User talk:David Eppstein

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Hi, and welcome to my User Talk page! For new discussions, I prefer you add your comments at the very bottom and use a section heading (e.g., by using the "New section" tab at the top of this page). I will respond on this page unless specifically requested otherwise. For discussions concerning specific Wikipedia articles, please include a link to the article, and also a link to any specific edits you wish to discuss. (You can find links for edits by using the "compare selected revisions" button on the history tab for any article.)

figuring out how to clean up a large semi-automated mangled citation mess[edit]

Thanks for fixing my revert mistake on Complex number‎. I made several other similar reverts somewhat quickly though I was trying to check them; hopefully I picked the right reversion point on all the others or if not hopefully someone watching those pages will double-check.

The overall problem here has a scope that is well beyond my capacity to handle on my own. I am guessing there are at least several hundred edits that need to be reverted or carefully manually fixed. I only checked/tried to revert the most recent handful. I opened a discussion at Wikipedia:Administrators' noticeboard#help cleaning up mangled citations but haven't gotten much feedback there yet. Ideas? –jacobolus (t) 07:42, 1 April 2023 (UTC)Reply[reply]

I've been trying to address the problem, but the amount of checking required is substantial. XOR'easter (talk) 16:27, 9 April 2023 (UTC)Reply[reply]

Temporally disrevert then revert the 'Power of 2' article[edit]

Stop icon

Your recent editing history at power of two shows that you are currently engaged in an edit war; that means that you are repeatedly changing content back to how you think it should be, when you have seen that other editors disagree. To resolve the content dispute, please do not revert or change the edits of others when you are reverted. Instead of reverting, please use the talk page to work toward making a version that represents consensus among editors. The best practice at this stage is to discuss, not edit-war; read about how this is done. If discussions reach an impasse, you can then post a request for help at a relevant noticeboard or seek dispute resolution. In some cases, you may wish to request temporary page protection.

Can you disrevert the 'power of 2' article, copy-paste the changes I made and paste it to your user page, then revert it? See Power of two. Faster328 (talk) 11:20, 1 April 2023 (UTC)Reply[reply]

The reason I added 2^160 and 2^2^2^n section is to provide more knowledge. 2^160 is approximately 1.461501637x10^48. You can make the new section better and improve it. Faster328 (talk) 11:23, 1 April 2023 (UTC)Reply[reply]
No, I'm not going to stuff up my user page with useless powers of two. If you want to see the text you added, look at the article history. —David Eppstein (talk) 16:11, 1 April 2023 (UTC)Reply[reply]
Can you improve my secondary page - Draft:2^n, 2^2^n, 2^2^2^n? Add more ciations for the secondary article. Faster328 (talk) 00:57, 2 April 2023 (UTC)Reply[reply]
More? There are zero citations. See WP:NOR. You should be working from sources, not piling up calculations in the hope that maybe someone else can source them for you. —David Eppstein (talk) 00:58, 2 April 2023 (UTC)Reply[reply]
I have proof that my calculations are true. I found it using the logarithm method. Below I will show you the logarithm method. The more precise the value is, the more accuracy. I can find a maximum of 4157 digits.
My calculations have been proven true because using a small change in value to log10(2), there is also a small change in the digits of the log10. When I remove a few digits, like 4 digits removed, the logarithm of the value change by the same amount of digits. I find the decimal point, copied the digits after it, then add a '0.' to the first digits of the number. I then used the 10^() function to convert the logarithm to the value. After that, I copied the digits before the decimal point and paste it between the brackets in the 10^(). The final result is the value times 10^(number of digits). I also added a fact for you guys that every tetration of 2, a new fixed digit appears. The fixed digits then converge to a 10-adic number(...5075353432948736). Faster328 (talk) 01:49, 2 April 2023 (UTC)Reply[reply]
This vague description does not fill me with confidence. —David Eppstein (talk) 01:54, 2 April 2023 (UTC)Reply[reply]
Can you make a simpler description? Faster328 (talk) 01:57, 2 April 2023 (UTC)Reply[reply]
Can you answer the question of what you think the point of adding this sort of calculation to Wikipedia would be? What information is a reader of the article likely to be looking for that would cause them to see this and think "that's what I was looking for"? How is it not just a useless pile of digits making the articles longer and making the actual information in the articles harder to find? —David Eppstein (talk) 02:05, 2 April 2023 (UTC)Reply[reply]
You can disrevert the page, then copy to the secondary article(Draft:2^n, 2^2^n, 2^2^2^n), then I can thank you. Faster328 (talk) 02:12, 2 April 2023 (UTC)Reply[reply]
So you still haven't read anything anyone has responded to you here or elsewhere, including my statement that I will not copy this material for you and my suggestion that if you want to find it again you look in the article history? —David Eppstein (talk) 02:13, 2 April 2023 (UTC)Reply[reply]
I agree with David on this one, I don't see the point of of listing 2^2^2^n all the way to n=11, for instance. What does this accomplish? I would suggest that this kind of research belongs in a Math journal, but are not interesting enough to the general public on Wikipedia. Try publishing your own paper. Dhrm77 (talk) 17:05, 13 April 2023 (UTC)Reply[reply]

Your GA nomination of Beckman–Quarles theorem[edit]

Hi there, I'm pleased to inform you that I've begun reviewing the article Beckman–Quarles theorem you nominated for GA-status according to the criteria. Time2wait.svg This process may take up to 7 days. Feel free to contact me with any questions or comments you might have during this period. Message delivered by ChristieBot, on behalf of The person who loves reading -- The person who loves reading (talk) 16:21, 1 April 2023 (UTC)Reply[reply]

Your GA nomination of Beckman–Quarles theorem[edit]

The article Beckman–Quarles theorem you nominated as a good article has passed Symbol support vote.svg; see Talk:Beckman–Quarles theorem for comments about the article, and Talk:Beckman–Quarles theorem/GA1 for the nomination. Well done! If the article has not already appeared on the main page as a "Did you know" item, or as a bold link under "In the News" or in the "On This Day" prose section, you can nominate it within the next seven days to appear in DYK. Bolded names with dates listed at the bottom of the "On This Day" column do not affect DYK eligibility. Message delivered by ChristieBot, on behalf of The person who loves reading -- The person who loves reading (talk) 16:44, 1 April 2023 (UTC)Reply[reply]

Powers of 5 article[edit]

I hope you can create a draft article named 'Powers of 5' and add values and sources from reliable ciations like the OEIS. You can use the ciation bot if you want. Faster328 (talk) 02:24, 2 April 2023 (UTC)Reply[reply]

Get used to disappointment. —David Eppstein (talk) 06:42, 2 April 2023 (UTC)Reply[reply]
Possible. Faster328 (talk) 06:50, 2 April 2023 (UTC)Reply[reply]


Hi. Have you ever thought of doing work over at NPP? You could focus only on scholars and other academic subjects. You are definitely the best one at assessing their notability. Onel5969 TT me 12:18, 2 April 2023 (UTC)Reply[reply]

I do regularly look through the new pages listed at User:AlexNewArtBot/MathSearchResult and User:AlexNewArtBot/WomenScientistsSearchResult. That tends to focus more on articles I'm likely to have an interest in. —David Eppstein (talk) 15:19, 2 April 2023 (UTC)Reply[reply]
Thanks for that. I always value your opinion on all articles, but most especially on those types. I often say to myself when reviewing a borderline NSCHOLAR article, "self, what would David Eppstein think?". Onel5969 TT me 01:17, 3 April 2023 (UTC)Reply[reply]

I liked ...[edit]

... your response here, "a more convincing argument" indeed ;-) Paul August 00:35, 3 April 2023 (UTC)Reply[reply]

Thanks! Notch another one up for Socrates. —David Eppstein (talk) 05:12, 3 April 2023 (UTC)Reply[reply]

Each pile of calculations needs an article.[edit]

I need you to create an article called 'Power of 2 Values' and add links from and to Power of two and 'Power of 2 Values'. Faster328 (talk) 06:11, 4 April 2023 (UTC)Reply[reply]

du Chatelet and journal publication[edit]

Hi David, I noticed your edit on Eunice Foote. I thought it was a good edit, and we should indeed not forget about Émilie du Châtelet. In particular I was ignorant of her publication "Dissertation sur la nature et la propagation du feu," and am glad to know more about her work - your edit inspired me to do so, and so I just wanted to say thanks for that. Qflib, aka KeeYou Flib (talk) 12:58, 7 April 2023 (UTC)Reply[reply]

Hello, David Eppstein,

You nominated this article for deletion last year and I closed the AFD as "Delete". I was just notified that a discussion has been started at Deletion review and an editor is asking for the article history to be restored. I'm not sure if you have an opinion about this but I think you probably have a stronger view than I did since you nominated the article for deletion and I simply assessed the comments that had been made by the participants. I hope you had a pleasant weekend. Liz Read! Talk! 00:26, 10 April 2023 (UTC)Reply[reply]

Thanks for the heads up. I don't think I have strong feelings about this either way. The history is a bit of a mess regardless. —David Eppstein (talk) 07:30, 10 April 2023 (UTC)Reply[reply]

Revert on Big O notation?[edit]

Can you explain the revert on Big O notation? Your comment makes it sound like it should be obvious but I don't see why, based on the talk page and WP:NOTTEXTBOOK. Assuming I'm missing something. - Procyonidae (talk) 06:48, 11 April 2023 (UTC)Reply[reply]

NOTTEXTBOOK means that we should not include certain types of unencyclopedic content, such as user manuals for pieces of software, game play-throughs, or recipes for cooking meals. It does not mean we should avoid saying anything at all that a textbook might also way. It does not require us to omit claims that might be useful to readers, merely on the basis that they are useful and therefore might be found in a textbook. In particular, it does not require us to omit the observation that 0 is never used for O-notation even though it looks a lot like O. It also does not require us to delete content merely because it is phrased in imperative. That phrasing is common throughout mathematical writing, especially as a way of avoiding the use of passive voice (as I used above instead) or second-person plural. Its use is merely a matter of grammatical choice, not an indication that the sentence is forbidden by NOTTEXTBOOK. Even descriptions of step-by-step processes for doing things are not always forbidden by NOTTEXTBOOK. For instance, many algorithms are notable, should be covered by Wikipedia, and cannot be covered without describing the steps of the algorithms. More often than not, when I see NOTTEXTBOOK used, it is used incorrectly. My edit summary was intended to indicate that your use of it was not merely incorrect, but more incorrect than usual. —David Eppstein (talk) 07:10, 11 April 2023 (UTC)Reply[reply]
Followup: ok so even if NOTTEXTBOOK is the incorrect reason, I still think it makes sense to remove the statement "The digit zero should not be used" from the article, no? - Procyonidae (talk) 07:38, 11 April 2023 (UTC)Reply[reply]
Why? —David Eppstein (talk) 15:47, 11 April 2023 (UTC)Reply[reply]
As written, it's just an uncited personal suggestion. Even if we add refs supporting the advice to not use a zero, I think it would need a qualifier, to the effect of "within [field/subfield], other symbols like a zero or theta are not considered equivalent". And then on top of that, I would think there still needs to be more to justify its inclusion in the article. For example, the page for π wouldn't include the sentence "The character n should not be used", unless there was some recorded justification, e.g. an observed trend of people using n for π, despite the two being similar in form. Does that reasoning make sense? What do you think? - Procyonidae (talk) 20:54, 11 April 2023 (UTC)Reply[reply]
It is very easy to source:
And since this is an article about notation, material that is precisely about notation is very much on-topic. —David Eppstein (talk) 21:02, 11 April 2023 (UTC)Reply[reply]
@Procyonidae I think your intuition here is "Wikipedia is descriptive not prescriptive". Idea: you could change the wording to "The digit zero is never used"—that was its wording before this edit. 'wɪnd (talk) 02:17, 16 April 2023 (UTC)Reply[reply]
@'wɪnd that wording looks better to me, I agree. —David Eppstein (talk) 05:13, 16 April 2023 (UTC)Reply[reply]

Sourced positive descriptions and vacuous puffery[edit]

I'd like some clarity on your undo here: Editors should refrain from this sort of effort until they learn to tell the difference between sourced positive descriptions of something and vacuous puffery.

And for context, the two relevant phrases are (emphasis mine):

  • Euler is held to be one of the greatest mathematicians in history and the greatest of the 18th century.
  • Euler is also widely considered to be the most prolific [mathematician]

Two things:

  1. I checked the given sources before editing and also where the edit was introduced. It's not clear to me where the source is.
  2. As I see it, greatest is a peacock term here in my reading of MOS:PUFFERY. It seems to me, our disagreement arises from the word pointing to different things for different people when applied to humans. For some of us, "greatness" is a factual term and can be measured by certain (implied) criteria: e.g. number of times cited, number of ideas used today that can be traced back to their writings, etc. For others, "greatness of humans" is not factual or sharp, but a subjective opinion/ranking based on unclear criteria. I believe, that's why MOS:PUFFERY points out Instead of making subjective proclamations about a subject's importance, use facts and attribution to demonstrate it.

The facts implying importance here are the sourced opinions of Laplace and Gauss and the "number of publications". Moreover, and arguably most importantly, the various ideas introduced by Euler which are listed. 'wɪnd (talk) 21:15, 15 April 2023 (UTC)Reply[reply]

@David Eppstein Or... did I misunderstand the MOS:PUFFERY and WP:PUFFERY page? It seems to me, it doesn't make the distinction between vacuous and unvacuous use of peacock terms, rather it says to avoid them all together... with the exception of attributions such as 'Laplace said: "Lisez Euler, lisez Euler, c'est notre maître à tous."'. What do you think? 'wɪnd (talk) 17:31, 17 April 2023 (UTC)Reply[reply]


In all seriousness, if you (or any lurker) have any suggestions about how I could have edited better before or during or conducted myself better in relation to this filing, I would be very appreciative of that - here, at my own Talk, or anywhere really. Newimpartial (talk) 01:52, 16 April 2023 (UTC)Reply[reply]

For a start, you could maybe have paid attention to the complaints that, by responding on the talk pages of everyone who left a negative opinion, you were bludgeoning the discussion, instead of continuing to do exactly the same thing, as you have done here, after that behavior had already started to become a significant topic of discussion? —David Eppstein (talk) 02:01, 16 April 2023 (UTC)Reply[reply]
Well, you seem to be saying that asking seasoned folks like you and Ealdgyth how I can improve is a form of bludgeon? Is that because I didn’t wait long enough after the !votes were cast, or something? My understanding is that BLUDGEON is about trying to convince people of something, usually to influence !votes, and my comments (at Talk pages of five of the ten non-support !voters) weren't aiming to do that at all - two of the five were simple questions about how to improve, and the other three were clarifications I wanted to make, not intended to convince anyone of anything.
So maybe BLUDGEON means something that I don't get from reading WP:BLUDGEON? And I do get (partly thanks to a helpful comment by Beccaynr) that I need to take more reflective time before deciding what is worth saying in response to others. But what else am I missing? You don't owe me an answer, of course, but I am asking for insight, I'm not trying to influence anything (especially not the current AN outcome, which seems foregone). I just want to grow and learn, here, as I have (sooner or later) from prior experiences on-wiki. Newimpartial (talk) 02:17, 16 April 2023 (UTC)Reply[reply]

Asking again[edit]

I've waited 24 hours after the closure of my ANI filing, and since you have not (yet?) banned me from your Talk I want to ask again how I can improve my editing, in relation primarily to your comment here. I have taken on your reccomendation here (and I have expressed my realisation about that and some of my plans going forward), so I should be able to a avoid recidivism on that specific matter.

To clear the air, I have to disclose that I felt rather hurt by your comments, especially this one. I consider that almost everything I understand about NPROF, and a large part of my understanding of WP:N, came from reading your contributions, and in a previous brief interaction at ANI, [1] I felt that you were standing against bullying I was experiencing, whereas this time you stood on the other side (I have explained some of my experience - with no aspersions about any other editor - here). It is difficult to be dismissed by an editor I have seen as a kind of exemplar.

Anyway, with that air cleared, I just want to know what (apart from staying away from the Talk pages of discussion participants while a discussion is going on) you think I ought to be changing in my editing? I am trying to keep the question as open-ended as possible, so that if you choose to answer the reply will be affected as little as possible by my own blinders and preconceptions. Newimpartial (talk) 18:09, 18 April 2023 (UTC)Reply[reply]

Your GA nomination of Rook's graph[edit]

Hi there, I'm pleased to inform you that I've begun reviewing the article Rook's graph you nominated for GA-status according to the criteria. Time2wait.svg This process may take up to 7 days. Feel free to contact me with any questions or comments you might have during this period. Message delivered by ChristieBot, on behalf of TheTrueSauce -- TheTrueSauce (talk) 13:42, 21 April 2023 (UTC)Reply[reply]

Apologies for the confusion on Rotating Calipers[edit]

Hey! Sorry about my incorrect """fix""" on Rotating calipers. I glanced at the citation and saw that it needed a journal and didn't think before I edited. Apologies and thanks for correcting me. Moon motif (talk) 22:30, 22 April 2023 (UTC)Reply[reply]

No problem. Sadly, I have been unable to find a usable link to the actual paper; you can find copies online but they're typically fair-use links from someone's course readings, probably fair use for the person putting it there but a violation of WP:ELNEVER for us. —David Eppstein (talk) 00:42, 23 April 2023 (UTC)Reply[reply]

Not sure my ping in the edit summary worked. Could you please take a look at this one? Thanks. Onel5969 TT me 11:11, 25 April 2023 (UTC)Reply[reply]

Talk page stalker here. Without looking deeply, this one looks a bit WP:TOOSOON for NPROF. We've got 2 moderately cited papers, in what I don't think is an extremely low citation field. I think that's short of WP:NPROF C1. She does not appear to be editor-in-chief of the Journal of Digital and Social Media Marketing (none such is listed, but Simon Beckett is the one writing the editorials for the last couple of issues of the journal), and I am anyway uncertain as to whether this would be a well-established journal for WP:NPROF C8. A Canadian Research Chair I understand to be more along the lines of a grant; anyway, WP:NPROF C5 explicitly does not give a pass for associate professors. No books are apparent for NAUTHOR. I didn't look too hard for GNG. Russ Woodroofe (talk) 11:57, 25 April 2023 (UTC)Reply[reply]
I don't really see anything to disagree with here. I might try to use #C5 for a tier 1 Canada Research Chair, but this one is tier 2. —David Eppstein (talk) 15:28, 25 April 2023 (UTC)Reply[reply]
Missed this. Thanks to you both.Onel5969 TT me 10:28, 2 May 2023 (UTC)Reply[reply]


I'm not a vandal, and not that vandal, whoever you were refering to. Sorry if my edit was considered non-constructive by you. Best regards, 2804:14D:5C32:4673:93CE:2F22:61C6:ABCC (talk) 05:39, 26 April 2023 (UTC)Reply[reply]

Herschel graph[edit]

The section "Polyhedron" in your pre-nomination GA Herschel graph may be added the name "Herschel enneahedron"? Because some of the sites and platforms (such as Youtube) used the term "Herschel enneahedron", but sadly it seems that in the Google Scholars there are only two sites that mentioned about that name: [2] and [3].

The second one is from MathWorld, but I don't think you wouldn't agree to use this source, so I'm waiting for your comment. Regards, Dedhert.Jr (talk) 12:31, 27 April 2023 (UTC)Reply[reply]

Why not wait until that name is used in scholarly work, not just YouTube? —David Eppstein (talk) 16:10, 27 April 2023 (UTC)Reply[reply]
Okay :D Dedhert.Jr (talk) 05:26, 28 April 2023 (UTC)Reply[reply]

Women in Red May 2023[edit]

WiR education 2023.png
Women in Red May 2023, Vol 9, Iss 5, Nos 251, 252, 267, 268, 269, 270

Online events:

See also:

Tip of the month:

  • Use the Google translate app and camera on your phone to translate text from an article or book

Other ways to participate:

Facebook icon.jpg Facebook | Instagram.svg Instagram | Pinterest Shiny Icon.svg Pinterest | Twitter icon.png Twitter

--Lajmmoore (talk) 18:27, 27 April 2023 (UTC) via MassMessagingReply[reply]

Your GA nomination of Midsphere[edit]

Hi there, I'm pleased to inform you that I've begun reviewing the article Midsphere you nominated for GA-status according to the criteria. Time2wait.svg This process may take up to 7 days. Feel free to contact me with any questions or comments you might have during this period. Message delivered by ChristieBot, on behalf of Pi.1415926535 -- Pi.1415926535 (talk) 05:22, 1 May 2023 (UTC)Reply[reply]

Your GA nomination of Rook's graph[edit]

Hi there, I'm pleased to inform you that I've begun reviewing the article Rook's graph you nominated for GA-status according to the criteria. Time2wait.svg This process may take up to 7 days. Feel free to contact me with any questions or comments you might have during this period. Message delivered by ChristieBot, on behalf of Hawkeye7 -- Hawkeye7 (talk) 23:01, 2 May 2023 (UTC)Reply[reply]

Your GA nomination of Rook's graph[edit]

The article Rook's graph you nominated as a good article has been placed on hold Symbol wait.svg. The article is close to meeting the good article criteria, but there are some minor changes or clarifications needing to be addressed. If these are fixed within 7 days, the article will pass; otherwise it may fail. See Talk:Rook's graph and Talk:Rook's graph/GA1 for issues which need to be addressed. Message delivered by ChristieBot, on behalf of Hawkeye7 -- Hawkeye7 (talk) 00:01, 3 May 2023 (UTC)Reply[reply]

Your GA nomination of Rook's graph[edit]

The article Rook's graph you nominated as a good article has passed Symbol support vote.svg; see Talk:Rook's graph for comments about the article, and Talk:Rook's graph/GA1 for the nomination. Well done! If the article has not already appeared on the main page as a "Did you know" item, or as a bold link under "In the News" or in the "On This Day" prose section, you can nominate it within the next seven days to appear in DYK. Bolded names with dates listed at the bottom of the "On This Day" column do not affect DYK eligibility. Message delivered by ChristieBot, on behalf of Hawkeye7 -- Hawkeye7 (talk) 08:43, 4 May 2023 (UTC)Reply[reply]

Your GA nomination of Midsphere[edit]

The article Midsphere you nominated as a good article has passed Symbol support vote.svg; see Talk:Midsphere for comments about the article, and Talk:Midsphere/GA1 for the nomination. Well done! If the article has not already appeared on the main page as a "Did you know" item, or as a bold link under "In the News" or in the "On This Day" prose section, you can nominate it within the next seven days to appear in DYK. Bolded names with dates listed at the bottom of the "On This Day" column do not affect DYK eligibility. Message delivered by ChristieBot, on behalf of Pi.1415926535 -- Pi.1415926535 (talk) 21:42, 4 May 2023 (UTC)Reply[reply]

Objection on proposed deletion Permutational_Number_System[edit]

Good day Professor David Eppstein. This is regarding your proposal for deletion of article Permutational Number System. You are absolutely right that the referred eJournal is not peer reviewed. This is because of the fact that at the time of writing the paper, I decided not to spend a lot of money on publishing in peer reviewed journal so I selected free eJournal. But I would be glad and honored if you can peer review it, and guide me for publishing in a peer-reviewed journal.

Here is a very quick summary for you:

[1] Factorial number system along with Lehmer-code can be used to rapidly compute nth unique permutation.

[2] Similar to point 1, Permutational Number System which is an extension/generalisation of Factorial Number System can be used along with Deep-Code(generalisation of Lehmer-code) to rapidly compute nth k-permutation.


Deepesh Pateldeepesh (talk) 23:42, 8 May 2023 (UTC)Reply[reply]

Turing test article[edit]

Combinational logicFinite-state machinePushdown automatonTuring machineAutomata theoryAutomata theory.svg
About this image
Classes of automata
(Clicking on each layer gets an article on that subject)

I've been thinking of expanding the Wikipedia article for Turing test to include a new section dealing with the sub-question of when a Turing machine is able to discern if the output of a responding machine is limited to being a pushdown automaton, a finite state machine, or a conventional machine limited to logical connectives as shown in the diagram. I'm currently aware of an exponential algorithm which solves this problem for whether a responding machine is limited to using only logical connectives, but do not have data for the other higher classes of automata which I just listed. Have you seen any algorithms published for these types of questions? ErnestKrause (talk) 20:55, 10 May 2023 (UTC)Reply[reply]

There's some relevant material in the computational learning theory / machine learning literature on automaton inference. For example, see: Rivest, Ronald L.; Schapire, Robert E. (1993), "Inference of finite automata using homing sequences", Information and Computation, 103 (2): 299–347, doi:10.1006/inco.1993.1021, MR 1216458David Eppstein (talk) 20:59, 10 May 2023 (UTC)Reply[reply]
Really interesting article by Rivest which I was able to download; they appear to be claiming an algorithm of cubic complexity for determining if the input/output behavior is from a finite state machine. Do you know if there are any articles about the other automata classes (for example, push-down automata) as well? It seems for the Rivest article that their claim to cubic complexity for FSM is substantially better that the exponential algorithm I've located for solving this problem when limited to automata using only logical connectives. ErnestKrause (talk) 21:19, 10 May 2023 (UTC)Reply[reply]
I don't know that literature very well but I think the keywords to use in searching for analogous results for push-down automata (or maybe rather context-free grammars) are grammar inference. We have a Wikipedia article on a closely related topic, grammar-based code, a data compression method based on finding a context-free grammar that accurately represents a given input. —David Eppstein (talk) 21:28, 10 May 2023 (UTC)Reply[reply]
I'm finding the following two books and I don't know if they are what you think best: (1) 'Grammatical Inference: Learning Automata and Grammars' by Colin de la Higuera, and (2) 'Grammatical Inference: Algorithms, Routines and Applications' (Studies in Computational Intelligence Book 673) by Wojciech Wieczorek. I don't know either author and maybe you can direct me to the better book since these are hard to get books. Your idea about switching to the study of grammatical inference in CFG grammars (rather than formal machines as I suggested) might be useful though there appears to be no mention of these algorithms in my edition of Hopcroft and Ullman's automata theory book. It might be useful if you could locate any insight into whether this question is even solvable/unsolvable for CFG grammars? ErnestKrause (talk) 15:19, 11 May 2023 (UTC)Reply[reply]
Sorry, no idea about the books.
The focus of learning theory tends to be less about whether a fixed input has a solution and more on whether some principle like parsimony or approximate probable correctness will lead you to eventually find a good solution over a stream of inputs. So while the theory of automata in Hopcroft and Ullman is essential for this area, the decidability or not of problems in this area may be less so. —David Eppstein (talk) 16:22, 11 May 2023 (UTC)Reply[reply]
The article for this at Wikipedia appears to be the start article for Grammar induction, which seems to be in need of some enhancement. From your previous comments, that is still leaving the questions about whether the question for a push-down automaton is still open as to being either P=NP or even decidable; do you have any thoughts on how the proof might be sketched out for either showing that its P=NP or if its even decidable? At present that would fill in the gap in the Wikipedia Turing Test article, for which I currently have citations only for FSMs and automata limited to logical connectives, but not for push-down automata. Sketching a proof for either P=NP or a statement about decidability would fill in that gap if you have any thoughts or references for this question. ErnestKrause (talk) 14:09, 13 May 2023 (UTC)Reply[reply]
What is the precise formulation of the decision problem whose decidability you think should be open? It cannot be "can the input be described by a context free grammar" because the answer is always yes. It cannot be "can the input be described by a context free grammar of complexity at most K" because that's obviously in NP. —David Eppstein (talk) 16:00, 13 May 2023 (UTC)Reply[reply]
The problem is stated as a subsection planned for the Wikipedia Turing test article, which deals with the subproblem of the Turing test when limited to formal machines. First, is the question decidable for looking at the input/output sequences for automata limited to using logical connectives alone; the answer here is that its decidable and an exponential algorithm can be given for it. Second, Rivest's paper which you linked states that at the next level of the automata hierarchy, that looking at input/output sequences, then there exists a cubic algorithm for determining if it is an FSM. Third, and unanswered here, is if there exists an algorithm that looks at input/output sequences which determines if the sequence is coming from a pushdown automaton; is this question decidable, and if it is, then is it NP-complete for PDA? That I think would be the precise formulation for this missing detail. ErnestKrause (talk) 14:26, 15 May 2023 (UTC)Reply[reply]
You're still not being precise enough. EVERY sequence can be generated by a FSM. EVERY sequence can be generated by a pushdown automaton. So the question "can this sequence be generated by this kind of automaton" is trivially decidable. It is decided by an algorithm that ignores its input and always answers yes. The only question is, do those automata provide any explanatory power by being significantly smaller than the sequence itself. —David Eppstein (talk) 15:55, 15 May 2023 (UTC)Reply[reply]
In the case of automata using only logical connectives, then a proper formation of the question would state, without loss of generality, to start with three binary input variables for the input, and then to observe the output. One then enumerates the inputs for all 3 digit combinations of zeroes and ones, and then collecting the observed outputs into a table mapping these inputs into the observed output values. If the input/output behavior is consistent, then the inference is that for 3 input binary variables, that the automaton being observed can be modeled by logical connectives. If not, however, then an automaton more complex (more than just logical connectives) is needed and possibly an FSM could then be constructed which does model and predict more complex input-output behavior. Rivest says he has an order three polynomial-time algorithm to determine if there exists an FSM which does successfully model observed input-output behavior for a given sample. If Rivest's algorithm fails, then an FSM does not exist by his construction which models the input-output behavior being sampled, and either a pushdown automaton or Turing machine is needed to perform the task. Rivest does not give the algorithm for doing this for either Pushdown automata or Turing machines, implying that he has set up his input-output experiment using a separate perspective from the one which you have just presented. Can Rivest's FSM approach be restated for pushdown automata in a way comparable to what he did accomplish for FSM in his paper which you linked above. ErnestKrause (talk) 23:44, 15 May 2023 (UTC)Reply[reply]

Removal of BLP PRODs[edit]

Hey @David Eppstein, could you correct me if I'm wrong here but I believe the articles I tagged that you reverted are covered under WP:BLP. I based my decision to PROD those articles on WP:BDP "Anyone born within the past 115 years (on or after 17 May 1908) is covered by this policy unless a reliable source has confirmed their death.". Since there wasn't a source I figured this was alright, if I'm wrong could you point me to the correct resource so this doesn't happen again. Thanks for your time and helping me understand things better. Dr vulpes (💬📝) 01:58, 19 May 2023 (UTC)Reply[reply]

You do know what the L in BLP stands for, right? It stands for living. I removed your BLP prods, which are intended only for people who are living or very recently dead, because they were on biographies of people who were not living and were not even recently dead. Zhao Wenfu died in 1990, Wen Minsheng in 1997, Song Yuxi in 2000, etc. They are not eligible for BLP prod. WP:BLPPROD clearly states "To be eligible for a BLPPROD tag, the entry must be a biography of a living person". The material you are quoting for WP:BDP is for people who might reasonably be considered living because we do not know what happened to them. There is no reasonable way to construe it as being about people with clearly stated death dates long ago. —David Eppstein (talk) 05:11, 19 May 2023 (UTC)Reply[reply]
Ahhh thanks for clearing that up makes complete sense. Guess I misunderstood the requirements. Thank you. Dr vulpes (💬📝) 07:01, 19 May 2023 (UTC)Reply[reply]

Appeal withdrawn[edit]

Appeal withdrawn, as per your advice Jack4576 (talk) 08:07, 22 May 2023 (UTC)Reply[reply]

Good choice. —David Eppstein (talk) 16:30, 22 May 2023 (UTC)Reply[reply]

Would you mind having a look at this declined AfC draft I just came across? It was declined under GNG but the subject has an h-index (64 according to google scholar) that looks high enough for WP:NPROF to me. I figure as a computer scientist you'd have a better idea of what actually is a high h-index in CS. -- asilvering (talk) 21:18, 25 May 2023 (UTC)Reply[reply]

Thanks for your feedback,but please recheck your facts[edit]

Every line is considered as a curve. This is proved in mathematics.I confess that I don't explain it in great depth but still it's a proved fact.Please recheck your facts.Appreciated your valuable feedback.Hope you next time don't revert it when I publish it.

Yuthfghds Yuthfghds (talk) 07:05, 28 May 2023 (UTC)Reply[reply]

It is not the facts, but their presentation, that I considered to be unhelpful. —David Eppstein (talk) 07:08, 28 May 2023 (UTC)Reply[reply]

Thank you, understood[edit]

Appreciated Yuthfghds (talk) 07:22, 28 May 2023 (UTC)Reply[reply]

Women in Red - June 2023[edit]

WiR Pride June 2022.png
Women in Red June 2023, Vol 9, Iss 6, Nos 251, 252, 271, 272, 273

Online events:

See also:

Tip of the month:

  • Looking for new red links? Keep an eye out for interesting and notable friends, family, or associates of your last article subject, and re-examine group photos for other women who may still need an article.

Other ways to participate:

Facebook icon.jpg Facebook | Instagram.svg Instagram | Pinterest Shiny Icon.svg Pinterest | Twitter icon.png Twitter

--Lajmmoore (talk) 09:15, 28 May 2023 (UTC) via MassMessagingReply[reply]

Dear David, I thank so much for clarifying that! I guess now, finally, the penny has dropped with me, and the fact that a fair dice is uniformly distributed confused me. Don't you think it would make sense to clarify in the main part (!) as well as in the Examples section of that article, that this has nothing to do with uniform distribution, as you pointed? I think this is really important but I don't want to include potential fallacies, so I guess I'd rather keep my hands off for the time being. Hoping for your kind support, Hildeoc (talk) 21:31, 31 May 2023 (UTC)Reply[reply]

Thank you very, very much for that swift and important improvement! I do highly appreciate this, as it brings essential clarification to thwart fallacies like mine. Well done. Per aspera ad astra! Best wishes, Hildeoc (talk) 13:59, 1 June 2023 (UTC)Reply[reply]

Descartes' theorem and Soddy circles[edit]

I notice you were just doing some cleanup of Descartes' theorem. Some of my own experimenting about the ellipses/hyperbolas with two triangle vertices as foci and passing through the third ( led me to notice that the 3-way intersections of these hyperbolas looked like they might be the two "Soddy centers". I did a quick literature search and there's apparently (as of only last year!) a paper discussing this, doi:10.31896/k.26.2. It might be worth mentioning in the relevant section there. I'm happy to make some figure(s) if there's anything that would be helpful.

It doesn't seem like these authors mentioned it but there are several other points worth exploring / noting. One that should probably get published somewhere if it hasn't been previously is that if you apply what John Conway called "Extraversion" to the triangle (treat one of the side lengths as negative) you get another set of three mutually tangent circles whose centers are the triangle vertices but with one enclosing the other two. Overall there are 4 different sets of such circles including the standard set. For each set of circles from an "extraverted" triangle, there are 2 additional tangent circles; these are centered at the 3-way intersection between two of the "vertex ellipses" and the third "vertex hyperbola" for the triangle. –jacobolus (t) 00:53, 2 June 2023 (UTC)Reply[reply]

Also see Lemoine (1890) (t) 01:22, 2 June 2023 (UTC)Reply[reply]
I added that, but is KoG really a reliable journal? Also, triangle geometry is a maze of many many lines and points and curves that coincide in different ways and I'm trying not to let that section overrun the rest of the article, as it's a bit tangential to the main topic. So while I would like to mention significant points of connection I would also like to keep it concise by confining much of the detail to the other Wikipedia articles on the specific centers and curves mentioned here. (Also not mentioned: the triangle centers in the final bullet point of that section are called "Eppstein points"). —David Eppstein (talk) 01:53, 2 June 2023 (UTC)Reply[reply]
I don't really know much about Croatian journals' editorial policies etc., though I don't really think it hurts anything to cite this kind of paper even if the journal is easy to get published in, as long as nobody is trying to use Wikipedia as a vehicle for self promotion or build whole articles out of it. –jacobolus (t) 01:59, 2 June 2023 (UTC)Reply[reply]
You should probably cite your paper there, whether or not you attach your name to these points. –jacobolus (t) 02:01, 2 June 2023 (UTC)Reply[reply]
If you want to play around with an interactive diagram:
Try toggling visibility on the different folders of circles in the list to the left. –jacobolus (t) 08:04, 2 June 2023 (UTC)Reply[reply]
There is discussion of the additional three "extraverted" sets of "Soddy circles" and "Soddy lines", including the intersection of the four lines at the de Longchamps point in
Vandeghen, A. (1964), "Soddy's Circles and the De Longchamps Point of a Triangle", American Mathematical Monthly, 71 (2): 176–179, doi:10.2307/2311750, JSTOR 2311750
I think it's worth mentioning that every triangle generates these 4 sets of circles. –jacobolus (t) 17:23, 2 June 2023 (UTC)Reply[reply]
I wonder if it would be worth splitting Soddy circle into its own article, merging Soddy line into there, and discussing more features of triangle geometry per se there, while just leaving a summary at Descartes's theorem. –jacobolus (t) 19:17, 2 June 2023 (UTC)Reply[reply]
In the current version, the first and last third of the "From triangles" section are very much on-topic for Descartes' theorem itself: the first third gives a closely related formula for the circle radius, and the last third sketches a proof of Descartes' theorem based on triangle geometry. It is only the middle third that is more tangential, and that is already more or less just a summary of related topics. So I'm not sure what savings would be possible by carving off another article. —David Eppstein (talk) 20:42, 2 June 2023 (UTC)Reply[reply]
Yes, but it seems like there is more to say here, but it gets increasingly out of scope and off topic at this article. So I am wonder if there's a good place to try to put a bunch of closely related topics in clearer context (semiperimeter, Soddy circles and lines, incircle and excircles, Gergonne triangle, Gergonne point, Extouch triangle, Nagel point, Heron's formula, Conway's "extraversion", these vertex conics, etc.) Maybe an article at extraversion (triangle) could be an okay place for it.
There is another similar cluster of topics related to altitudes and orthocenters etc., for which orthocentric system seems like a reasonable home (though that article is currently very poorly sourced and could be substantially expanded). –jacobolus (t) 22:49, 2 June 2023 (UTC)Reply[reply]

Math query[edit]

I would normally ask this of Gog the Mild, but he's on vacation and currency calculations boggle my mind. I am hoping that you will be able to help me, but if you are too busy or not interested, no worries. My subject earned 18 sous per day ca. 1865. According to this a sou = 1/20 of a livre, which was almost the same value as a franc (81 livres=80 francs) At the time 25.25 francs = 1 £, so what would be the conversion of 18 sous to GBP? If you can help, I will be very thankful. SusunW (talk) 21:24, 2 June 2023 (UTC)Reply[reply]

18 sou × (1 sou / 20 livre) × (80 frank / 81 livre) × (1 £ / 25.25 frank)
= 18 × (1/20) × (80/81) × (1/25.25) £
≈ 0.035 £
But in 1865 they would have used £/s/d, not decimal £.
So one more level of conversion: 240 d/£ gives about 8 1/2 pence.
A cross-check on the calculation: in 1837 Canada a sou and a ha'penny were the same thing [4], so the fact that we started with 18 sou and ended up with 17 ha'pennies checks out. —David Eppstein (talk) 21:33, 2 June 2023 (UTC)Reply[reply]
I appreciate your help so much. So if I cited both sources and said she earned the equivalent of 8 1/2 British pence or 17 Canadian half pennies, at the time, I'd be good? SusunW (talk) 21:50, 2 June 2023 (UTC)Reply[reply]
I would just go with the British unit and not the Canadian one (Canada wasn't even really a separate country at that time), but yes. —David Eppstein (talk) 21:51, 2 June 2023 (UTC)Reply[reply]
Very good. I am truly thankful for your help. SusunW (talk) 21:56, 2 June 2023 (UTC)Reply[reply]