Talk:meanless

Request for verification
"Without a mathematical mean." All I could find was a 2004 book: "the average is used to get a meanless dataset". Equinox ◑ 16:29, 29 May 2009 (UTC)
 * Statistics context makes no sense. Any real sample is finite and therefore has a mean. A probability distribution without a mean can exist in a mathematician's imagination. I don't understand the quote. DCDuring TALK 17:24, 29 May 2009 (UTC)


 * In your cite it seems to mean roughly "normalized, such that the mean is zero and the expressions can be written without a mean". In this one it also seems to mean "having a mean of zero", but I'm really not sure. By contrast, in this one it really does mean "having no mean whatsoever": if you set up the improper integral, you'll find that it diverges to infinity; see Power law distribution. (I suppose one could also say "having a mean of +∞", but "having no mean" is also a valid interpretation.) —Ruakh TALK 17:28, 29 May 2009 (UTC)


 * There seems to be be some sense that comes up in signal detection and/or encryption that may be the given sense, but confirming it and entering the formulas are both beyond my paygrade. To say "meanless" means "having a mean of zero" really bothers me, especially with a mathematics context tag. It seems so literary. DCDuring TALK 17:58, 29 May 2009 (UTC)
 * It shouldn't bother you. Mathematicians frequently speak of "no|without holonomy", "no|without homotopy", etc., meaning that the holonomy or homotopy is zero. Brevity, though the soul of inaccuracy in mathematics, still makes for easier reading.—msh210 ℠  16:12, 3 June 2009 (UTC)


 * Perhaps someone can extract something from these 4 from b.g.c. DCDuring TALK 18:01, 29 May 2009 (UTC)


 * I'm not certain about this, but I think that in all four of those, it means "meaningless". One says that L represents blah-blah-blah when it's at least 3, but that it's "meanless" otherwise; one says that you can use dual-rail logic (which encodes 0s and 1s in a higher-order stream with equal numbers of 0s and 1s — it's usually used to combat channel bias) to make Hamming-weights (numbers of 1s in strings of 0s and 1s) "meanless" to would-be decrypters; one describes various kinds of image analysis that are limited by certain assumptions or requirements, including one that's useless in situation foo and another that's "meanless" in situation bar; and the remaining one explains that they're not bothering to distinguish horizontal from vertical because said distinction is "meanless" when you're just dealing with numerical images. —Ruakh TALK 18:30, 29 May 2009 (UTC)


 * Perhaps: the number of "meanlesses" on Google Groups (Usenet) as errors for "meaningless" is very high indeed. Equinox ◑ 18:40, 29 May 2009 (UTC)


 * Statisticians in writing would generally use more formal language, saying a distribution had an "undefined mean" (as with the Cauchy distribution or certain parameterizations of the F-distribution), but I've heard "meanless" used in sense orally (sorry, don't have time to find cites). --Bequw → ¢ • τ 22:12, 29 May 2009 (UTC)


 * Perhaps try Scholar if you have the chance. There seemed to be some, but I couldn't get enough access to be sure. I looked through b.g.c. without much luck, but you might have better google-fu. DCDuring TALK 22:37, 29 May 2009 (UTC)


 * The OED has two senses, one from the noun (ie "without the means") meaning "without agency or aid" (a citation mentions "meanless miracles" which ceased since Christ's ascension). The other from the verb, meaning "meaningless".  Both marked obsolete, and the first rare as well.  < class="latinx">Ƿidsiþ 16:21, 3 June 2009 (UTC)

Cited, I think. I combined "mean = 0" and "mean DNE" into one sense, which may be cheating, but I'm pretty sure that all three cites must mean one or the other of those. —Ruakh TALK 21:42, 27 December 2009 (UTC)