Skip to content
#small language model Open access

One and the Same "Proportional To" Carries Four Different Kinds of Constant ── a sphere's S/V^2/3 is fixed by geometry, a pendulum's Tsqrt(g/L) moves with amplitude, a power law fitted to the primes moves with the window, and the Catalan coefficient reaches only 0.99988751 of its limit even at n=10000 ── [Paper 530]

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)

Abstract

When one writes "A is proportional to B", the standing of the constant in front is rarely asked about. What this paper shows is that such constants come in four kinds, some of which move when something is changed and some of which do not. No new theorem or law is claimed. Scope of this paper (scope note): No new theorem or law is claimed──dimensional analysis, power-law fitting and the asymptotics of Catalan numbers are all standard. The axis differs from Paper 117──that paper divided the ways an exponent is chosen into identity, dimension matrix, measurement and fixed point. This paper leaves the exponent alone and divides the standing of the constant. The two fourfold splits are not the same. The ubiquity of power laws is not discussed──why nature offers so many of them is not entered. Fitting methods are not compared──least squares against maximum likelihood is not the subject. No claim about prime distribution is made──the prime counting function is used as a specimen of fitting a power law to something that is not one; nothing in number theory is asserted. No asymptotic expansion is offered──the next term for Catalan numbers is not treated; only that the ratio tends to one. It is not claimed the four exhaust the possibilities──four were counted, and no claim is made that there are no others. Relation to earlier papers: Paper 117 showed that scale invariance fixes an exponent in four different ways──this paper looks at the coefficient instead, so the contents of the four differ, and that difference is declared at the outset. Paper 326 showed that the square-cube law is an identity──that is the first kind here, a constant fixed by geometry. Paper 372 treated an exponent relating area and species──that exponent is a fitted value, and the third kind here measures what such a value is like. Paper 144 showed that an exponent is the signature of what is conserved──it can be read as a signature when the constant's standing is geometric or asymptotic. Paper 300 showed that sharing a root is decidable──the four standings are separate roots under the criterion of what a quantity carries. What is added is setting the four side by side and measuring what moves each, measuring the pendulum's group moving by 1.1803, measuring a fitted exponent moving from 0.827369 to 0.912949, showing a limit unreached at 0.99988751, and placing the separator on what must be changed. First, some constants are fixed by geometry──a sphere's S/V^2/3 is 4.835976, agreeing with an independently assembled (36pi)^1/3 to 12 digits, while a cube gives 6.000000 (Section 2). Second, dimensional analysis fixes no constant──a pendulum's Tsqrt(g/L) is 6.283186 at small amplitude, matching 2pi=6.283185, but 7.416299 at 90 degrees, a factor of 1.1803 (Section 3). Third, and this is the core. A fitted constant moves with the window it is fitted in──fitting a power law to the prime counting function gives 0.827369 on [100,1000] and 0.912949 on [10^5,10^6] (Section 4). Fourth, that is because the truth is not a power law──the local slope is 1-1/ln x, which is 0.855235 at x=10^3 and 0.927618 at x=10^6 and keeps moving (Section 4). Fifth, a limit is never reached at finite size──the Catalan coefficient is 0.89780266 of 1/sqrt(pi) at n=10 and still only 0.99988751 at n=10000 (Section 5). Sixth, the separator is what must be changed to move the constant──shape, or range, or window, or n: the four answer to different things (Section 6). When one writes that A is proportional to B, the standing of the constant in front is rarely asked about. A geometric constant ignores size and answers to shape──a sphere's S/V^2/3 is 4.835976, agreeing to 12 digits with (36pi)^1/3 assembled without the area and volume formulas. A cube stays at 6.000000 from a=1 to a=7.3, which is an identity rather than a measured agreement. And yet sphere and cube differ, so the constant is fixed again whenever the geometry changes and is not universal. Dimensional analysis fixes no constant──the form T proportional to sqrt(L/g) follows from dimensions while the number does not, and 2pi appears only on solving the equation of motion. That 2pi is itself conditional: 6.283186 at small amplitude becomes 7.416299 at 90 degrees, a factor of 1.1803. It moves because the amplitude is itself dimensionless and supplies a second group, the very case Paper 117 excluded by requiring exactly one. A fitted value moves with its window──fitting a power law to the prime counting function gives 0.827369 on [10^2,10^3] and 0.912949 on [10^5,10^6], all below one, so the fit looks successful while the value refuses to settle. The truth is x/ln x, whose local slope 1-1/ln x runs from 0.855235 to 0.927618, so the fits were averaging a moving quantity. A fitted value therefore obliges one to state the window, and the test is easy: split it and fit both halves. A limit is not reached at finite size──the Catalan coefficient is 0.89780266 of 1/sqrt(pi) at n=10 and still 0.99988751 at n=10000, which is the definition of this standing rather than an error. This is where it parts from a fitted value: both disagree at finite size, but only a limit has a destination. The separator is what must be changed to move the constant──shape, the value of a second group, the window, or n. Under the criterion of Paper 300, a geometric constant carries a shape and a fitted value carries a window, and quoting the number without its belongings erases the standing. Paper 144 read an exponent as the signature of a conservation law, which works when the standing is geometric or asymptotic and fails when it is fitted. To be honest──Paper 117 divided the ways an exponent is chosen; this paper divides the standing of the coefficient, so the two fourfold splits differ in content. No claim is made that four exhausts the list, since exhaustion requires first fixing how to exhaust. The ubiquity of power laws and the merits of fitting methods were not discussed, the prime counting function served only as a specimen with no claim in number theory, and the next term for Catalan numbers was not treated. What can be said is that four constants of different standing stand in front of the same symbol, and each answers to a different change, and no further. On the making of this work: The ideas and content of this work stem from the author's own considerations. Assistance from an AI (a large language model) was used for structuring, English translation, and checking the algebra. Any remaining errors or misinterpretations are solely the author's. Feedback and corrections are sincerely appreciated. Keywords: constant of proportionality, dimensional analysis, power law, Catalan numbers, scaling. ----- 「A は B に比例する」と書くとき、前に立つ定数の身分は問われないことが多い。本稿が示すのは、その定数に四つの身分があり、動かせるものと動かせないものが混じっていることである。新しい定理も法則も主張しない。 本稿の射程(射程注記):新しい定理も法則も主張しない──次元解析も、冪則の当てはめも、カタラン数の漸近も標準的である。論文117 と軸が違う──あちらは指数がどう選ばれるかを恒等式・次元行列・経験・不動点の四通りに分けた。本稿は指数ではなく比例定数の身分を分ける。同じ四分割ではない。冪則の普遍性を論じない──自然界に冪則が多いことの理由には立ち入らない。当てはめの手法を評価しない──最小二乗と最尤のどちらがよいかは主題ではない。素数分布を主張しない──素数計数は「冪則でないものに冪則を当てはめるとどうなるか」の見本として使う。数論の主張はしない。漸近展開を出さない──カタラン数の次の項は扱わず、比が 1 に寄ることだけを見る。四つで尽きるとは書かない──本稿が数えたのは四つで、他に無いとは主張しない。既刊との関係:論文117 はスケール不変性が四通りに指数を選ぶと示した──本稿は指数ではなく係数を見るので、四つの中身が違う。冒頭でその差を宣言する。論文326 は二乗三乗の法則が恒等式だと示した──本稿の第一の身分がそれで、定数が幾何で決まる。論文372 は面積と種数の冪を扱った──あちらの指数は当てはめ値である。本稿の第三の身分が、その身分の性質を測る。論文144 は指数が何が保存しているかの署名だと示した──署名として読めるのは、定数の身分が「幾何」か「極限」のときである。論文300 は同根か別根かが判定できると示した──四つの身分は「量の持ち物」の基準で別根である。加えたのは、四つの身分を並べて、それぞれ動かすと動く相手を測ったこと、振り子の Pi 群が振幅で 1.1803 倍動くと出したこと、当てはめた指数が窓で 0.827369 から 0.912949 まで動くと測ったこと、極限値が有限の n では届かないことを 0.99988751 という数で示したこと、分離子を「何を動かすと定数が動くか」に置いたことである。 第一に、定数が幾何で決まる場合がある──球の S/V^2/3 は 4.835976 で、独立に組んだ (36pi)^1/3 と 12 桁一致し、立方体では 6.000000 になる(第2節)。 第二に、次元解析は定数を決めない──振り子の Tsqrt(g/L) は小角で 6.283186(2pi=6.283185)だが、振幅 90 度では 7.416299 で、1.1803 倍になる(第3節)。 第三に、これが本稿の芯である。当てはめた定数は、当てはめる窓で動く──素数計数に冪則を当てはめると、窓 [100,1000] で 0.827369、[10^5,10^6] で 0.912949 になる(第4節)。 第四に、それは真の姿が冪則でないからである──局所の傾きは 1-1/ln x で、x=10^3 で 0.855235、x=10^6 で 0.927618 と動き続ける(第4節)。 第五に、極限値は、有限では届かない──カタラン数の係数は n=10 で 1/sqrt(pi) の 0.89780266 倍、n=10000 でも 0.99988751 倍で、厳密には一致しない(第5節)。 第六に、分離子は「何を動かすと定数が動くか」である──大きさか、範囲か、窓か、n か。四つで動く相手が違う(第6節)。 「A は B に比例する」と書くとき、前に立つ定数の身分は問われないことが多い。幾何で決まる定数は、大きさでは動かず形で動く──球の S/V^2/3 は 4.835976 で、面積と体積の式を通さずに組んだ (36pi)^1/3 と 12 桁一致する。立方体では a を 1 から 7.3 まで動かしても 6.000000 のままで、これは測った一致ではなく恒等式である。それでも球と立方体で値が違うので、幾何を変えれば決まり直す。普遍定数ではない。次元解析は定数を決めない──振り子の T proportional to sqrt(L/g) という形は次元だけで出るが、前の数は出ず、2pi は運動方程式を解いて初めて現れる。しかもその 2pi は条件付きで、小角の 6.283186 が振幅 90 度では 7.416299、1.1803 倍になる。動く理由は、振幅そのものが無次元で Pi 群が二つあるからである。論文117 が「Pi 群がちょうど一つ」という条件を立てたのは、この場合を除くためである。当てはめ値は、当てはめる窓で動く──素数計数に冪則を当てはめると、窓 [10^2,10^3] で 0.827369、[10^5,10^6] で 0.912949 になる。どの窓でも 1 より小さいので当てはめは成功して見えるが、値は一つに定まらない。真の姿は冪則ではなく x/ln x で、局所の傾きは 1-1/ln x、x=10^3 で 0.855235、x=10^6 で 0.927618 と動き続ける。当てはめた値はこれを窓の中で平均していただけである。だから当てはめ値には窓を書く義務がある。見分ける手続きも簡単で、窓を二つに割って両方で当てはめ、値がずれれば当てはめ値である。極限値は、有限では届かない──カタラン数の係数は n=10 で 1/sqrt(pi) の 0.89780266 倍、n=10000 でも 0.99988751 倍で、厳密には一致しない。これは誤差ではなく、この身分の定義である。当てはめ値との違いはここで見える。当てはめ値は窓を変えると動いてどこにも収束しないが、極限値は n とともに一つの値へ寄る。寄る先が在るかどうかが分かれ目である。分離子は、何を動かすと定数が動くかである──幾何なら形、次元解析なら無次元群の値、当てはめなら窓、極限なら n。四つで動く相手が違う。論文300 の「量の持ち物」でいえば、幾何の定数は形を持ち物にし、当てはめ値は窓を持ち物にする。持ち物を書かずに数だけを引用すると、身分が消える。論文144 は指数が保存則の署名だと示したが、署名として読めるのは身分が幾何か極限のときで、当てはめ値の指数は保存則ではなく窓を語っている。正直に言えば──論文117 は指数がどう選ばれるかを四通りに分けた。本稿は指数ではなく係数の身分を分けたので、同じ「四つ」でも中身は別である。四つで尽きるとも書かない。数えたのは四つで、尽くしたと言うには尽くし方を先に決めなければならない。冪則の普遍性も、当てはめ手法の優劣も論じておらず、素数計数は冪則でないものに冪則を当てはめる見本として使っただけで、数論の主張はしていない。カタラン数の次の項にも触れていない。言えるのは、同じ記号の前に身分の違う定数が四つ立っており、それぞれ別のものを動かすと動く、そこまでである。 作成にあたって:本稿の着想と内容は、著者自身の考察に基づくものです。文章の構成整理や英訳、数式の確認には AI(大規模言語モデル)の助力を得ました。最終的な内容の解釈や誤りがあれば、それらはすべて著者の責に帰します。お気づきの点があれば、ご教示いただければ幸いです。 キーワード:比例定数、次元解析、冪則、カタラン数、スケーリング。

View source

Similar papers

#small language model Dataset Open access Oct 2026

Socratic guiding questions in synthetic arithmetic data: matched LoRA runs (revision v2)

Supporting data, adapters, predictions and code for the article *Low-Cost LoRA Fine-Tuning of Small Language Models for Multi-Step Arithmetic Reasoning* by Jake O'Grady, Asena Isik Gürhan, Chee Fong Ting and Effirul Ramlan (University of Galway). We generated 20,000 GSM8K-derived arithmetic problems with step-by-step s...

O'Grady, Jake, Gürhan, Asena Isik, Chee, Fong Ting et al. · 465 citations
#computer vision Open access Jun 2016

Software Development in Startup Companies: The Greenfield Startup Model

The results are packaged in the Greenfield Startup Model (GSM), which explains the priority of startups to release the product as quickly as possible, and the need to shorten time-to-market, by speeding up the development through low-precision engineering activities.

Carmine Giardino, Nicolò Paternoster, M. Unterkalmsteiner et al. · 178 citations · ⚡14
#computer vision Open access Oct 2016

Software Startups - A Research Agenda

Software startup companies develop innovative, software-intensive products within limited timeframes and with few resources, searching for sustainable and scalable business models.

M. Unterkalmsteiner, P. Abrahamsson, Xiaofeng Wang et al. · 157 citations · ⚡17
#machine learning Review Open access Oct 2016

“Failures” to be celebrated: an analysis of major pivots of software startups

This study conducts a case survey study based on the secondary data of the major pivots happened in 49 software startups, and demonstrates that customer need pivot is the most common among all pivot types.

Sohaib Shahid Bajwa, Xiaofeng Wang, Anh Nguyen-Duc et al. · 127 citations · ⚡15
#computer vision Review Open access May 2015

A survey study on major technical barriers affecting the decision to adopt cloud services

The comparison of adopter and non-adopter sample reveals three potential adoption inhibitor, security, data privacy, and portability, which underlines the importance of the technical and security perspectives for research investigating the adoption of technology.

Nattakarn Phaphoom, Xiaofeng Wang, S. Samuel et al. · 111 citations · ⚡8
#computer vision Open access Feb 2018

Lean Internal Startups for Software Product Innovation in Large Companies: Enablers and Inhibitors

This study investigates how Lean internal startup facilitates software product innovation in large companies and identifies its enablers and inhibitors, and shows the potential of the method-in-action framework to investigate the Lean startup approach in non-startup context.

Henry Edison, Nina M. Smørsgård, Xiaofeng Wang et al. · 78 citations · ⚡6

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.