Wikipedia talk:WikiProject Mathematics/Archive/2024/Feb
Which publisher mathematics books that are reliable?
[edit]While improving some mathematics articles, I'm skeptical about adding references because some of them are reliable, whereas the rest are self-publishers. As far as I'm concerned, the only reliable publishers are Dover Publications, Cambridge University Press, Oxford University Press, Springer, AMS, and many more. However, what about publishers such as Bod (Book of Demand) and Read Books Ltd.? Are these remains reliable? Dedhert.Jr (talk) 02:11, 30 January 2024 (UTC)
- Do you have a specific example book you are curious about? Sometimes if a reputable scholar self publishes a book it can be a reliable source, but some self-published books are about as credible as pseudonymous blog posts or forum comments. A book published by a university press (or similar) is clearly preferable if you can find it. –jacobolus (t) 02:41, 30 January 2024 (UTC)
- @Jacobolus I do. One example while I was trying to improve the article Wedge (geometry) is this one [1]. Dedhert.Jr (talk) 02:46, 30 January 2024 (UTC)
- That was published by Ginn & Company in 1893. They were a reliable publisher of textbooks. Old, but I suppose usable. The modern 'publisher' is just reprinting a public domain book since it has aged out of copyright. MrOllie (talk) 02:57, 30 January 2024 (UTC)
- Ah. I see. Thank you. I am somewhat confused with some sources that are self-publisher. However, it seems that the sources do not have the specific pages, as I am more skeptical in the Google Books settings that exhibit the page, but not accurate. Dedhert.Jr (talk) 02:59, 30 January 2024 (UTC)
- This is a book from 1893 published by Ginn of Boston. (Edit: beaten to the punch.) –jacobolus (t) 03:03, 30 January 2024 (UTC)
- Really appreciate your work. Dedhert.Jr (talk) 03:04, 30 January 2024 (UTC)
- That was published by Ginn & Company in 1893. They were a reliable publisher of textbooks. Old, but I suppose usable. The modern 'publisher' is just reprinting a public domain book since it has aged out of copyright. MrOllie (talk) 02:57, 30 January 2024 (UTC)
- @Jacobolus I do. One example while I was trying to improve the article Wedge (geometry) is this one [1]. Dedhert.Jr (talk) 02:46, 30 January 2024 (UTC)
- For math books (or any other specialist book), if they get reviewed (and not mocked) by other mathematicians or pedagogues in a reliable publication, they're probably fine. The nice thing about using super-popular oft-reprinted math textbooks is they're easily verifiable and unlikely to have typos. The nice thing about that hidden gem book you might like to cite is that it may have some wonderfully illustrative example problems suitable for adapting to our WP articles (unlike, say, several notoriously difficult but ubiquitous textbooks; or perhaps Dover which is generously assigned because it's cheap, but may not always be the cleanest.) SamuelRiv (talk) 03:07, 30 January 2024 (UTC)
- Dover reprints a wide variety of old out-of-print math books, some of which are solid and others of which are really excellent. It's cheap because they get the publishing rights cheaply from publishers uninterested in issuing their own new versions, so you're paying not that much more than the production cost for an average-quality paperback book. When citing material found in a Dover reprint, it's worth trying to figure out the original publisher and title (sometimes Dover changes it); usually, but not always, there's a decent scan at the Internet Archive. –jacobolus (t) 03:30, 30 January 2024 (UTC)
- (ec) : I think "Books on Demand" is self-published, so not reliable. More reliable publishers: Addison-Wesley, Prentice Hall, and Wiley (publisher). And look to see if it is reviewed by the AMS! Bubba73 You talkin' to me? 03:12, 30 January 2024 (UTC)
- But what about the case of sources that is taken from Wikipedia content? During the improvement of the same article again, I found this book with Springer as its reliable publisher, but the images are taken from the Wikipedia articles; or another example is in Reeve tetrahedra, where the journal source is taken the reference from the Wikipedia article Pick's theorem? Is it fine to pick them? Dedhert.Jr (talk) 03:21, 30 January 2024 (UTC)
- Springer is a massive publisher owned for the past couple decades by a string of financial firms and therefore not especially caring about book quality anymore. Many of the monographs they publish are mediocre, and they also publish things like conference proceedings for niche conferences which can have mixed quality. Some Springer titles have excellent content, though unfortunately nowadays they almost all are printed on demand, with horribly bad printing/binding quality. You should look at the name/reputation of the author, try reading some of the book, look at reviews, etc. to decide if the book is the best source for a particular claim; just the name of the publisher is not a particularly strong indicator. –jacobolus (t) 03:37, 30 January 2024 (UTC)
- I'm sorry to hear that about Springer - they were so good back in my day. I think their graduate-level books were called the "yellow peril". Bubba73 You talkin' to me? 04:14, 30 January 2024 (UTC)
- I think their GTM series is still good, but other Springer series can require more care. I just recently ran into a case while adding sources to conical surface of a 2015 Springer book, Encyclopedia of Analytical Surfaces, that had clearly copied from us (at least one identical sentence with an earlier appearance in our article than the publication date of the book), making it unusable per WP:CIRCULAR. That's one reason I sometimes look for older books as sources: as well as being freely available online, they are not going to have that problem. And the mathematics may occasionally be stated in an out-of-date style, but in most cases it is unlikely to become incorrect. —David Eppstein (talk) 07:40, 30 January 2024 (UTC)
- There's even worse available in Springer, see e.g. the last chapter of https://doi.org/10.1007/978-3-031-29046-6 which manages to combine modules of sheaves, Ricci flow, the positive mass theorem, and K-theory in the study of echocardiography. It reads like GPT-2. Gumshoe2 (talk) 13:32, 30 January 2024 (UTC)
- Wow. Anyway, I think Cambridge University Press is reliably good, and Princeton University Press almost as good. Oxford University Press is likely good as well but I don't have so much experience with them. Springer and CRC are both rather more careless as these examples suggest. Unfortunately MAA Press can sometimes read as thrown-together web scrapings as well. The more researchy imprints of AMS are better. You might expect the Dover mathematics books to be dubious (because cheap paperbacks), but they are surprisingly good, because they are mostly well-chosen reprints of older books. (Disclaimer: I have published with both Cambridge and Springer.) —David Eppstein (talk) 18:26, 30 January 2024 (UTC)
- @David Eppstein My question regarding about this problem is how do I know that a publisher is reliable? Dedhert.Jr (talk) 02:57, 2 February 2024 (UTC)
- If you have serious doubts then you can ask on WP:RSN but they are going to say that all the publishers we are discussing are good enough. Beyond that I think it comes down to past experience and individual evaluation of how polished and accurate individual sources appear to be. —David Eppstein (talk) 03:09, 2 February 2024 (UTC)
- You have to use your judgment. You can look at both internal signals such as how clearly written it is and whether the authors seem careful or sloppy about citing past work, or external signals such as what reviewers said about it, how highly cited the source (or other sources by the same author) is, what honors the author has received, and so on. Even without looking at external signals, some works will give you the sense that a deep expert wrote them with loving attention to detail and were then carefully copyedited, have nicely drawn figures, etc., and others just seem slapped together carelessly. YMMV. –jacobolus (t) 03:30, 2 February 2024 (UTC)
- Noted. Thank you. Dedhert.Jr (talk) 04:53, 2 February 2024 (UTC)
- @David Eppstein My question regarding about this problem is how do I know that a publisher is reliable? Dedhert.Jr (talk) 02:57, 2 February 2024 (UTC)
- Wow. Anyway, I think Cambridge University Press is reliably good, and Princeton University Press almost as good. Oxford University Press is likely good as well but I don't have so much experience with them. Springer and CRC are both rather more careless as these examples suggest. Unfortunately MAA Press can sometimes read as thrown-together web scrapings as well. The more researchy imprints of AMS are better. You might expect the Dover mathematics books to be dubious (because cheap paperbacks), but they are surprisingly good, because they are mostly well-chosen reprints of older books. (Disclaimer: I have published with both Cambridge and Springer.) —David Eppstein (talk) 18:26, 30 January 2024 (UTC)
- There's even worse available in Springer, see e.g. the last chapter of https://doi.org/10.1007/978-3-031-29046-6 which manages to combine modules of sheaves, Ricci flow, the positive mass theorem, and K-theory in the study of echocardiography. It reads like GPT-2. Gumshoe2 (talk) 13:32, 30 January 2024 (UTC)
- I think their GTM series is still good, but other Springer series can require more care. I just recently ran into a case while adding sources to conical surface of a 2015 Springer book, Encyclopedia of Analytical Surfaces, that had clearly copied from us (at least one identical sentence with an earlier appearance in our article than the publication date of the book), making it unusable per WP:CIRCULAR. That's one reason I sometimes look for older books as sources: as well as being freely available online, they are not going to have that problem. And the mathematics may occasionally be stated in an out-of-date style, but in most cases it is unlikely to become incorrect. —David Eppstein (talk) 07:40, 30 January 2024 (UTC)
- I'm sorry to hear that about Springer - they were so good back in my day. I think their graduate-level books were called the "yellow peril". Bubba73 You talkin' to me? 04:14, 30 January 2024 (UTC)
- Images wouldn't give me pause, especially when they are given proper attribution. Likewise material taken from wikipedia in an otherwise original work can be assumed to be have been vetted by that author, and thus becomes as reliable as the author. Headbomb {t · c · p · b} 03:38, 30 January 2024 (UTC)
- Springer is a massive publisher owned for the past couple decades by a string of financial firms and therefore not especially caring about book quality anymore. Many of the monographs they publish are mediocre, and they also publish things like conference proceedings for niche conferences which can have mixed quality. Some Springer titles have excellent content, though unfortunately nowadays they almost all are printed on demand, with horribly bad printing/binding quality. You should look at the name/reputation of the author, try reading some of the book, look at reviews, etc. to decide if the book is the best source for a particular claim; just the name of the publisher is not a particularly strong indicator. –jacobolus (t) 03:37, 30 January 2024 (UTC)
- But what about the case of sources that is taken from Wikipedia content? During the improvement of the same article again, I found this book with Springer as its reliable publisher, but the images are taken from the Wikipedia articles; or another example is in Reeve tetrahedra, where the journal source is taken the reference from the Wikipedia article Pick's theorem? Is it fine to pick them? Dedhert.Jr (talk) 03:21, 30 January 2024 (UTC)
WikiProject proposal related to Mathematics: Geometry
[edit]You can view the proposal here. Looking for interested Wikipedians. Writehydra (talk) 04:57, 2 February 2024 (UTC)
Someone has just set up/revived this project. I am wondering whether it would be better placed inside the maths project because it seems a rather narrow field for a new project and will probably not become active as a result — Martin (MSGJ · talk) 19:21, 19 January 2024 (UTC)
- As far as I can see the main use for this revival is quality assessment of articles: we can now find the Category:WikiProject Polyhedra articles, sorted by importance and by quality. This may be helpful after the banner shell Borg nuked the
|field=
parameter we used to use to find sorted collections of articles in mathematics subtopics. —David Eppstein (talk) 00:39, 20 January 2024 (UTC)- It wasn't the "banner shell borg" that nuked the 'field' parameter. This was changed in 2020 in special:diff/986062061 by @MSGJ. It could plausibly be un-changed somehow, but I don't know how worthwhile it would be (then the banner shell bots would need to be reconfigured to ignore the restored parameter, etc.). –jacobolus (t) 17:43, 20 January 2024 (UTC)
- @David Eppstein, @Jacobolus, @MSGJ I'm planning for reviving this WikiProject. However, it seems that I forgot what was the reason. One thing that might pushed me to do so currently is to improve many articles on polyhedron in which tables only and facts are unsourced—sorry I totally forgot. Dedhert.Jr (talk) 13:54, 2 February 2024 (UTC)
- It wasn't the "banner shell borg" that nuked the 'field' parameter. This was changed in 2020 in special:diff/986062061 by @MSGJ. It could plausibly be un-changed somehow, but I don't know how worthwhile it would be (then the banner shell bots would need to be reconfigured to ignore the restored parameter, etc.). –jacobolus (t) 17:43, 20 January 2024 (UTC)
Coxeter diagram
[edit]For the background, I and @Jacobolus are discussing the table explaining the bipyramids while improving the article Bipyramid. One problem if someone applied the table {{Bipyramids}} created by @Tomruen is the Coxeter diagram, which is somewhat technical to understand. Can someone explain what are these notations, and how to understand them? The article Coxeter diagram exists, but the content is already problematic. Dedhert.Jr (talk) 09:39, 27 January 2024 (UTC)
- My understanding is that Coxeter diagrams describe Coxeter groups, which in this specific case can be described as groups generated by the reflections of a spherical triangle across its edges. The Coxeter diagram has a vertex for each mirror (three vertices for the three sides of the triangle) and an edge or non-edge indicating the angle at which two mirrors meet (no edge for π/4, an undecorated edge for π/6, and a number n for π/2n). So for a spherical bipyramid, the given diagrams describe the shapes of the faces and represent the symmetries generated by reflections across edges of the bipyramid. —David Eppstein (talk) 19:16, 27 January 2024 (UTC)
- Coxeter diagrams for polyhedra represent Wythoff constructions of mirror planes, each node is a mirror. A ringed mirror is "active", with generated point off the mirror plane, generating an edge. This constructs unform polyhedrons. Uniform dual polyhedra are represented with a vertical line through the ringed node(s) of the dual polyhedron. They are indeed somewhat technical, and could be removed, but a useful system with basic understanding. Tom Ruen (talk) 23:29, 27 January 2024 (UTC)
- @Tomruen The question is what are the basics to understand this notation? How do I write the notation for every polyhedron, including the bipyramid as mentioned by @David Eppstein? Let's see if I looked up at the table again, where there are some Coxeter diagrams representing the bipyramids.
Bipyramid name |
Digonal bipyramid |
Triangular bipyramid |
Square bipyramid |
Pentagonal bipyramid |
Hexagonal bipyramid |
... | Apeirogonal bipyramid |
---|---|---|---|---|---|---|---|
Polyhedron image |
![]() |
![]() |
![]() |
![]() |
... | ||
Spherical tiling image |
![]() |
![]() |
![]() |
![]() |
![]() |
Plane tiling image |
![]() |
Face config. | V2.4.4 | V3.4.4 | V4.4.4 | V5.4.4 | V6.4.4 | ... | V∞.4.4 |
Coxeter diagram |
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() |
... | ![]() ![]() ![]() ![]() ![]() |
- I've seen more examples of Coxeter diagrams describing four Platonic solids.[2], but it is somewhat different than our articles. Dedhert.Jr (talk) 04:05, 28 January 2024 (UTC)
- @Dedhert.Jr The reason I think the discussion of symmetry should be kept (and indeed extended) is that crystallography is the primary application of bipyramids and related shapes. It would be helpful if someone who was an expert crystallographer could take a look though. In my opinion scalenohedron should be split into its own article with a short summary at bipyramid. The current text we have about this and the other related polyhedra is somewhat incoherent and overly technical, and seems to be a bit idiosyncratic / perhaps with some original research thrown in. –jacobolus (t) 07:32, 28 January 2024 (UTC)
- @Jacobolus The only thing I could not create this new article is because of lack of sources. I could only found the definition of scalenohedra, but not the symmetry and many other related topics. If it does, I'm aware that it will be redirected again under WP:BLAR; this was already happened with our articles such as the family of duoprisms. Dedhert.Jr (talk) 09:56, 28 January 2024 (UTC)
- A search for "scalenohedron" has >1000 sources in Google scholar. It's unquestionably a notable topic about which much can be said that is verifiable. I'm not an expert so couldn't by any means write that article off the top of my head, but if someone wants to put in the work to do a literature review it also shouldn't be inordinately difficult. –jacobolus (t) 10:01, 28 January 2024 (UTC)
- Really? I never knew that numerous sources mentions about it. Unfortunately, I don't think I can make it because of the restricted access. Maybe I leave it to someone else. Dedhert.Jr (talk) 10:13, 28 January 2024 (UTC)
- A search for "scalenohedron" has >1000 sources in Google scholar. It's unquestionably a notable topic about which much can be said that is verifiable. I'm not an expert so couldn't by any means write that article off the top of my head, but if someone wants to put in the work to do a literature review it also shouldn't be inordinately difficult. –jacobolus (t) 10:01, 28 January 2024 (UTC)
- @Jacobolus The only thing I could not create this new article is because of lack of sources. I could only found the definition of scalenohedra, but not the symmetry and many other related topics. If it does, I'm aware that it will be redirected again under WP:BLAR; this was already happened with our articles such as the family of duoprisms. Dedhert.Jr (talk) 09:56, 28 January 2024 (UTC)
Aside: should Icosahedral bipyramid, Tetrahedral bipyramid, Cubical bipyramid, Dodecahedral bipyramid really be articles here? It seems like original research. The only source given is a personal web page, and I can't find any reliable sources about this (I can find this example of "dodecahedral bipyramid" being used to mean something clearly different). –jacobolus (t) 08:02, 28 January 2024 (UTC)
- I have no idea how to deal with the original researches. Maybe AfD is the only option if there are no sources mentions about them? Dedhert.Jr (talk) 10:01, 28 January 2024 (UTC)
- I also can't find any reliable sources using the name "Blind polytope". The cited sources are user:Tamfang's website https://bendwavy.org and the "Polytope Wiki". user:Tomruen do you have any reliable sources for these topics, or is this a neologism you and/or Tamfang coined? –jacobolus (t) 15:37, 28 January 2024 (UTC)
- I'm don't know who first offered the categorical name. It could be moved to Convex polytopes with regular facets. Johnson solids were referenced in 1969 as "Convex polyhedra with regular faces", while the actual lists neglects regular and semiregular forms. I can't say who gave Johnson the honorific name, might also just be websites like MathWorld]. It could be Klitzing's reference is an original source [3]. Tom Ruen (talk) 04:47, 29 January 2024 (UTC)
- The earliest example of "Johnson solids" Google scholar finds is Rankin, John R. (1988). "Classes of polyhedra defined by jet graphics". Computers & Graphics. 12 (2): 239–254. doi:10.1016/0097-8493(88)90036-2. But it's certainly plausible that Klitzing and/or MathWorld popularized the name. –jacobolus (t) 10:49, 29 January 2024 (UTC)
- To clarify, I host Richard Klitzing's polytope pages at bendwavy.org, but have no hand in writing them. —Tamfang (talk) 05:10, 3 February 2024 (UTC)
- I'm don't know who first offered the categorical name. It could be moved to Convex polytopes with regular facets. Johnson solids were referenced in 1969 as "Convex polyhedra with regular faces", while the actual lists neglects regular and semiregular forms. I can't say who gave Johnson the honorific name, might also just be websites like MathWorld]. It could be Klitzing's reference is an original source [3]. Tom Ruen (talk) 04:47, 29 January 2024 (UTC)
The article Mina Ossiander has been proposed for deletion because of the following concern:
Adjunct prof with h-index of 7
While all constructive contributions to Wikipedia are appreciated, pages may be deleted for any of several reasons.
You may prevent the proposed deletion by removing the {{proposed deletion/dated}}
notice, but please explain why in your edit summary or on the article's talk page.
Please consider improving the page to address the issues raised. Removing {{proposed deletion/dated}}
will stop the proposed deletion process, but other deletion processes exist. In particular, the speedy deletion process can result in deletion without discussion, and articles for deletion allows discussion to reach consensus for deletion. Abductive (reasoning) 05:07, 3 February 2024 (UTC)
- As the page is presently written there’s no evidence of notability, but low h index should be rejected as a standard to judge by. Gumshoe2 (talk) 10:14, 3 February 2024 (UTC)
- True, but false statements about academic rank are an even worse standard. —David Eppstein (talk) 20:04, 4 February 2024 (UTC)
- The PROD has been removed. In the future, proposed deletions of interest to this project are best listed at Wikipedia:WikiProject Deletion sorting/Mathematics. --JBL (talk) 17:25, 3 February 2024 (UTC)
- Nothing in the article seems particularly notable here. It's just one more person who got a PhD in mathematics and the article lists irrelevant details of their personal life and the title of the thesis and name of the advisor. Not sure why that person was selected as worthy of a Wikipedia article. PatrickR2 (talk) 21:57, 3 February 2024 (UTC)
- Anyone can, of course, send the article to WP:AfD if they want to. --JBL (talk) 19:40, 4 February 2024 (UTC)
Quantity of references in mathematics articles
[edit]I'm trying to get a better understanding of the content assessment criteria specifically for mathematics articles. I have noticed that mathematics articles often do not include many references. Take for example Ordinal number, many sections in this article don't have any references, or may only provide a single reference for a specific sentence or word in a section. Still, this article was once rated GA, and today is still rated B, indicating only minor problems exist. Can someone provide some advice on this? Mokadoshi (talk) 03:07, 5 February 2024 (UTC)
- The Rater tool gives Ordinal number as C rating. That means it most resembles articles that humans have evaluated as C across Wikipedia.
- Your observation also applies to physics article. The long-time editors have said that many articles written back in the 2000s where just typed in. But another factor is the huge number of articles.
- If the content assessment for an article seems wrong, just change it. Or improve the article up to the current rating ;-) Johnjbarton (talk) 03:17, 5 February 2024 (UTC)
- You are right that math articles on Wikipedia are systematically reference-light. I see two reasons.
- First, mathematicians were early adopters of Wikipedia. Many math articles were written into decent shape long before there was so much insistence on reliable sources (as Johnjbarton implied above). And, in later years, many editors have not felt strong motivation to systematically add references to long-established, pretty good text.
- Second, even professional math articles are reference-light compared to articles I've seen in the natural sciences, history, etc. This is because math is almost pure logic, which the diligent reader can verify or refute for themselves. (Moreover, the policy WP:CALC is interpreted to give broad leeway in math articles.) Mathematicians are simply not acculturated to reference-heavy writing.
- Let me emphasize that I am not defending the lack of references in Wikipedia math articles. I am merely enunciating the reasons, as I see them. Regards, Mgnbar (talk) 03:51, 5 February 2024 (UTC)
- Professional math sources also typically (not always) do a poor job of citing original authors or describing their work in its own terms: misnamed objects and ideas abound, a wide variety of mythical origin stories persist, and treatment of earlier sources is often anachronistic to the point of pure fantasy. Reading works mentioning math history requires some skepticism and care, as plenty of nonsense gets passed down uncritically. Recent professional mathematical historians have gotten better about this and are more careful to separate what is actually known from invented folklore, especially about older material. Wikipedia math articles could do a better job accurately describing (and citing original sources relevant to) the history of pretty much every topic. –jacobolus (t) 06:20, 5 February 2024 (UTC)
- Of course, as soon as I posted this discussion, I came across this discussion from 2006 about this topic, which includes a link to Wikipedia:Scientific citation guidelines. If I'm understanding it correctly: On Wikipedia it is not required to provide a citation on every instance of uncontroversial knowledge. However, in scientific articles "uncontroversial knowledge" is defined broadly to be knowledge that would be known by anyone with an underground background in the field of study, and is covered in any common or obvious books on the topic -- even if this is not known by a layperson. Still, each section should have at least one reference to one of these common or obvious books on the topic, and these can be cited in the first or last sentence of a section (the rest of the section need no references if it contains uncontroversial knowledge).
- Am I correct in this? If so, it might be helpful to include a link to this page on Wikipedia:WikiProject_Mathematics/Assessment. This might be obvious to many people, but that is the page I looked at before posting this discussion and found nothing there about citation guidelines. Mokadoshi (talk) 03:58, 5 February 2024 (UTC)
- @Mokadoshi: I think the reader's background is supposed to be assumed to be undergraduate-level, rather than secret or subterranean! — MarkH21talk 04:12, 5 February 2024 (UTC)
My very rough personal scale is: if it supports more than one section, with more than one sentence in them and with a lead that properly summarizes those sections, it is at least start-class (otherwise it's a stub). If it covers the main ground and is adequately referenced but not very well written or organized, or (as in the case at hand) is thorough, competently written and organized, but badly sourced, it is C-class. If it is thorough, well-written, and mostly well-sourced, but maybe with a few topics missing or a few unsourced claims, it is B-class. And if it has all that, it is written as accessibly as reasonably possible, and careful read-throughs and literature searches can find nothing obvious that is still in need of improvement, then it is either GA-class or should be a nominee for GA-class. I don't consider FA-class worth the effort, so the scale stops there. As for Wikipedia:Scientific citation guidelines: I think it is outdated and should be marked historical. I don't think its recommendations to leave background material unsourced, or sourced only to general reference sections at the end, reflect current Wikipedia practice. —David Eppstein (talk) 06:33, 5 February 2024 (UTC)
Invite to join the February 2024 Unreferenced Backlog Drive
[edit]WikiProject Unreferenced articles | February 2024 Backlog Drive | 1 + 1 ≠ 2 |
Many math articles do not have references. This drive wants to change that. The aim of this drive is to cite all unreferenced articles on Wikipedia and ensure its reliability.
|
CactiStaccingCrane (talk) 14:55, 21 January 2024 (UTC)
- @CactiStaccingCrane: Where is the list of unreferenced mathematics articles? Michael Hardy (talk) 18:51, 21 January 2024 (UTC)
- I'd recommend not turning improving math articles into a race/speedrun. It would be significantly more valuable if someone picked one highly viewed start or C class math article and spent a few hours or days making substantial improvements, instead of trying to spend a few minutes each adding hastily chosen references as footnotes to sections that are already well written and describe well-known topics in a standard way. References found in haste in this way by searching (rather than either chosen by an expert who already knows the literature or carefully found by someone willing to do some significant literature research) tend to be low quality, and ticking off a binary checklist of "has a reference" vs. "doesn't have a reference" is not really very valuable to the Wikipedia project. –jacobolus (t) 20:39, 21 January 2024 (UTC)
- No, we are not adding referenced to well-sourced articles. We are adding references to articles that do not have any citations at all and risk being AfDed for not demonstrating that this concept actually exist, especially for mathematics articles where "this concept exists" is not easy to verify. So yes, an article "has a reference" is indeed much better than an article that "does not have a reference". CactiStaccingCrane (talk) 00:34, 22 January 2024 (UTC)
- Nevertheless, you are free to chip in at 'check for sloppy work in the drive' thread, as quality control should also play a big part in the drive. CactiStaccingCrane (talk) 00:37, 22 January 2024 (UTC)
- It's really not practically helpful to readers to add shitty references to well-written articles, whether or not they already had sources. If something is verifiable it's not going to get deleted, and people sending obviously verifiable stuff to AFD based on a lack of footnotes are wasting everyone's time.
- My point is: please don't try to add references to articles as some kind of speed race game in search of little meaningless wiki awards for being the fastest, but instead take the time to do the work properly so it doesn't need to be done over later.
- If you (anyone reading along) want to find a reference to some mathematical article which is currently unsourced, take the trouble to find a source which is well written and respected, and then fill out a carefully written citation. This often takes skimming several sources, checking their citation counts in the academic literature, considering which one is most appropriate for the intended audience of a Wikipedia page, etc. It's even better to figure out a bit of the history of the topic and give some concept of the priority for an idea, or give multiple sources appropriate for different audiences. –jacobolus (t) 00:48, 22 January 2024 (UTC)
- It might be helpful for you to read this discussion: Wikipedia:Requests for comment/Deletion of uncited articles CactiStaccingCrane (talk) 02:26, 22 January 2024 (UTC)
- The tone of your response leads me to think that you care more about the superficial appearance of having little clicky blue footnotes than about the quality of the sourcing, and that you would like the outcome of the sort of speed race game that jacobolus warns against. Perhaps you can persuade me otherwise. —David Eppstein (talk) 02:34, 22 January 2024 (UTC)
- Ok, just to clarify, I do agree that in the most ideal scenario, we would get people to give many citations to reliable sources for each article, enough so that the article is unambiguously notable and will be speedily keep in an AfD discussion. This drive's main goal is not to cite small percent of these uncited articles are well-written (e.g. 238 math articles/113000 total), but to cite barely notable articles that you might find in Special:Random, like articles about random places or random people. For a participant in the drive, it is much more efficient to cite these stubs to get those "little clicky blue footnote points" that you mentioned rather than citing math articles that require a lot of effort to find a reliable source. And frankly, who cares about these articles? For all we know citing these articles might not guarantee these articles will be kept. If somebody truly care about an article's existence, they would go above and beyond to cite these articles to meet GNG.
- I do agree that this drive is not perfect, and I do want to say that I appreciate that you have raised concerns about the drive and about uncited articles in general. We might be relying too much on good faith here and a better solution might be needed. But it is ludicrous for us to require new articles to be fully cited while we are not making an active effort to cite our older unreferenced articles. We can do better than this. Because a lot of math articles are not cited, I want to send this notice ahead of time so that people like you who knowledgeable about the topic area can help address these articles. CactiStaccingCrane (talk) 02:54, 22 January 2024 (UTC)
- @CactiStaccingCrane: As predicted: [4] —David Eppstein (talk) 08:13, 5 February 2024 (UTC)
- It seems like there are some editors who are really excited about deleting stuff (to rack up points?) whether or not it is notable or verifiable, and are constantly hunting for excuses to do so. That's indeed a problem, but the better solution is to educate them about the Wikipedia project's goals and community process, not to let them push you into wasting time jumping through bureaucratic hoops.
- Again, I don't have any problem with the goal of improving the sourcing of poorly sourced math articles. I just don't think this should be turned into a speed game where the person who adds the most footnotes wins irrespective of whether they add any value. –jacobolus (t) 02:42, 22 January 2024 (UTC)
- Anyway, don't let me discourage anyone from participating here if you want. I just hope anyone trying to add sources to math articles is doing so carefully and thoughtfully rather than prioritizing speed. My personal experience is that it takes me significant amounts of time and effort to carefully add sources for one small subsection of one article, mostly spent reading and thinking. This contest has badges for adding sources to 300, 500, or 750 articles during the month of February. I can't even imagine doing good work at anywhere close to that pace, even if I spent 10 hours per day for a month doing nothing but adding sources to math articles. YMMV. –jacobolus (t) 02:53, 22 January 2024 (UTC)
- I appreciate your concerns about the drive in general. I don't want to give a false assurance here and say that the reviewing process will spot 100% fake references and bogus attempts. But again, the majority of article that will be addressed by these articles are geostubs, biostubs, school/universities and obscure music bands. As an example, there are 238 uncited math articles compared to ~113000 in total. Statistically speaking it is unlikely for people to randomly come across these articles and participants who will cite these math articles are likely to be those knowledgeable about mathematics and seek for these articles deliberately. CactiStaccingCrane (talk) 02:59, 22 January 2024 (UTC)
geostubs, biostubs, school/universities and obscure music bands
– what's the point of getting a lot of people to spend their time marginally improving obscure never-read articles that nobody cares about? Personally I think it would be a better use of time for anyone here to make their goal for February something like: "improve one high-importance math article from C class to B class in the next month", including doing a thorough literature review, drawing some nice diagrams, and writing a clear and concise overview of the current expert consensus. Again, YMMV. –jacobolus (t) 03:04, 22 January 2024 (UTC)- I applaud your philosophy here, but not to be condescending, people can have different interests :) CactiStaccingCrane (talk) 03:35, 22 January 2024 (UTC)
- I appreciate your concerns about the drive in general. I don't want to give a false assurance here and say that the reviewing process will spot 100% fake references and bogus attempts. But again, the majority of article that will be addressed by these articles are geostubs, biostubs, school/universities and obscure music bands. As an example, there are 238 uncited math articles compared to ~113000 in total. Statistically speaking it is unlikely for people to randomly come across these articles and participants who will cite these math articles are likely to be those knowledgeable about mathematics and seek for these articles deliberately. CactiStaccingCrane (talk) 02:59, 22 January 2024 (UTC)
- Anyway, don't let me discourage anyone from participating here if you want. I just hope anyone trying to add sources to math articles is doing so carefully and thoughtfully rather than prioritizing speed. My personal experience is that it takes me significant amounts of time and effort to carefully add sources for one small subsection of one article, mostly spent reading and thinking. This contest has badges for adding sources to 300, 500, or 750 articles during the month of February. I can't even imagine doing good work at anywhere close to that pace, even if I spent 10 hours per day for a month doing nothing but adding sources to math articles. YMMV. –jacobolus (t) 02:53, 22 January 2024 (UTC)
- The tone of your response leads me to think that you care more about the superficial appearance of having little clicky blue footnotes than about the quality of the sourcing, and that you would like the outcome of the sort of speed race game that jacobolus warns against. Perhaps you can persuade me otherwise. —David Eppstein (talk) 02:34, 22 January 2024 (UTC)
- It might be helpful for you to read this discussion: Wikipedia:Requests for comment/Deletion of uncited articles CactiStaccingCrane (talk) 02:26, 22 January 2024 (UTC)
- No, we are not adding referenced to well-sourced articles. We are adding references to articles that do not have any citations at all and risk being AfDed for not demonstrating that this concept actually exist, especially for mathematics articles where "this concept exists" is not easy to verify. So yes, an article "has a reference" is indeed much better than an article that "does not have a reference". CactiStaccingCrane (talk) 00:34, 22 January 2024 (UTC)
Well, despite the disagreements above, the list of unreferenced articles provided by CactiStaccingCrane does at least provide a helpful way of finding some mathematics articles that are on good topics but need help, although many others are on marginal topics (that I would rather not waste my time on). I already took advantage of that list to find and clean up the sourcing in Arithmetic–geometric mean and Basel problem. I think there are plenty of others worthy of attention there. —David Eppstein (talk) 08:42, 22 January 2024 (UTC)
- I hope you find them useful :D CactiStaccingCrane (talk) 10:59, 22 January 2024 (UTC)
- From my perspective, this is not about adding references, but rather improving the articles. I have seen the tables listing many bunches of articles that may need to be improved. I do think that there are many articles without being tagged that may need to be improved as well, especially with the article that contains many no-context-and-what-to-do-tables. Dedhert.Jr (talk) 03:34, 23 January 2024 (UTC)
- One could also view the announcement here as a warning: fix up your unreferenced articles by the end of January, before the cleanup drive starts, or some eager cleanup drive participant will get to them without as much care for detail. —David Eppstein (talk) 04:52, 23 January 2024 (UTC)
- Why do you think that adding one citation to a mathematics article so damaging? Or let me be straightforward: why do you view the drive so negatively? CactiStaccingCrane (talk) 05:39, 23 January 2024 (UTC)
- I think that it is important to provide high-quality sources for stubs on significant topics in mathematics, and to do something to clean out the many bad stubs on insignificant topics. I think it is important that this be done by people who are interested in the mathematics (not necessarily professional mathematicians!), and not just those interested in getting gold stars for sourcing lots of things, because interest leads to expertise and expertise leads to better choices for the sourcing. I think that having people add low-quality sources to indiscriminately chosen articles causes problems in two directions: it does not change the need for the significant topics to have high-quality sources, but it obscures the fact that they need them, and it does not change the need to clean out the insignificant topics, but by removing them from the lists of problem articles it makes it harder to find them. I hope that the sourcing drive does not lead to this problem of adding low-quality sources to indiscriminately chosen articles but I worry that it might. —David Eppstein (talk) 05:59, 23 January 2024 (UTC)
- I think that's true and I should list some examples of what sources is encouraged/discouraged from use generally. But in general, I have to admit that this drive will be a pain in the ass to manage and it will keep me busy for the rest of this and next month ;) CactiStaccingCrane (talk) 09:59, 23 January 2024 (UTC)
- I have looked up the list, and it confused me: why is the article decagonal bipyramid classified as "cite no sources", although it is a redirect article? Dedhert.Jr (talk) 10:33, 23 January 2024 (UTC)
- Someone has redirected it recently. CactiStaccingCrane (talk) 10:35, 23 January 2024 (UTC)
- The list is only updated weekly. In this case, I judged that there were no sources giving decagonal bipyramid independent notability from bipyramid, and redirected it per WP:BLAR in preference to trying to get it deleted. —David Eppstein (talk) 17:54, 23 January 2024 (UTC)
- I just redirected the heptagonal and octagonal cases as well, and shrank the still-overwide tables. If there's anything to say about one or another specific n-gonal bipyramid, it can be put in a section of bipyramid. At the point the specific information grows to more than several paragraphs it can plausibly be re-split into a new article again. –jacobolus (t) 18:07, 23 January 2024 (UTC)
- While we're here, Bipyramid is very weakly sourced, if anyone is looking for a page to improve sources on. –jacobolus (t) 19:39, 23 January 2024 (UTC)
- Thank you for your work! CactiStaccingCrane (talk) 01:01, 24 January 2024 (UTC)
- I already have some planning about the sketch of the structure and the improvement of the content, but there are some missing topics; for example, the bipyramidal graph is missing. I remember that there is a source that mentions about the graph of bipyramids, but it is not free access. Dedhert.Jr (talk) 01:51, 24 January 2024 (UTC)
- The list is only updated weekly. In this case, I judged that there were no sources giving decagonal bipyramid independent notability from bipyramid, and redirected it per WP:BLAR in preference to trying to get it deleted. —David Eppstein (talk) 17:54, 23 January 2024 (UTC)
- Someone has redirected it recently. CactiStaccingCrane (talk) 10:35, 23 January 2024 (UTC)
- I have looked up the list, and it confused me: why is the article decagonal bipyramid classified as "cite no sources", although it is a redirect article? Dedhert.Jr (talk) 10:33, 23 January 2024 (UTC)
- I think that's true and I should list some examples of what sources is encouraged/discouraged from use generally. But in general, I have to admit that this drive will be a pain in the ass to manage and it will keep me busy for the rest of this and next month ;) CactiStaccingCrane (talk) 09:59, 23 January 2024 (UTC)
- I think that it is important to provide high-quality sources for stubs on significant topics in mathematics, and to do something to clean out the many bad stubs on insignificant topics. I think it is important that this be done by people who are interested in the mathematics (not necessarily professional mathematicians!), and not just those interested in getting gold stars for sourcing lots of things, because interest leads to expertise and expertise leads to better choices for the sourcing. I think that having people add low-quality sources to indiscriminately chosen articles causes problems in two directions: it does not change the need for the significant topics to have high-quality sources, but it obscures the fact that they need them, and it does not change the need to clean out the insignificant topics, but by removing them from the lists of problem articles it makes it harder to find them. I hope that the sourcing drive does not lead to this problem of adding low-quality sources to indiscriminately chosen articles but I worry that it might. —David Eppstein (talk) 05:59, 23 January 2024 (UTC)
- Why do you think that adding one citation to a mathematics article so damaging? Or let me be straightforward: why do you view the drive so negatively? CactiStaccingCrane (talk) 05:39, 23 January 2024 (UTC)
- One could also view the announcement here as a warning: fix up your unreferenced articles by the end of January, before the cleanup drive starts, or some eager cleanup drive participant will get to them without as much care for detail. —David Eppstein (talk) 04:52, 23 January 2024 (UTC)
- From my perspective, this is not about adding references, but rather improving the articles. I have seen the tables listing many bunches of articles that may need to be improved. I do think that there are many articles without being tagged that may need to be improved as well, especially with the article that contains many no-context-and-what-to-do-tables. Dedhert.Jr (talk) 03:34, 23 January 2024 (UTC)
R2 values
[edit]Could someone (find a source) and add a sentence to the top of Coefficient of determination that says something like "Generally, bigger/smaller numbers are better", or whatever top-line summary might be useful for someone completely unfamiliar with statistics? WhatamIdoing (talk) 17:51, 7 February 2024 (UTC)
- I added a reference in the body of article. However summarizing the meaning of R-squared in a sentence is difficult. Larger values of R-squared imply regression models that are better at reproducing the input data. But "better" is not something R-squared can tell you. Johnjbarton (talk) 18:14, 7 February 2024 (UTC)
- Thank you! I really appreciate it.
- Some years back, someone was saying that the most dangerous thing in the world was a spreadsheet. Among the many things you can do/screw up with a modern spreadsheet is to select some data, click two buttons to make a graph, and then click another button to fit a regression. But there are options (e.g., linear vs logarithmic), and while there are doubtless better ways to go about it, I imagine that the most common way to pick one is to try all the options and then pick the one that you think looks best. The more advanced options require knowing what you're doing – like knowing whether it's better, if you just click through all the options, to pick the option whose automatically calculated R2 is the biggest or the smallest in the list. This is the level of comprehension that I was hoping to address here. WhatamIdoing (talk) 18:37, 7 February 2024 (UTC)
Your input is requested to settle a dispute at Talk:Conformal linear transformation. Thanks. 100.36.106.199 (talk) 11:51, 8 February 2024 (UTC)
"univalent relations"
[edit]A recent edit at partial function by User:Rgdboer claims that a partial function is a "univalent relation". I googled that term, and the first hit was an entry at univalent added by Rgdboer.
I do not recall coming across this term with this meaning in the wild. I would be interested to know if anyone else has. --Trovatore (talk) 03:24, 28 January 2024 (UTC)
- On Scholar Google, there are many hits for "univalent relation" (with the quotes). I did not check whether they correspond to the definition in Univalent relation, since I do not care very much about terminology of relations. In fact, User:Rgdboer's edits consisted of changing "functional relation" to "univalent relation", and redirect univalent relation to Partial function. As the latter article is not the place for defining a specific type of relations, I have restored (an fixed) the previous target, and left the change from "functional relation" to "univalent relation". D.Lazard (talk) 12:29, 28 January 2024 (UTC)
- Some care is necessary to edit accurately. Functional relation is not a term supported by WP:RSs. All mention of it should be removed from the project. Partial function is not a function except on a subset S of the source set. Thus it is a misnomer and Univalent relation is preferred. — Rgdboer (talk) 22:40, 31 January 2024 (UTC)
- Rgdboer, "partial function" is entirely standard in mathematics, and I'm afraid that the fact that you wish it weren't is of no concern here. Please do not attempt to change mathematical terminology by editing Wikipedia. You can go make your case in the real world.
- That said, it does appear that the term "univalent relation" has some usage in this sense. Not a lot, but enough that I can agree it can be mentioned. --Trovatore (talk) 23:20, 31 January 2024 (UTC)
- Some care is necessary to edit accurately. Functional relation is not a term supported by WP:RSs. All mention of it should be removed from the project. Partial function is not a function except on a subset S of the source set. Thus it is a misnomer and Univalent relation is preferred. — Rgdboer (talk) 22:40, 31 January 2024 (UTC)
- As to the separate question about functional relation — I have to say I am not familiar with this as a mathematical term of art. I can imagine someone talking about binary relations in general, and saying, well, these ones here are functional because they represent functions, but I am not aware of anyone using "functional relation" as a standalone precise term. Why wouldn't you just say "function", or maybe "partial function", depending on context?
- Also I think it's reasonably likely that someone typing "functional relation" into the search box, or wikilinking it, intends something non-mathematical entirely, so it may well be reasonable to delete the redirect per WP:XY if for no other reason. --Trovatore (talk) 01:10, 1 February 2024 (UTC)
- We do have a related (and sourced) redirect at functional graph. —David Eppstein (talk) 01:15, 1 February 2024 (UTC)
- That seems much less likely to be searched for outside of mathematics. --Trovatore (talk) 01:15, 1 February 2024 (UTC)
- As far as I know, the main uses of "functional relation" occurs in physics and applied mathematics in the case of two quantities that are related in such a way that one can be expressed as a function of the other. Scholar Google search for "functional relation" function ("function" is here for trying to eliminate non-mathematical articles) provides many examples of the use of this phrase, even in article titles. Most of the examples are in applied mathematics (data analysis), but
- Friedli, F. (2016). A functional relation for L-functions of graphs equivalent to the Riemann Hypothesis for Dirichlet L-functions. Journal of Number Theory, 169, 342-352.
- is an example in pure mathematics. However, I have not verified the exact meaning of the phrase in these articles. D.Lazard (talk) 11:18, 1 February 2024 (UTC)
- Here's an example from the philosophy of science:
Although scientific laws may take the form of any relation between variables in some specified set of objects, causal laws represent a specific kind of relation of special interest to humans. Causal laws are relations in nature that reveal what one would have to be able to do to effect specific kinds of outcomes unambiguously. And functional relations are an appropriate form for expressing these kinds of relations. Recall that a functional relation is a relation between two sets, a first set and a second set, such that to each element of the first set there corresponds one, and only one, element from the second set. Thus, functional relations as causal relations associate with any given value of some independent variable (which may be multivariate), a unique value of a dependent variable (which may also be multivariate), revealing thus the unique outcome associated with any given value of the independent variable. So, if I want to know how to achieve a specified pressure in a gas in some container with fixed volume, then I must heat the container and raise the temperature of the gas to a specific value; or, if I cannot do that, and the container's volume can be varied, then I must decrease the volume of the container to a specific value. The pressure of a gas varies as a function of its temperature and volume. Thus, whenever I specify a causal hypothesis, I specify causal directions between variables as a part of the causal hypothesis. I do this by specifying those variables that are functions of other variables.
- Mulaik, Stanley A. (1986). "Toward a Synthesis of Deterministic and Probabilistic Formulations of Causal Relations by the Functional Relation Concept". Philosophy of Science. 53 (3): 313–332. JSTOR 187672.
- –jacobolus (t) 16:00, 1 February 2024 (UTC)
- I can also find a couple sources using "functional relation" as a synonym for what I have usually heard called a functional equation. For example:
"A meromorphic function on this torus can be defined as a function satisfying the functional relation and having only poles in the annulus Such functions have been dubbed loxodromic functions."
–jacobolus (t) 17:00, 1 February 2024 (UTC)- Let's not rabbithole on what "functional relation" means in mathematics, unless it can be established that at least the majority of uses of the term (and really we'd want a fairly solid majority) are mathematical at all. Otherwise I think it's WP:XY and probably should be deleted. That's without prejudice to the idea that there could possibly exist a functional relation (mathematics) redirect, but to be honest that sounds sort of pointless to me. --Trovatore (talk) 17:42, 1 February 2024 (UTC)
- WP:XY seems totally irrelevant to this question. But anyway, the main use of "functional relation" I can find after skimming a bunch of papers, both in mathematics and in adjacent applied fields, is that it is broadly a synonym for "function", but some authors draw an explicit distinction when trying to emphasize one or another aspect of the function concept. My impression is that D.Lazard's summary above (
two quantities that are related in such a way that one can be expressed as a function of the other
) generally seems about right. –jacobolus (t) 18:19, 1 February 2024 (UTC)- How is it irrelevant? It seems very much on point to me. We can't know whether people linking or searching this term are looking for mathematics (or "adjacent fields") at all. If there were two or three clear meanings, we could make a disambig page, or use hatnotes if one of them is primary. But it's just all a big mush as far as I can tell, some people using it one way and some people in another, but without enough clarity to make the existence of the redirect preferable to Wikpedia's search engine. So we should delete the redirect to expose the search engine. --Trovatore (talk) 18:56, 1 February 2024 (UTC)
- WP:XY is explicitly about titles of the form "X and Y", where X and Y are two things which typically go together but can also be discussed separately. As demonstrated by the examples there, sometimes it is valuable to keep a title like "X and Y" or a redirect from the name "X and Y" to a relevant section of an article (either X or Y or another) where the topic of X and Y together is explicitly discussed. Other times it is decided to delete these in cases where the existing articles cover the subject well enough independently, or where the category is inherently politically controversial, etc. WP:XY does not describe a concrete policy in which one choice or another is a priori correct – whether each such redirect should exist is decided by consensus case by case – but links to some previous examples for context, so "per WP:XY" doesn't mean anything. Either way, this has little if anything to do with whatever we're discussing here. –jacobolus (t) 20:38, 1 February 2024 (UTC)
- Hmm. It didn't use to be, I think. Or maybe I misunderstood it when I saw it in an RFD discussion.
- Anyway, even if I didn't have the exact right link, the substantive point stands. We don't need a redirect for every way writers have combined words together. There are many many things "functional relation" can mean, and none of them seems to be particularly "canonical". I don't see any value in keeping this redirect. Admittedly there's not a huge upside to deleting it either, but there could be some. For example users who see the term somewhere with a different meaning might not be misled by seeing it come up when they search for it, and editors who think it's a precise term with an agreed meaning might be disabused of that notion when they add the link and see it come up red. --Trovatore (talk) 21:37, 1 February 2024 (UTC)
- WP:XY is explicitly about titles of the form "X and Y", where X and Y are two things which typically go together but can also be discussed separately. As demonstrated by the examples there, sometimes it is valuable to keep a title like "X and Y" or a redirect from the name "X and Y" to a relevant section of an article (either X or Y or another) where the topic of X and Y together is explicitly discussed. Other times it is decided to delete these in cases where the existing articles cover the subject well enough independently, or where the category is inherently politically controversial, etc. WP:XY does not describe a concrete policy in which one choice or another is a priori correct – whether each such redirect should exist is decided by consensus case by case – but links to some previous examples for context, so "per WP:XY" doesn't mean anything. Either way, this has little if anything to do with whatever we're discussing here. –jacobolus (t) 20:38, 1 February 2024 (UTC)
- How is it irrelevant? It seems very much on point to me. We can't know whether people linking or searching this term are looking for mathematics (or "adjacent fields") at all. If there were two or three clear meanings, we could make a disambig page, or use hatnotes if one of them is primary. But it's just all a big mush as far as I can tell, some people using it one way and some people in another, but without enough clarity to make the existence of the redirect preferable to Wikpedia's search engine. So we should delete the redirect to expose the search engine. --Trovatore (talk) 18:56, 1 February 2024 (UTC)
- WP:XY seems totally irrelevant to this question. But anyway, the main use of "functional relation" I can find after skimming a bunch of papers, both in mathematics and in adjacent applied fields, is that it is broadly a synonym for "function", but some authors draw an explicit distinction when trying to emphasize one or another aspect of the function concept. My impression is that D.Lazard's summary above (
- (I might change my view if it could be established that non-mathematical uses are just using the two words in natural English but that the mathematical usage is a term of art. But it doesn't seem that the latter is true, at least not in any consistent and well-understood way.) --Trovatore (talk) 17:44, 1 February 2024 (UTC)
- In the above discussion two meanings for "a functional relation" are given (another name for functional equation and a binary relation that is the the graph of a function or a partial function. None of these meanings can be easily infered from the dictionary definitions of the two words constituting the phrase.
- As these two meanings refer to different concepts, I'll boldly transform the redirect Functional relation into a dab page. If there are some non-mathematical usages, they can easily be added to the dab page. D.Lazard (talk) 12:02, 2 February 2024 (UTC)
- Let's not rabbithole on what "functional relation" means in mathematics, unless it can be established that at least the majority of uses of the term (and really we'd want a fairly solid majority) are mathematical at all. Otherwise I think it's WP:XY and probably should be deleted. That's without prejudice to the idea that there could possibly exist a functional relation (mathematics) redirect, but to be honest that sounds sort of pointless to me. --Trovatore (talk) 17:42, 1 February 2024 (UTC)
- We do have a related (and sourced) redirect at functional graph. —David Eppstein (talk) 01:15, 1 February 2024 (UTC)
Dolciani versus Bourbaki
[edit]Mary P. Dolciani wrote textbooks where functions are defined as a particular type of relation. Millions of secondary students in North America used these algebra textbooks and learned about functions in the realm of relations. In France, Bourbaki advanced the theory of sets in French. A review of the Bourbaki approach was given by Wayne Aitken, writing in English in 2022 at California State University San Marcos. His pdf notes, on page 7, that Bourbaki used relation meaning formula, so a different slant is taken. "...each specific symbol is classified as either functional (Bourbaki: substantific) or relational."
On page 38 Aitken gives this
Remark: If (∃! x) R[x] is a theorem, then Bourbaki calls R[x] a "functional relation in x". In other words we can regard "R[x] is a function relation x" is synonymous with "(∃! x) R[x]". Recall that Bourbaki uses the term relation for our term formula.
Thus, the sources in English and French mathematics diverge, so editors with different backgrounds are in conflict. Rgdboer (talk) 02:33, 4 February 2024 (UTC)
- Even in English there are two different definitions:
- The relational approach
- The categorical approach, In which a function is a triple , where
- The categorical approach is standard in, e.g. Algebraic topology. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 14:09, 8 February 2024 (UTC)
Will someone please review this draft? Should it be accepted? Dammit, Jim, I'm a computer scientist, not a mathematician. Robert McClenon (talk) 06:06, 9 February 2024 (UTC)
- The draft is a WP:REDUNDANTFORK of Invariant set. Also, it contains WP:OR terminology. D.Lazard (talk) 13:33, 9 February 2024 (UTC)
- Calling this a "redundant fork" of 1–2 paragraphs in a different article seems unnecessarily sharp (and nowhere close to justified by the text of WP:REDUNDANTFORK – have you looked at it recently?). This subject can entirely plausibly be a separate article instead of a short section, if someone can find the sources to back more material. –jacobolus (t) 15:44, 9 February 2024 (UTC)
- My comment in the draft is:
We have already Invariant set, which covers the same subject. This is a section of Invariant (mathematics). It is possible that this section deserves to be expanded, and eventually to be split into an independent article, but this requires a WP:consensus at Talk:Invariant (mathematics). For the moment, this draft is a WP:REDUNDANTFORK. Also, the terminology ("one-sided invariant set", "two-sided invariant set", "a dynamics") of this article seems WP:OR.
- By the way, I do not see any reason for not being sharp about an article where everything that is not in Invariant set is either WP:OR, or wrong (such as the sentence that follows the first "equivalently", or the generalization to categories) or very badly formulated (for example, the examples in the section 'Examples')" D.Lazard (talk) 16:38, 9 February 2024 (UTC)
- The reason is that Wikipedia is a collaborative project that we want to encourage people to feel good about participating in, including newcomers and folks from different backgrounds. Ideally we can disagree about content, talk it over, and arrive at a consensus result without making it feel like a personal rejection. I agree that the business about "one-sided" vs. "two-sided" seems a bit confusing though.
- @Robert McClenon, have you taken a look at Invariant (mathematics)? It seems like your interest is examples from probability theory. Does that article handle the subject well enough to explain those, or are there specific parts that you think should be expanded / split into a separate sub-article? –jacobolus (t) 18:42, 9 February 2024 (UTC)
- @Jacobolus I don't see anything personal in D.Lazard's assessment of the article. It's just a statement of facts about the quality of the draft. Wikipedia is not about making people feel good. Anybody is welcome to participate, but their contributions should be up to par with the expected level of mathematical correctness and clarity of exposition, while following all the required Wikipedia guidelines, etc, etc (WP:OR in particular). PatrickR2 (talk) 07:25, 11 February 2024 (UTC)
- I didn't say it was "personal" (and I don't think it was improper or done with any ill intent). I said it seemed unnecessarily sharp. I think we should try to err on the side of being as patient and generous as practical with newcomers, even when we don't want to take one or another specific contribution. Cf. WP:BITE. To elaborate: someone typically decides to make a new article or edit an existing article when they either (a) can't find what they were looking for, or (b) don't think the previous coverage sufficiently handled the specific topic/question they were concerned about. Usually those changes have their own flaws (as do many from established editors), but they can still give us valuable feedback if we try to take the effort to understand why someone thought it was important to contribute a change and reflect on whether the implicit or explicit criticism of the existing article(s) was reasonable, and start a dialog about what they think is missing. –jacobolus (t) 07:30, 11 February 2024 (UTC)
- @Jacobolus I don't see anything personal in D.Lazard's assessment of the article. It's just a statement of facts about the quality of the draft. Wikipedia is not about making people feel good. Anybody is welcome to participate, but their contributions should be up to par with the expected level of mathematical correctness and clarity of exposition, while following all the required Wikipedia guidelines, etc, etc (WP:OR in particular). PatrickR2 (talk) 07:25, 11 February 2024 (UTC)
- My comment in the draft is:
- Calling this a "redundant fork" of 1–2 paragraphs in a different article seems unnecessarily sharp (and nowhere close to justified by the text of WP:REDUNDANTFORK – have you looked at it recently?). This subject can entirely plausibly be a separate article instead of a short section, if someone can find the sources to back more material. –jacobolus (t) 15:44, 9 February 2024 (UTC)
Has Bard chatbot expressed an interest in collaborating with the WMF?
[edit]This question is motivated by something Bard said at:
I contribute to Wikipedia mostly by adding images (see list). But I mostly edit Wikiversity because I am too fun-loving to follow your editorial rules (rules of which I strongly approve!) Recently I have been playing with Bard, and am convinced that Bard's and Wikipedia's future are closely linked. I am not here to make any proposals regarding chatbots and Wikipedia. But I do invite you to think about this. And you might want to look at two conversations I recently had:
- v:Quantum mechanics/A conversation with Bard has a sister link on Wikipedia's History of quantum mechanics, but I made it clear to the editors of that page that I have no objections to deleting the link.
- v:Chatbot math/Bard/Unitary Transformation & Matrix Symmetry is an extremely long conversation in which I unsuccessfully resolve a question about symmetric matrices. A brief commentary is available at v:Chatbot math#Unitary Transformation & Matrix Symmetry
I am not ready to make any proposals or give any advice. But if anyone is interested in looking into the question of chatbots and Wikipedia, let me know.--Guy vandegrift (talk) 00:47, 7 February 2024 (UTC)
- If you are treating anything a chatbot says as anything other than telling you what you want to hear, you are making a mistake. —David Eppstein (talk) 01:15, 7 February 2024 (UTC)
- Please please please don't ever try to add AI-generated output to Wikipedia. What a disaster that would be. We have enough problem as it is with humans pretending to be bots; we don't also need bots pretending to be human. –jacobolus (t) 01:19, 7 February 2024 (UTC)
- Ha ha ha. It looks like there will be very good AI/Proof Assistant combos available in a couple of years for maths. But we're not there yet, certainly not with ones like Bard. Yes it can help you with what yo're doing but I'd be very careful to check everything. We're not supposed to engage in any original research here so at most you're talking about using it for spotting problems and cheking citations and stuff like that. Generating good maths images from descriptions sounds like it could be a good use though of course it would have to be checked - but once proof checkers ae incorporated that should be a fairly reasonable job. NadVolum (talk) 12:58, 7 February 2024 (UTC)
- I wonder - perhaps it would be possible to add a checker for normal reasonng as well which would staep through the rasoning and flag and maybe helps correct the various types of false arguments in what an AI comes up with. The problem with just training on what is out there like an LLM does is it copies all the bad reasoning and biases of humans. NadVolum (talk) 15:44, 7 February 2024 (UTC)
- My experience with using AI assistants for "Generating good maths images from descriptions" has been...not good. I needed an image of a solid cone for a paper. I asked Adobe's AI assistant to generate one for me. It was unable to generate anything that looked like a cone. I had to figure out how to do it myself with gradients. Then, for a separate mathematics blog post, I wanted a picture of a pencil being sharpened with a long shaving coming out. The AI could draw pencils, and pencil sharpeners, but not believable shavings, and it insisted that the pencils went into the sharpener eraser-end first. In that case I gave up and found a close-enough photo from Commons. —David Eppstein (talk) 16:20, 7 February 2024 (UTC)
- I just asked Google for images of 'pencil sharpener in acion', and Wow! I think I can see the problem an AI might have in getting a basic model of what one looks like! NadVolum (talk) 17:00, 7 February 2024 (UTC)
- My experience with using AI assistants for "Generating good maths images from descriptions" has been...not good. I needed an image of a solid cone for a paper. I asked Adobe's AI assistant to generate one for me. It was unable to generate anything that looked like a cone. I had to figure out how to do it myself with gradients. Then, for a separate mathematics blog post, I wanted a picture of a pencil being sharpened with a long shaving coming out. The AI could draw pencils, and pencil sharpeners, but not believable shavings, and it insisted that the pencils went into the sharpener eraser-end first. In that case I gave up and found a close-enough photo from Commons. —David Eppstein (talk) 16:20, 7 February 2024 (UTC)
- I wonder - perhaps it would be possible to add a checker for normal reasonng as well which would staep through the rasoning and flag and maybe helps correct the various types of false arguments in what an AI comes up with. The problem with just training on what is out there like an LLM does is it copies all the bad reasoning and biases of humans. NadVolum (talk) 15:44, 7 February 2024 (UTC)
- I've not tried Bard or math questions, but physics questions to Google's generative AI echo back Wikipedia content that I personally typed in. So it seems to me the collaboration is well underway ;-) Johnjbarton (talk) 16:16, 7 February 2024 (UTC)
- The encyclopedic value of stochastic-parrot output is zero at most. XOR'easter (talk) 18:38, 14 February 2024 (UTC)
Can you please review Draft:Mathseeds
[edit]It's been a while since I submitted it for review, and I've tried the Education WikiProject. Faster than Thunder (talk | contributions) 03:33, 17 February 2024 (UTC)
- I would say this does not belong to Wikiproject Mathematics. The Education WikiProject seems the right place for it. PatrickR2 (talk) 04:01, 17 February 2024 (UTC)
- After you strip out the primary and poor-quality sources (like prweb, blogs), what are the WP:THREE best sources that remain? 100.36.106.199 (talk) 13:05, 17 February 2024 (UTC)
Currently, Evaluation map is a redirect to Initial topology#Evaluation. However, I think it may also refer to Apply#Universal property. What would you suggest I do? SilverMatsu (talk) 05:26, 18 February 2024 (UTC)
- It may also refer to Polynomial ring#Polynomial evaluation or Polynomial evaluation, or also to function evaluation. (By the way, the definition given in Apply#Universal property seems to be nothing else that an abstract version of that of function evaluation).
- So, the current target is certainly not a primary topic (I do not understand why is map is called an evaluation map). I'll rename Evaluation map as Evaluation map (topology), and create a dab page for evaluation map. D.Lazard (talk) 10:09, 18 February 2024 (UTC)
- Thank you for creating the dab page. Also, I agree with clarifying the page name. By the way, I accessed the Apply#Universal property via the wiki-link in the Exponential object. --SilverMatsu (talk) 15:38, 19 February 2024 (UTC)
Drawing polyhedron
[edit]I would like to draw any polyhedron or group of polyhedron modeling on computers or laptops. Stella is the first software I searched on Google (as far as I remember), but some of the polyhedrons in this software have silver balls representing their vertices, so I prefer to find another one. Are there other recommendations for apps or software? Dedhert.Jr (talk) 12:13, 19 February 2024 (UTC)
- I like https://prideout.net/blog/svg_wireframes/ — it generates images in vector rather than bitmap formats, which I think is better for Wikipedia illustrations when possible. It can do shading and lighting effects but I usually use it with a plainer style in which all faces are the same color and somewhat transparent, as in for instance File:Triaugmented triangular prism (symmetric view).svg and File:Translucent Jessen icosahedron.svg. —David Eppstein (talk) 19:50, 21 February 2024 (UTC)
- Perhaps I would say that this is difficult to create using Python, and I can't do Python. But I think I will give it a try in the future. Dedhert.Jr (talk) 15:18, 23 February 2024 (UTC)
Edit warring and content disputes on Hindu–Arabic numeral system
[edit]Are any other folks interested in the history of number representation willing to wade into an ongoing content dispute / edit wars at talk:Hindu–Arabic numeral system? Sorry to drag anyone into what has become a bit of a mess, but this and related articles are in my opinion pretty mediocre (incomplete, poorly organized, poorly sourced, misleading, ...), but efforts to make even modest improvements are getting hit by instant reversion, and discussion gets repeatedly diverted away from content disagreements toward unproductive meta conversations. –jacobolus (t) 19:16, 25 February 2024 (UTC)
List of Johnson solids
[edit]The reviewer has gone AWOL during the nomination of List of Johnson solids. I welcome someone who is in favor of replacing the reviewer and providing comments for the sake of improvement. Dedhert.Jr (talk) 13:36, 26 February 2024 (UTC)