Wikipedia:Articles for deletion/Wilkinson's polynomial

From Wikipedia, the free encyclopedia

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was keep. Mailer Diablo 08:32, 30 July 2006 (UTC)

[edit] Wilkinson's polynomial

Article appears to be nonsense. There actually is something called a "Wilkinson's polynomial", which can be googled, but this article appears to describe something else; it appears to be original research of some sort. Although an article describing the real, actual "Wilkinson's polynomial" can be and should be written, it seems easier to do that by starting from a clean slate -- there appears to be nothing salaveageble from this article. linas 03:58, 25 July 2006 (UTC)

  • Keep. The definition is correct. See for instance Hong Zhang, Numerical condition of polynomials in different forms, Electronic Transactions of Numerical Analysis, 12:66, PDF, or the original in Wilkinson, Rounding Errors in Algebraic Processes, published somewhere in the 60s. -- Jitse Niesen (talk) 04:11, 25 July 2006 (UTC)
  • Delete unless immediately cleaned up. I showed it to a mathematician friend and he says the article is partially correct, but mostly nonsense. Without any valid sources or exposition in layman's terms, there's no way to prove this isn't a hoax. A Google search turned up very little. Crabapplecove 04:18, 25 July 2006 (UTC)
    • There is no such option. We're trying to determine whether a topic is worthy of inclusion in Wikipedia, not whether an article is any good in its present form. The topic is real, definitely not a hoax. So all that's left is deciding whether it is of sufficient interest for an article. --MarkSweep (call me collect) 05:04, 25 July 2006 (UTC)
      • What do you mean, "no such option"? This is a discussion, not a vote; we're debating what should be done with the article, not selecting from a list of predefined outcomes. It's perfectly legitimate to make a recommendation conditional. It's also perfectly legitimate to demand that a particularly poor article on an encyclopedic topic be deleted, since this hardly prevents a good one being written in future; it's better to have missing articles than misleading articles. — Haeleth Talk 11:13, 25 July 2006 (UTC)
        • Read the deletion policy. Deleting an article just because it's partly inaccurate is not an option. Holding an article hostage by saying "delete unless cleaned up" is not in accordance with deletion policy and such opinions are usually discounted by the closing admin. --MarkSweep (call me collect) 17:26, 25 July 2006 (UTC)
          • Linas is perfectly justified in proposing to start over if necessary. Fortunately, the present text, as cleaned up, seems to be satisfactory. Septentrionalis 20:13, 26 July 2006 (UTC)
  • Keep and rewrite/expand. Wilkinson's polynomial and the numerical issues it illustrates are quite real. This case should have been referred to fact and reference checking instead of AfD. I'll add a few references now. --MarkSweep (call me collect) 05:04, 25 July 2006 (UTC)
  • Keep per Jitse. He is a numerical analyst. It is described as Wikinson's Pathologic polynomial in "Numerical Methods that work" by F. S. Acton, Harper 1970. However the second equation is not correct and not a Lagrange form. It needs fixing by someone more expert than I. --Bduke 08:49, 25 July 2006 (UTC)
Sorry but the second equation is indeed a lagrange form: just use the roots as "interpolation points" plus any other point and you got a "special" lagrange form of the polynomial; the point here is that all coefficients but one equal zero: this could explain why you failed to see the match with a lagrange form... Julien Tuerlinckx 17:02, 25 July 2006 (UTC)
Well, it is now, but it was'nt when I wrote the above. It looked like unfinished work starting from a copy of the first eqn. --Bduke 00:10, 26 July 2006 (UTC)
  • Keep for now, but if the above concerns are accurate and the article is incorrect, it must be fixed promptly. It's far too common for people to recommend keeping an article on the assumption that someone else will fix it, only to find that a year later it's barely been touched. I suggest that this should be given an {{expert}} tag, with a view to bringing it back to AfD in a month or two if no expert has deigned to check the article and correct any inaccuracies there may be. — Haeleth Talk 11:13, 25 July 2006 (UTC)
  • Keep: Looks interesting. Perhaps tag the incorrect part for correction. Stephen B Streater 16:49, 25 July 2006 (UTC)
  • Keep: I fail to see the inaccurate or incorrect part (maybe the numbers added recently can be tagged as inaccurate since it does not use the ad hoc notation (we only have 5 decimals...)). I feel more like a lot more could be said on this topic, so the expert tag seems a good idea. Julien Tuerlinckx 17:02, 25 July 2006 (UTC)
  • Strong keep. The topic is important in numerical analysis. I have given the article some of the attention it seemed to need, perhaps enough to sway the few delete votes. --KSmrqT 17:21, 25 July 2006 (UTC)
  • Keep It seems important, and I fully trust the other professional mathematicians here. —Mets501 (talk) 09:47, 26 July 2006 (UTC)
  • Keep, maybe speedy keep as I am the one who opened the AfD. My apologies. The article is now a fine article. linas 00:35, 27 July 2006 (UTC)
  • Speedy Keep, modified article looks fine now, all seem to agree. --Salix alba (talk) 08:15, 27 July 2006 (UTC)
  • speedy keep this please the article is now modified thank you linas
  • Strong keep This is my first time seeing the article, and it is well-done. Acyso 06:44, 30 July 2006 (UTC)
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.