Talk:Fibonacci sequence/Archive 2: Difference between revisions

Revision as of 05:49, 2 May 2011

Revision as of 06:22, 27 July 2011

Line 572:

:Okay with me. —[[User:Dominus|Mark Dominus]] ([[User talk:Dominus|talk]]) 04:19, 13 December 2010 (UTC)

== Systematic bias in revert by David Eppstein ==

This is related to the revert by David Eppstein and the prejudiced remarks with which a cited

fact was reverted, along with other edits.

The name [[Gopala-Hemachandra number]] is gaining currency as a general form of the Fibonacci

number. Two mathematics papers, one from the J of Number theory, were cited to show

how this term is used.

Whether the term is "a trivial extension" is not an issue on Wikipedia. The fact is that

it is being used, and we need to report it here.

If you had argued about the citations, said that they are non-authoritative, showed other

evidence that these are not coherent or otherwise, I would have understood this. On wikipedia

it doesn't suffice to say that the editor has an Indian-sounding

name and hence this is Nationalistic edit.

In fact, there is stronger evidence for [[Wikipedia:Systemic bias]] (eurocentric bias) in

this revert than any factual stance.

The revert also had other edits other than this issue, such as the incorrect "disputed" characterization.

Please re-edit the text, if needed, do not revert.

At the very least, such issues should be discussed here before reverting edits.

I am reverting the revert and will expect a civilized response on this page rather than a revert war. [[User:Mukerjee|mukerjee]] ([[User talk:Mukerjee|talk]]) 06:41, 26 March 2011 (UTC)

:The phrase "Gopala-Hemachandra number" has exactly zero hits in Google scholar. And if it had two hits, that would still be a tiny number compared to the vast literature on Fibonacci numbers. As for the disputed part, please read the paragraph in the "Origins" section, which clearly outlines the dispute among scholars about the timing of the discovery in India of the Fibonacci numbers, and clearly contradicts your poorly-sourced edits in the lead section.. —[[User:David Eppstein|David Eppstein]] ([[User talk:David Eppstein|talk]]) 06:49, 26 March 2011 (UTC)

::I get four hits, including the Basu-Prasad paper, the Thomas paper, and two more. Is gscholar geography-biased? [[User:Mukerjee|mukerjee]] ([[User talk:Mukerjee|talk]]) 07:16, 26 March 2011 (UTC)

:I reviewed the first revert carefully, as it seemed that it might be better to include this material in the article. The more I looked at it though, the more it became clear that a much more complete and well referenced treatment of the contributions of these two guys was already in the article. There's no need to put this obscure and controversial bit into the lead as well. [[User:Dicklyon|Dicklyon]] ([[User talk:Dicklyon|talk]]) 06:54, 26 March 2011 (UTC)

:David:

:You make the point

::nationalist claims of priority that are sourceable but also disputed in the literature.

:Pls cite a source that "disputes" the Indian prior claim. I can't find any and would be happy to have some. The three articles cited in the article all dispute whether the Indian work dates from 400BC or 700BC, no one argues against any priority.

:But this is an aside. The main point is that there are references using this name.

::I am attaching some segments of the paper. It is available through elsevier at http://www.sciencedirect.com/science?_ob=MImg&_imagekey=B6WKD-4YT7KMP-1-1&_cdi=6904&_user=489944&_pii=S0022314X10000533&_origin=gateway&_coverDate=09%2F30%2F2010&_sk=998699990&view=c&wchp=dGLzVlz-zSkzk&md5=f2a86d5e45ce40cf5875f53017c9befb&ie=/sdarticle.pdf

::: Here sre some extracts from the Basu-Prasad paper:

:::Journal of Number Theory 130 (2010) 1925–1931

::: Long range variations on the Fibonacci universal code

::: Manjusri Basu, Bandhu Prasad Received 15 September 2009 Revised 6 January 2010 Available online 8 April 2010 Communicated by David Goss

:::Abstract

:::Fibonacci coding is based on Fibonacci numbers and was defined by Apostolico nd Fraenkel (1987) [1]. Fibonacci numbers are generated by the recurrence relation Fi = Fi−1 + Fi−2 ∀i >= 2 with initial terms F0 = 1, F1 = 1. Variations on the Fibonacci coding are used in source coding as well as in cryptography. this paper, we have extended the table given by Thomas [8] We have found that there is no Gopala–Hemachandra code for a particular positive integer n and for a particular value of a ∈ Z . We conclude that for n = 1, 2, 3, 4, Gopala–Hemachandra code exists for a = −2, −3, . . . , −20. Also, for 1 � n � 100, there is at most m consecutive not available (N/A) Gopala–Hemachandra code in GH−(4+m) column where 1 � m � 16. And, for 1 � n � 100, as m increases the availability of Gopala–Hemachandra code decreases in GH−(4+m) column where 1 � m � 16.

:::... And section 2 has:

:::2. Gopala–Hemachandra (GH) sequence and codes

:::A variation to the Fibonacci sequence is the more general GH sequence [6] {a, b, a + b, a + 2b, 2a + 3b, 3a + 5b, . . .} for any pair a, b which for the case a = 1, b = 2 represents the Fibonacci numbers [4,5,7] gives historical details of these sequences.

::The paper is available from Elsevier at http://www.sciencedirect.com/science?_ob=MImg&_imagekey=B6WKD-4YT7KMP-1-1&_cdi=6904 (or at most univ libraries)[[User:Mukerjee|mukerjee]] ([[User talk:Mukerjee|talk]]) 07:16, 26 March 2011 (UTC)

===Origins section===

I hate edit wars.

re-reading the whole article, I find that the origin section has the sentence, which is perhaps part of what Dicklyon is referring to:

:However, according to Knuth, neither of these works concern the more specific problem of counting patterns with a fixed total duration, for which the answer is the Fibonacci numbers, and for which Knuth cites instead the anonymous Prakrta Paingala (c. 1320 AD), which uses [[Fibonacci coding]] to solve the problem.

But what Knuth says in the relevant part is:

:... consider the set of all sequences of L and S that have exactly m beats. ... there are eactly Fm+1 of them. For example the 21 sequences when m=7 are: [List]

:In this way Indian prosodists were led to discover the Fibonacci sequence, as we have observed in Section 1.2.8

:Moreover, the anonymous author of prAkr.ta paingala (c.1320) discovered elegant algorithms for ranking and unranking w.r.t. m-beat rhythms. p. 50

Thus the statement "neither of these works concern the more specific problem of counting patterns with a fixed total duration, for which the answer is the Fibonacci numbers" appears to be a deliberate misleading the reader. In fact, Knuth says unequivocally in v.1 p.100:

: Before Fibonacci wrote his work, the sequence Fn had already been discussed by Indian scholars, who had long been interested in rhythmic patterns... both Gopala (before 1135AD) and Hemachandra (c.1150) mentioned the numbers 1,2,3,5,8,13,21 ... explicitly. [See P. Singh Historia Math 12 (1985) 229-244] AOP v.1 3d ed. p. 100

I am not sure who inserted this sentence, but if there is evidence of prejudice it is in this sentence, which is also used to make an incorrect claim about what was disputed in the lead.

Would anyone object if I edited the origins section, or is the above text justified? [[User:Mukerjee|mukerjee]] ([[User talk:Mukerjee|talk]]) 09:01, 26 March 2011 (UTC)

:I don't have immediate access to a recent edition of Knuth Vol.I which is what you seem to be quoting; the part about Gopala and Hemachandra certainly isn't in the part of Knuth that is cited in the article. Maybe someone else who does have a recent edition and can check what Mukerjee says would like to weigh in? —[[User:David Eppstein|David Eppstein]] ([[User talk:David Eppstein|talk]]) 23:11, 26 March 2011 (UTC)

::there is no problem accessing [[AOP]] - i have also provided the google books links.

::Even if others haven't accessed it yet, it follows wikipedia verifiability norms so I am going ahead and editing in the changes.

:: Pls note that wikipedia [[wikipedia:Editing policy|policy]] suggests: "Please boldly add information to Wikipedia, either by creating new articles or adding to existing articles, and exercise particular caution when considering removing information."

::So pls take added care in removing edits that contain verifiable facts.

::I will next be re-inserting some of the original text in the lead. [[User:Mukerjee|mukerjee]] ([[User talk:Mukerjee|talk]])

== Sunflower image ==

I removed the sunflower image since when I tried to verify the spiral count I kept getting different results. I think the particular sunflower in the image had irregularities that made the spirals too unstable to make the counts meaningful. The other sunflower images I found weren't much better but I did find a chamomile image where I could draw in the spirals. I also removed the seashell image, the connection to the Fibonacci number has been claimed by some authors but my understanding is that this has been debunked.--[[User:RDBury|RDBury]] ([[User talk:RDBury|talk]]) 02:05, 28 April 2011 (UTC)