Talk:Least-squares spectral analysis

Move proposal

I propose that Vaníček analysis be moved to Least-squares spectral analysis, a more descriptive, neutral, and widely used and known term for the topic. The article has a lot of good well-sourced information on a notable topic, but is written in a biased way aimed apparently at making sure that Vaníček is the main name associated with the method, and that the contributions of others are denied or minimized. It written by a WP:SPA who has not denied that he is Vaníček's ex-student who wrote the thesis that's used as a source to verify such POV statements as that the "method has been mistakenly called by some the Lomb-Scargle or simply the Lomb method." Dicklyon 16:11, 7 October 2007 (UTC)

I am rather surprised by these allegations. First of, you first call me an expert and then YOU arbitrarily pick (out of at least five names for the same method) the one that YOU deem to be "a more descriptive, neutral, and widely used and known term". Here is a perfectly logical explanation for my way of presenting the method's name by using just two references (while stating that the name you selected has many references but without stating those specifically -- this doens't mean that the name you picked has more references than the name as I stated it -- don't mix apples and oranges!): the reason why I included only the first known (from 1970-ies) and the last known (from 2006) reference that cited the method as Vanicek's, was that I hoped this would have been sufficient for an encyclopedia type of article. If you insist, and if you are some kind of authority as you seem to be, I can provide a reference that calls the method "Vanicek's" for every single year between 1970 and 2007! But why flood the article like that? Second, why are you assuming that not admitting to be some person means you are actually that person? This makes no sense to me. Last -- before accusing me of making up the word "mistakenly", please do read the reference first, before passing accusations like these. I am not here to be insulted like that.

I appreciate the comment you made on the method authors' biography (I just included the biography source in there, so you should check it). But it seems to me like your contributions to the method article are made not in good faith (not to mention you are not an expert, as you indicated yourself), but in order to say just about anything that comes across your mind. It doesn't seem to me like improvements can be made like that. Please use your powers sparingly. I am here to improve the Wikipedia in the aspect of geodesy, not to waste time and energy fighting you. You should also read the link on biographies that you stated to me: see in there what the Wikipedia inventor says about the Wikipedia rules that obstruct improvements to Wikipedia: he says -- IGNORE THEM. --Geoeg 20:18, 7 October 2007 (UTC)

I have no power, just trying to help wikipedia be the best it can be, and to learn new stuff along the way; I assume that you would want the attention to your article that you brought by linking it from a whole bunch of signal processing articles. You, on the other hand, seem to be a WP:SPA, and you'll written an article that's very slanted. I'm proposing we fix that; I don't see why you would question my good faith here; check my contribs history.

But let's talk about the article, instead of making it personal. Don't you think the widely-used generic name would be more appropriate? Here are some book refs you can check to see how frequently it is called what; compare these and these and these. Dicklyon 00:17, 8 October 2007 (UTC)

Indeed let's. By the way, the links you provided list mostly articles, not books. Now to the point: when you don't use dash, the number of returns increases hundredfold or more (I counted in hundreds). So one gets this and this, returning the number of papers that properly credit Vanicek for his method.

So now it turns out that not only you pretend to be an expert on something you are not, but you also use biased Google searches to make a point. I have no idea what your real intentions here are, but by now you do seem like you are taking this personal because I caught you making those mistakes. Why don't you just cool daown, and stay away from the whole thing? Leave it to someone more skilled, I'm sure there are skilled editors out there who can actually add to the discussion. --Geoeg 02:42, 8 October 2007 (UTC)

The searches were to illustrate what the method is called. I would certainly not suggest that Vanicek should not be given full credit for his contributions. But you've written the article in a way that specifically denies credit to the other contributors. If you review those search hits you found, do many of them call it Vanicek analysis? or Vanicek method? or Vanicek spectral analysis? No, they don't; and that's what I was illustrating by the searches I showed you. So stop confusing the history and credit with the naming; we should represent the method by what it's most often called, and list alternative names as alternative; including the Lomb and Scargle and whoever else it may be called after. And I have not pretended to be an expert on any of this. Dicklyon 02:53, 8 October 2007 (UTC)

"Illustrate" is a convenient word with numerous meanings... But back to the point: is that why you now bolded those other names used for the same thing? Strange reasoning, because in that case you also forgot to delete the word "mistakenly" for Lomb and Scargle (as you seem to be there advocate?). Of course, you'd be in conflict with regulations then, since that term was properly cited... --Geoeg 03:20, 8 October 2007 (UTC)

I have no idea what you're referring to. Dicklyon 03:27, 8 October 2007 (UTC)

You sure you don't? Here it is: a paper may refer to Vanicek in a context, it doesn't have to name the method precisely either "Vanicek method" or "Vanicek analysis". Here are some of the infinitely many possibilities: "method by Vanicek", "as Vanicek showed", "as it was demonstrated by Vanicek", "analysis technique by P. Vanicek", "spectrum was computed after Vanicek", "data were assessed using Vanicek's", "Vanicek has developed", "Vanicek derived... You get the point? --Geoeg 08:53, 8 October 2007 (UTC)

The question is not whether the name exists. I think we can all agree that both Vanicek analysis and least squares spectral analysis are used in the literature. The question is, which is used more often? A Google hit count has its drawbacks but it is one way of establishing popularity of a term. --Zvika 10:57, 8 October 2007 (UTC)

No, that's not the question, but a question for there are various criteria according to which relevance can be assessed. Who says that simply counting of the exact phrase is the criterion of all criteria? For instance, I could pose following "weighted" criteria, and easily claim that they are superior to just simply counting the phrases (quality before quantity!): (1) Only references that most closely approach the uniform distribution (say, per year-window) over 38 years since the method was invented, or, if you insist on counting why not try this (more fair to the author): (2) Only references that credit Vanicek, regardless of the order of the words they use to credit him, or say (3) Only references that mention the method during mid five years since the method was invented (sort of moving average). To me, Vanicek analysis is the common and simplest denominator for all the possible versions. Besides being a nonjudgmental choice (people have already called it after the inventor, so we're not rediscovering America here are we), it is also linguistically justified and simple -- compare with Fourier analysis. --Geoeg 14:55, 8 October 2007 (UTC)

Yes, and it would be prefectly appropriate for you to call it Vanicek analysis in all your own publications. But in wikipedia, our job is not to decide what it ought to be called, but rather to report on it in terms of what it is called. And applying Vanicek's name as part of the name of the method is actually rather rare, though he is almost always given credit for originating it. Your own POV, however, jumps out at us in your writings, when instead of following the historical credit given in refs such as Press, you rewrite the history slanted to Vanicek as "Although these latter authors claim no new method (for instance, abstract in [9]), the Vaníček method has been mistakenly called by some the Lomb-Scargle or simply the Lomb method, as well." This is what we need to work on neutralizing, and it's clear that you'll need help from other editors to get around your Vanicek-centric view. I think it's fair to presume that you are an ex-student of his, most likely the author of the cited thesis, since you didn't bother to deny that; so why not fess up and help us make progess? Dicklyon 14:59, 8 October 2007 (UTC)

I don't have any such publications. Also, I am not trying to decide on anything here; did you read my previous post at all? In a nutshell: Vanicek method has already been called that, and your potatoes-counting assessment is, if not worse, then at least as good as any other assessment is. As I already told you, if you have a problem with the word "mistakenly" then remove it... if you dare judge a scientific reference, that is. But as far as other semantics is concerned, be my guest. The only one who is jumping at us here is you, with your "identify yourself... identify yourself" (the more you sound like a broken record the more fun it is to ignore your illegitimate requests). But since this isn't funny any more, here it is: I deny it. Of course, the reason why I "didn't bother to deny that" is because I was not supposed to. I'll take "Vanicek-centric" as a compliment. --Geoeg 16:04, 8 October 2007 (UTC)

What is it?

Without interfering with your fascinating mutual bashing above, I would like to ask if one of you could add a section to the article here, to explain what it actually is. I mean a technical explanation (equations and all) of what you do when you do "Vanicek analysis" (or whatever you end up deciding it should be called). Right now the article describes historical background, some "properties" (which are more like an advertisement), and applications, but doesn't say what I have to do to perform Vanicek analysis. I think this will also make it easier to reach consensus on what is the most common name in the scientific community for this method. --Zvika 06:13, 8 October 2007 (UTC)

I haven't studied it enough to know the details yet, but this book gives a pretty good description. Basically, choose sinusoid coefficients to minimize the sum of squared errors at the data points, for every frequency of interest. What's not clear to me is what should be the frequencies of interest, and what to do about the fact that the sinusoids may not be orthogonal when considered just at the data points. There are standard techniques using big-matrix algebra for fitting sums of non-orthogonal functions to samples, so once you decide what the frequencies of interest are, it should be easy enough; it doesn't have the regular structure that makes the FFT possible, which is why it's computationally more expensive. I'm sure it's all explained in there somewhere, and I'll let you know if/when I figure it out. Dicklyon 06:48, 8 October 2007 (UTC)

Oh, and did I really sound like I was bashing? Sorry. I tried to gently determine what was behind his single-purpose-account behavior, but he didn't cooperate; I tried to keep it civil and on topic, and made constructive comments on how to improve the article. Should I back off? Dicklyon 06:50, 8 October 2007 (UTC)

Oh, you (again) don't read Wikipedia regulations, otherwise you would have noticed that there is nothing wrong with "single-purpose-account behavior", even if that was the case here (which it isn't, see my other contributions on geodesy). So you just proved (again) that yes -- you are bashing. Your comments were anything but civil, starting with your inquisition-style interrogation as to my identity. As soon as I refused to "show you my ID Mr. Officer", you started your bashing thing. --Geoeg 08:47, 8 October 2007 (UTC)

Zvika: more explanations including the technical ones are coming as soon as I have time, hopefully this week. I had no idea that people could get so impatient around here -- as if a Wikipedia printed edition is scheduled and deadlines have to be met. By the way, I don't see how technical details such as formulae could solve the naming dispute; the method has been called Vanicek's continuously ever since it was invented almost forty years ago. Finally, Dicklyon has started bashing first, unsatisfied with my identity being hidden from him (and they say Wikipedia can be edited by everyone), so I'm sorry if it sounded like I'm bashing too. --Geoeg 08:47, 8 October 2007 (UTC)

I am not being impatient, merely pointing out that this is a 1000-word article, yet I only began understanding what it is about after Dick's one-sentence explanation above. The technical details are necessary because, without them, I don't know what the method is, and so I can't say whether or not I know it by another name; all I can say is, I've never heard of Vanicek analysis (and neither have any of the four textbooks I've looked at so far).

As for bashing, Geoeg, I suggest you try to calm down. Dick has been an editor in WP for a long time and has contributed significantly to a large number of articles. That does not give him "powers" as you insinuate, but it does make your accusations of bad faith seem far-fetched. While he might not have always been civil (I don't think the vandalism templates were appropriate), he really is trying to help out. Please try to see his side of the argument as well. --Zvika 09:56, 8 October 2007 (UTC)

And I never questioned Geoeg's right to anonymity. I politely asked if he would reveal more about himself and/or his relationship to the subject because if he would then it would make it much easier to address the questions of WP:COI and WP:POV. I pointed out to him that as an anon WP:SPA with apparent close connection to the subject, his work would necessarily be subject to close scrutiny. That's the way it works. Dicklyon 13:36, 8 October 2007 (UTC)

As I said at the beginning, I have no problem with scrutiny. I just couldn't imagine that anyone would be enforcing some regulations so rigidly, but at the same time arbitrarily ignoring other relevant regulations, as well. Not to mention the use of term "vandal", and other blunders. But I guess we are all humans... --Geoeg 14:55, 8 October 2007 (UTC)

Indeed we are. I felt the vandalism warning was appropriate after the removed of the tag again after I explained that that was inappropriate. If there's a more effective way that I should have used to get your attention to that problem, please let me know. And I'm not sure what you mean by enforcing regulations; I have no powers of enforcement, and wikipedia mostly doesn't have regulations; but I do try to help follow guidelines, and I'm usually pretty flexible. But when new editors get stubborn, so do I. Dicklyon 15:02, 8 October 2007 (UTC)

Just to remind you, and for the record, it was you who started it -- with your illegitimate "ID yourself" requests. That's the source of your taking it personal (see my denial above). About the method name: I propose that it remains as is, because it is (A) linguistically acceptable, (B) scientifically correct, (C) properly referenced (regardless of a criteria choice for assessing those references' weights), (D) not precedent-setting (see Fourier analysis). As to the Vanicek page, I propose either (A) the removal of the tag along with the removal of the specific parts of his biography you have problem with, or (B) leaving the tag along with "citation-needed" tags on each part of his biography you have problem with. --Geoeg 16:04, 8 October 2007 (UTC)

When I asked on your talk page "Were you his grad student?", was that illegitimate? I thought it was a reasonable way to put the issue up front for discussion. Dicklyon 22:28, 8 October 2007 (UTC)

Yes, it was. And not only that: while attempting to invade my privacy, you put a pressure on me via misuse of your Wikipedia privileges/access. I am not sure if Wikipedia has a regulation on this, but I am quite certain that the US Criminal Code does. --Geoeg 00:20, 9 October 2007 (UTC)

You may want to review WP:THREAT, because if you do that again you'll be out of here. Dicklyon 03:57, 9 October 2007 (UTC)

Are you threatening me? --Geoeg 16:33, 9 October 2007 (UTC)

Yes, sort of. I'm trying to make it clear that if you do that again, I will report you and ask for you to be banned, as I should probably have done the first time. Dicklyon 21:45, 9 October 2007 (UTC)

I'm sorry, do what again? I was under the impression that you were pressuring me to reveal my identity to you, and that you were threatening me with legal action (sort of). Where did I threaten anyone with anything?????????????? --Geoeg 23:10, 9 October 2007 (UTC)

Images

Unrelated to the above: I would appreciate if anyone could please advise me as to the appropriate source of freeware or otherwise acceptable images? I would like to place a more attractive image at the top of geodesy page. The way it is right now, I think that many people who open the page get easily repelled (by that obsolete photo) rather than attracted to the subject. (And I would hope Wikipedia is in business of spreading free knowledge). Unfortunately, scanning through internet pages is reality of the fast-paced world we live in. --Geoeg 16:04, 8 October 2007 (UTC)

Actually, I have just found that Wikipedia has a compiled list of such links. --Geoeg 18:22, 11 October 2007 (UTC)

Opinion of 3O

As a simple 3O browser I'd like to weigh in. I think a cool-down period is in order here. Let's try and back off and get some perspective. This is a Wikipedia article, and as impressive as WP is it's not worth adding all this aggravation to your lives. Threats were made and words were tossed... let it go. I'm going to go do some research, let's all look for a way to step back and take a breath. Padillah 12:32, 10 October 2007 (UTC)

While you do that, it seems like minimum good will to remove at least some of the naughty repellant-tags on Petr Vaníček and Vaníček analysis (because Dickhawk has attacked both). Right now, the two pages look like a Christmas tree...in July. In the mean-time, here is some food for thoughts: please check out J. Tuzo Wilson Medal and ask yourself how many recipients of the highest geophysics recognition in the second largest nation on Earth (therefore: a geophysics world power) actually do have their biographies on Wikipedia? Answer: a mere 30%! If various hawks are allowed to continue massacring every new biography of a scientist just because it was not featured on MSNBC, then I am afraid there is no point to further discussions here. In that case, all Wikipedia pages about geophysics and geodesy would cease being serious. Speculations on whether they too are based on a popular vote would start creeping in. --Geoeg 18:48, 10 October 2007 (UTC)

Ok. Several things: personal attacks, such as calling someone 'dickhawk', have no place on wikipedia. Neither do thinly veiled legal threats, as I've mentioned here. Comment on content, not people. The same thing goes for the tags. They have been placed because editors feel there are legitimate issues. Those issues should be addressed, as wikipedia does operate by consensus. (While experts do make many valuable contributions, their arguments aren't intrinsically worth any more than anyone else's. Merit rules the day, and the best argument/evidence wins). And on that note, as much as I'm loathe to use google for determining importance, "Vaníček analysis" has 6 google hits whereas "least-squares spectral analysis" has 504. That would seem to me to support the move, unless there's significant reasons (backed by reliable sources) not to.

(Also, for the sake of neutrality, I do feel I should add that comments about your possible identity were inappropriate. A discussion of COI does not need to attempt any privacy violations). --Bfigura ^(talk) 01:49, 11 October 2007 (UTC)

Since you insist on the exact phrases count: Google search returns about 100 pages with the exact phrase "Vanicek spectral analysis", plus some 100 returns as well for "Gauss-Vanicek spectral analysis". Why are some 500 so much better than the competing 250? Even in the way you (Dicklyon) stated the argument (selective non-weighted counting), in most Google returns done in your way the long name has no dashes in "least squares", and the four initial letters are actually capitalized. So the present version ("Least-squares spectral analysis") is actually the least acceptable one according to the rule you (Dicklyon) suggested?! Finally, I hope someone will remove at least one of the tags from the author's biography page as there is no point of having such a scarecrow in here, is there. --Geoeg 04:02, 11 October 2007 (UTC)

Forget the Google counts; I never proposed looking at counts, and only suggested looking at book hits, since they tend to be much more reliable than web hits (books usually have the benefit on an editor). As to capitalization, wikipedia has a guideline about that: WP:NAME. As to the hyphenation, see hyphen#joining; it is very common for authors to violate the general rule, and some editors even explicitly allow it, but I prefer to follow it, to give the reader a good clue to the structure of the phrase. If it appears in any reasonable proportion of book hits (as it does), then I'm pretty confident that it's a correct and acceptable way to punctuate. If you disagree, then propose a move. Dicklyon 04:16, 11 October 2007 (UTC)

There are 20 returns without the dash v. 10 returns with the dash, in the Google book search you used. So based on your own criterion (which is fine by me), and given that there is no regulation giving your preferences priority over my preferences, as the page author I ask you to please remove the dash at your earliest convenience. Also (less relevant though) I believe that the dash actually decreases the number of search returns. If you are unable to comply with this request, please let me know and I will remove it myself. --Geoeg 14:02, 11 October 2007 (UTC)

My own criteria include "forget the counts" and "any reasonable proportion of book hits". When have I ever suggested looking at google hit counts? And there's no dash. Did you read the hyphen#joining that I referred you to? As I said above, if you want to propose a move, do so; start a new discussion section. Dicklyon 14:17, 11 October 2007 (UTC)

Call it an Indian Tiger if you like. The "book hits" come from Google Inc., do they not? So it's 20:10 in favor of no dash option. Please remove the dash (hyphen or whatever) or I shall do it. The regulation you cited is irrelevant as it refers to a style that is a matter of one's preference. I already explained this to you. Your preference is not any older than mine as the page author. --Geoeg 15:49, 11 October 2007 (UTC)

As I noted above, I don't think google hits determine notability. However, they do tend to reinforce my view that there's a more appropriate and neutral title than the Vanicek version. It seems that the most commonly used name is the least-squares one, so it makes the most sense to me. (I'll try and back that up with an ISI Web of Knowledge citation report at some point in the future, since that would be the definitive way of determining the more common name for this sort of thing). As far as the POV/COI tag on the author's page, I'll go look and see if it's still needed. Best, --Bfigura ^(talk) 04:51, 11 October 2007 (UTC)

I agree that ISI would be a good measure. Looking forward to your report. --Geoeg 14:02, 11 October 2007 (UTC)

I've removed request for sources and notability tags. Notability seems to be established, and the article seems sourced. (If more sources are needed, maybe a {{FACT}} tag could be added to indicate what needs sourcing?). PS: Will copy this comment to the author's talk page. Cheers, --Bfigura ^(talk) 05:10, 11 October 2007 (UTC)

You're referring to the Petr Vanicek article I presume, since this page never had such tags; it has been well sourced from the beginning; just not neutrally sourced, which is what the COI tag was about. Dicklyon 06:13, 11 October 2007 (UTC)

Just for the sake of clarity & completeness: my response is on the Vanicek talk page (since that seemed the logical place to continue the discussion). Cheers, --Bfigura ^(talk) 06:37, 11 October 2007 (UTC)

As far as I can see, the tags are still there? Also, I don't see any new posts by you on my talk page? --Geoeg 14:02, 11 October 2007 (UTC)

Was that directed at me? Sorry, by author's talk, I meant Vanicek's talk page. Best, --Bfigura ^(talk) 16:27, 11 October 2007 (UTC)

Double Redirects

There seem to be a fair number of double redirects created as a result of the recent page moves. (See here). I'm going to try fixing them to point here, since I can't think of a good reason to have double redirects. Cheers, --Bfigura ^(talk) 04:56, 11 October 2007 (UTC)

Yes, there are numerous redirects that were made to catch all the common misspellings of Vanicek; it doesn't actually hurt anything that they are double redirects, but there's a bot that will typically fix them within a day anyway, so don't spend too much energy on it. I already replaced the links that matter. Dicklyon 05:01, 11 October 2007 (UTC)

Oh well, done already. TOo bad I didn't read that a bit ago. --Bfigura ^(talk) 05:05, 11 October 2007 (UTC)

3RR

It would seem that WP:3RR is also about to apply here. Everyone, please slow down, and remember: we have plenty of time before we have to worry about hitting the editorial deadline. (Again, I have no interest in whether or not the current version is the right version, but this needs to be settled through consensus, not via edit warring). --Bfigura ^(talk) 17:10, 11 October 2007 (UTC)

What can I say? You saw I was ready to let it calm down, but he has just brought some partial reinforcement with him, and it is now two of them doing exactly the same thing as Dicklyon: frantically looking for more regulations that they interpret rigidly. I only wish they used that much energy on editing the actual two pages. --Geoeg 23:00, 11 October 2007 (UTC)

I've actually done some significant improvements to the page's lead and history sections, replacing deprecating statements about Scargle with actual quotes that better characterize his contribution. The reverts of Geoeg along the way are a distraction, to be sure, but I have not asked anyone to come help me. I had made a neutral request for a third opinion, as anyone can do. Dicklyon 23:42, 11 October 2007 (UTC)

Dicklyon said in the Least-squares spectral analysis edit field on 23:28, 11 October 2007: "what Scargle actually said is already quoted below; deprecating him in the lead is continued POV pushing". Then he added a whole new paragraph, well-written too. This is exciting: we finally see that indeed he can contribute something of value to the actual article(s)! Imagine all the contributions he could have made if he actually was editing the pages instead of talking enormously lot of how they should be edited... --Geoeg 23:46, 11 October 2007 (UTC)