get their fix

I've also seen people get dunked on for posting something from their blog which is relevant

copper.hat

Now if I could trade in my rep for a high quality MSE t-shirt...

Yorch

and people calling it self advertising and stuff

I printed my own mse flair

Koro

8:00 PM

Yess!! I remembered. This is the book: Willard Gibbs' Vector Analysis

@Ted @Leslie

Thorgott

same thing just happened here math.stackexchange.com/questions/4178339/…

whats going on

robjohn

@TedShifrin $f(x)=e^{2i\tan^{-1}(x)}$ maps the one point compactification of the reals to the unit circle, and mapping $f(\tan(x))=e^{2ix}$ is nicely continuous. $f(1/x)=-1/f(x)$ is also a nice consequence.

Yorch

that user got suspended for EOQS

Ted Shifrin

@Koro very old-fashioned

Cornman has become a vigilante. I told him so.

leslie townes

you can't go too far back with vector analysis or you get to books where they waste their time implicitly fighting quaternions vs. vectors and taking weird perspectives that have been improved upon.

Yorch

8:07 PM

lol, that answer has 5 upvotes now

he's not a vigilante

Ted Shifrin

@robjohn OK, so sometimes. I was thinking in the context of the original $\tan(3\pi/2)$.

Yorch

he is just lashing out after being wrongfully suspended

I'm sure he'd stop it after a while

it is clear he doesn't believe in meddling with people if you read his meta post

Astyx

Hello chat

Ted Shifrin

Cornman was suspended, so he’s now doing the same thing?

Yorch

he was suspended for something else, EOSQ

Ted Shifrin

8:11 PM

I don’t spend my life on meta.

leslie townes

ted, you are missing out on the full richness of human experience.

Ted Shifrin

What is EOSQ?

Hi @Astyx.

Thorgott

enforcement of quality standards

Koro

@TedShifrin +1

@Thorgott I see

Thorgott

he's had an argument with quid under a meta post earlier, I imagine this is him getting back in some way

Ted Shifrin

8:13 PM

Asshole behavior.

Astyx

How do people get so involved

Ted Shifrin

Now watch me get banned.

Koro

No Ted, please.

Yorch

what's the problem

he clearly can't get back at quid

as was clearly demonstrated

Ted Shifrin

This is like Trompies refusing to mask. Makes me want to quit MSE entirely.

Thorgott

8:14 PM

I will quote this again:
The Committee which was set up in Rome for the unification of vector notation
did not have the slightest success, which was only to have been expected.
-F. Klein, Elementary Mathematics
from an Advanced Standpoint, 1925

Axel

I am quite a newbie to this site, and I apologize if my question is not very relevant. This is a great community and people are putting a lof of efforts for EoQS to be respected, but mostly among the new contributors there are numerous people that don't meet the expected quality standards and they get immediately downvoted, but some of them seem not to be really aware that they post a low-quality question. So I wondered, is there a way to prevent this from happening beforehand?

Ted Shifrin

No. COVID has brought cheating to new levels.

leslie townes

the people who would read or click through any documents before posting a first question (and internalize the lessons) are probably people who would be asking good questions anyway.

2

i don't know how else to do it

Axel

@leslietownes I do agree with you

Koro

What about this: If X uploads a question, answer (posted as answer to X's question) should be visible to X only after say 15 mins. Meanwhile, rest all users can see the answer but the question asker can't.

Astyx

8:18 PM

Allow only people with more than 10k rep to post anything

Yorch

what percentage of questions are good?

like 0.1%?

Koro

This may avoid cheating to an extent

Yorch

it depends on what good means

I think I'm being lax

copper.hat

Well, smart cheaters can just add some bs 'context' to pass the squad test.

Koro

Practically, who would spend 15 mins doing nothing in the exam

?

Yorch

8:19 PM

smart cheaters don't use this site to cheat

it's too unreliable

Ted Shifrin

The whole thing strikes me as akin to the misbelief that the death penalty functions as a deterrent. That was handily disproved , and yet …

leslie townes

"i tried (my textbook's hint rewritten as an attempt)" is a classic edit.

Yorch

just get in a discord group

robjohn

@Astyx too much time for navel contemplation

leslie townes

one thing that struck me about cheaters, when i taught math - at least the most egregious ones - is that they are lower than average in awareness of the technological tools available to them.

copper.hat

8:20 PM

apart from ethical issues, the death penalty does not pass the basic test of unquestioned guilt in many cases.

leslie townes

discord would be a great idea. let's set up a cheaters' den discord server and refer cheaters to it

Ted Shifrin

That's a separate point from mine, copper.

copper.hat

i agree

Thorgott

@Astyx great suggestion, I approve

copper.hat

i can't say i fully understand the general goal of the pruning.

Thorgott

8:22 PM

cheaters tend to be not the brightest

love it when they delete the entire body of their question thinking that covers their tracks

Ted Shifrin

@Astyx haha

Yorch

The one thing that I can't understand is: "If pruning is important, why are the tools for pruning the site so incredibly bad"

It leads me to believe the corporate overlords don't care about pruning

Axel

I had another question too, why sometimes do I lose 10 or 20 points of reputation without being noticed? It happens that I have a notification saying "user has been removed" but other times I have no clue.

Yorch

go here and check "show deleted posts" math.stackexchange.com/users/745510/axel?tab=reputation

Axel

thank you!

robjohn

8:26 PM

@Astyx How do we get to 10K then?

Yorch

did it work?

Astyx

Precisely :)

Axel

@Yorch I don't know if you're talking to me, but if it is the case, yes, I have the details, it's great, thanks :-)

Yorch

dope xD

copper.hat

i view it as a sort of evolving social experiment.

Yorch

8:32 PM

I don't know

Axel

Did you have as much "low-quality" questions before COVID-19?

Sorry, I deleted my question by mistake

Yorch

I've never had a problem with the quality of the site

I can skim through everything posted in a day in 20 minutes so I don't care

robjohn

@Yorch You can? there is so much posted every day

copper.hat

I think part of the issue is searching for matching questions is a hit or miss affair.

Yorch

it's like 350 I think

last I checked

so 3.5 seconds per

yeah seems about right

robjohn

8:34 PM

@copper.hat definitely. That makes finding duplicates hard, but there are still PSQs that are pretty common.

Yorch

you can also block tags

oh, it's 424

copper.hat

@robjohn Maybe some automated dup detector for such stuff. I don't have any idea how prevalent they are or what the real site impact is in having the dups.

robjohn

There are about 900 posts per day (questions and answers)

Axel

I wondered also, what about the older questions that did not really come to light? Are they forgotten or does someone help to clean old posts?

Yorch

I was just saying questions

I don't think they clean old posts

unless you edit an old post to make it better

robjohn

8:37 PM

500 questions per day approx

Yorch

then it runs the risk of getting deleted

Axel

ok thank you

Yorch

stackexchange.com/sites#questionsperday

robjohn

Interesting the source I saw says the last 7 day average was slightly above 500

copper.hat

wow, SO is the largest by a large margin

robjohn

8:39 PM

close enough and variation is pretty wild

copper.hat

i guess that is the income generator

Yorch

you can also change the sorting parameter, it's pretty interesting :)

copper.hat

probably too complicated, but questions themselves should probably go through some 'ageing' process before rep can be gained by answering them.

robjohn

@copper.hat duplicates don't get deleted if they have answers, except if the questions can be merged together

copper.hat

makes sense.

Yorch

8:41 PM

you mean automatically?

I've seen loads of seemingly fine dupes get deleted

robjohn

@Yorch no, a mod needs to do it

@Yorch were they answered?

copper.hat

i certainly have found dups useful at times, a slightly different approach in an answer can be enlightening.

Yorch

math.meta.stackexchange.com/a/33612/33907

Yeah I tried to get one I had an answer on undeleted once

robjohn

If the question and answers have 0 or negative vote count, I think they may get autodeleted

Yorch

And failed hilariously

oh no, it got 10k power deleted

copper.hat

8:43 PM

reminds my why i avoid politics, school boards, city councils, etc

Yorch

yeah, I'm just waiting to start my masters, so I have too much free time

I've been getting lampooned by bureaucrats

robjohn

@Yorch that was deleted by other users. The answers might have been preserved if they had flagged for a moderator instead.

Yorch

Mike's answer is pretty dope

Mine isn't really worth saving to be completely honest

copper.hat

i think this context thing is bs.

robjohn

@Yorch yeah. It was closed by Mike after answering. Some people felt that was unfair.

Yorch

8:48 PM

oh yeah I had forgot about that

I also one time did that

closing a question after someone else voted to close

and gold badge-hammering it

robjohn

@copper.hat People want to prevent PSQs, where people simply copy the questions from their homework, and it is hard to codify how to stop that. The context thing is one attempt.

Yorch

at the end of the day I think I shouldn't worry too much

Axel

Thank you for your answers by the way, I am wishing everyone a great evening (or a good afternoon depending on where you live)

Yorch

maybe if at some point I get banned for like half a year

I need to really change how I use the site

I've been using comments a lot more now

also, I feel like I'm only really at risk during periods of boredome

robjohn

@Yorch I had to. I tried to help too many people and didn't worry about the quality of their questions.

copper.hat

8:52 PM

@robjohn i understand the intent, but what passes for context is weak, and there are lots of good questions that stand on their own.

Yorch

At least I'm thankful the whole context thing doesn't seem to be beiing applied retroactively to really old stuff (for now)

robjohn

Then you run into the people who nix questions and their answers based on their form, not necessarily their content. However, they don't know enough about the content to judge by the content, so they need to use the form instead.

@copper.hat I agree that some questions are good just as they stand, but it requires a keen eye to tell those that are from those that aren't and so the standards have to be pro forma.

copper.hat

i appreciate that.

robjohn

not all eyes were created equal

Yorch

I wish I could get a leaderboard on delete votes cast all-time, I'm pretty sure it would be hilarious

"the results will shock you"

robjohn

8:59 PM

@Yorch try this

You might be able to alter that to give delete votes.

Yorch

hmm, thanks !

deletevotes doesn't seem to be a parameter :/

although I was able to run a super heavy downvote qeque

And I was not surprised

well, actually I was

some really low rep people showed up :/

copper.hat

downvote $\neq$ - upvote

Yorch

9:14 PM

there's a user with 9 upvotes, over 3.5k downvotes and 800 rep. wut?

robjohn

@Yorch no, there are no parameters to that query. You would need to create a new query which counts delete votes instead of downvotes

Yorch

I was scrolling through the parameter options but didn't find the correct one

Oh but that user has been inactive since 2015 :'(

he honestly seemed to post good stuff

robjohn

@Yorch It does seem that one can check the up/downvote totals (and you can see those on the user activity page), but not the delete vote totals (because that is not publicly available data).

Yorch

thanks

leslie townes

10:06 PM

i shaved what had become a beard, and when my daughter saw me she said "no! what happened to your fur?" and "get your fur back"

robjohn

10:31 PM

@leslietownes how long had you been growing the fur?

leslie townes

maybe two months?

robjohn

So it wasn't the whole duration of covid.

leslie townes

no, i can never commit. i had a three month one, and then two two-month ones. math would suggest that now i have a 7-month beard, but for some reason i don't.

robjohn

there are beard toupees that you can get

leslie townes

i may have to.

copper.hat

10:41 PM

you need to maintain some element of control...

leslie townes

well it's a tie. my wife hates my beard, but my daughter loves it. i like it too, for about 2-3 months.

beard modulo 2 months might be the plan

PM 2Ring

11:59 PM

@Axel For some context on the PSQ policy, please see math.meta.stackexchange.com/q/33508/207316

Transcript for

$Mathematics$ Mathematics