Skip to content

Latest commit

 

History

History
290 lines (244 loc) · 25.9 KB

File metadata and controls

290 lines (244 loc) · 25.9 KB

Mathematical Derivations

$\Lambda(n)$ and Manifestations of $\Lambda(6) = 1/\sqrt{21}$

Continued in part B, part C, part D, and part E.

1. Algebraic and Analytical Foundations of $\Lambda(n)$

1.1. Discrete Evaluation

Fundamental formula: $$\Lambda(n) = \sqrt{\frac{2}{n(n+1)}} = \frac{1}{\sqrt{T(n)}}, \quad n \in \mathbb{N}_{>0}$$

For index $n = 6$:

  1. The sixth triangular number: $$T(6) = \frac{6(6+1)}{2} = \frac{42}{2} = 21$$
  2. Substitution into the formula: $$\Lambda(6) = \sqrt{\frac{2}{6(7)}} = \sqrt{\frac{2}{42}} = \sqrt{\frac{1}{21}} = \frac{1}{\sqrt{21}}$$
  3. High-precision numerical evaluation: $$\Lambda(6) = 0.2182178902359923812660974854156\ldots$$ $$100 \cdot \Lambda(6) = 21.82178902359923812660974854156\ldots$$

1.2. Partial Fraction Decomposition and Telescoping Sum

Given $\Lambda(n)^{2} = \frac{2}{n(n+1)}$, the decomposition is the standard partial fraction decomposition:

  1. Partial fraction decomposition: $$\frac{2}{n(n+1)} = \frac{A}{n} + \frac{B}{n+1} \implies 2 = A(n+1) + Bn$$ Setting $n = 0$ gives $A = 2$. Setting $n = -1$ gives $B = -2$. $$\Lambda(n)^{2} = 2\left(\frac{1}{n} - \frac{1}{n+1}\right)$$
  2. Summation up to upper limit $N$: $$\sum_{n=1}^{N} \Lambda(n)^{2} = 2 \sum_{n=1}^{N} \left(\frac{1}{n} - \frac{1}{n+1}\right) = 2\left[\left(1 - \frac{1}{2}\right) + \left(\frac{1}{2} - \frac{1}{3}\right) + \cdots + \left(\frac{1}{N} - \frac{1}{N+1}\right)\right]$$
  3. Telescoping cancellation of all intermediate terms: $$\sum_{n=1}^{N} \Lambda(n)^{2} = 2\left(1 - \frac{1}{N+1}\right) = \frac{2N}{N+1}$$
  4. Convergence in the limit as $N \to \infty$: $$\lim_{N \to \infty} \sum_{n=1}^{N} \Lambda(n)^{2} = \lim_{N \to \infty} \frac{2N}{N+1} = 2$$

1.3. Generating Function of Squared Terms $G(z)$

Power series definition: $$G(z) = \sum_{n=1}^\infty \Lambda(n)^{2} z^{n} = 2 \sum_{n=1}^\infty \left(\frac{1}{n} - \frac{1}{n+1}\right) z^{n}, \quad \lvert z\rvert < 1$$

  1. Splitting the series: $$G(z) = 2 \sum_{n=1}^\infty \frac{z^{n}}{n} - \frac{2}{z} \sum_{n=1}^\infty \frac{z^{n+1}}{n+1}$$
  2. Mercator series for $-\ln(1-z)$ (Mercator series): $$\sum_{n=1}^\infty \frac{z^{n}}{n} = -\ln(1-z)$$ $$\sum_{n=1}^\infty \frac{z^{n+1}}{n+1} = \sum_{k=2}^\infty \frac{z^{k}}{k} = -\ln(1-z) - z$$
  3. Substitution and algebraic simplification: $$G(z) = -2\ln(1-z) - \frac{2}{z}\left(-\ln(1-z) - z\right) = -2\ln(1-z) + \frac{2\ln(1-z)}{z} + 2$$ $$G(z) = 2 + \frac{2(1-z)\ln(1-z)}{z}$$

1.4. Continuous Generalization through the Euler Beta and Gamma Functions

  1. Definition of the Euler Beta function $B(x, y) = \frac{\Gamma(x)\Gamma(y)}{\Gamma(x+y)}$ (Beta function and Gamma function): Setting parameter $y = 2$: $$B(x, 2) = \frac{\Gamma(x)\Gamma(2)}{\Gamma(x+2)} = \frac{\Gamma(x) \cdot 1!}{(x+1)x,\Gamma(x)} = \frac{1}{x(x+1)}$$
  2. Continuous representation: $$\Lambda(x) = \sqrt{2,B(x, 2)} = \sqrt{\frac{2,\Gamma(x)}{\Gamma(x+2)}} = \sqrt{\frac{2}{x(x+1)}}, \quad x > 0$$
  3. Derivation of the logarithmic derivative $\frac{d}{dx} \ln \Lambda(x)$: $$\ln \Lambda(x) = \frac{1}{2}\ln 2 - \frac{1}{2}\ln x - \frac{1}{2}\ln(x+1)$$ $$\frac{d}{dx} \ln \Lambda(x) = -\frac{1}{2x} - \frac{1}{2(x+1)}$$

2. Ten Combinatorial Evaluations of 21

The integer $21$ has the ten evaluations below. They are not independent counts: $T(6)=\binom{7}{2}$, and both Stirling evaluations equal that binomial coefficient by the identities $S(n,n-1)=c(n,n-1)=\binom{n}{2}$ (unsigned Stirling numbers of the first kind and Stirling numbers of the second kind).

  1. Triangular Number ($T_{6}$) $$T(6) = \frac{6 \times 7}{2} = 21$$
  2. Binomial Coefficient $\binom{7}{2}$ $$\binom{7}{2} = \frac{7!}{2!,5!} = \frac{7 \times 6}{2} = 21$$
  3. Octagonal Number ($O_{3}$) The $s$-gonal formula is $P(s,n)=((s-2)n^{2}-(s-4)n)/2$ (polygonal number). For $s=8$: $$P(8,3) = 3(3\cdot 3-2) = 21$$
  4. Icosihenagonal (21-gonal) Number ($I_{21}(2)$) $$P(21, 2) = \frac{19(2^{2}) - 17(2)}{2} = \frac{76 - 34}{2} = 21$$
  5. Fibonacci Number ($F_{8}$) With $F_{0}=0$ and $F_{1}=1$ (Fibonacci sequence): $$F_{0}=0, F_{1}=1, F_{2}=1, F_{3}=2, F_{4}=3, F_{5}=5, F_{6}=8, F_{7}=13, F_{8}=21$$
  6. Stirling Number of the Second Kind $S(7, 6)$ $$S(n, n-1) = \binom{n}{2} \implies S(7, 6) = \binom{7}{2} = 21$$
  7. Unsigned Stirling Number of the First Kind $c(7, 6)$ $$c(n, n-1) = \binom{n}{2} \implies c(7, 6) = \binom{7}{2} = 21$$
  8. Motzkin Number ($M_{5}$) The recurrence is $M_{n} = M_{n-1} + \sum_{i=0}^{n-2} M_{i} M_{n-2-i}$, and the sequence begins $1,1,2,4,9,21$ (Motzkin number): $$M_{0}=1, M_{1}=1, M_{2}=2, M_{3}=4, M_{4}=9, M_{5}=21$$
  9. Standard Young Tableaux $f^{(3,3,1)}$ (Hook Length Formula) $$f^\lambda = \frac{n!}{\prod h_{\lambda}(i,j)}$$ (hook-length formula) For partition $\lambda = (3,3,1)$ with $n = 7$:
    • Row 1 hook lengths: $(5, 3, 2)$
    • Row 2 hook lengths: $(4, 2, 1)$
    • Row 3 hook lengths: $(1)$ $$\prod h_{(i,j)} = 5 \times 3 \times 2 \times 4 \times 2 \times 1 \times 1 = 240 \implies f^{(3,3,1)} = \frac{7!}{240} = \frac{5040}{240} = 21$$
  10. Standard Young Tableaux $f^{(3,2,2)}$ For partition $\lambda = (3,2,2)$ with $n = 7$:
    • Row 1 hook lengths: $(5, 4, 1)$
    • Row 2 hook lengths: $(3, 2)$
    • Row 3 hook lengths: $(2, 1)$ $$\prod h_{(i,j)} = 5 \times 4 \times 1 \times 3 \times 2 \times 2 \times 1 = 240 \implies f^{(3,2,2)} = \frac{5040}{240} = 21$$

3. Derivations in Analytic Number Theory

3.1. Quadratic Gauss Sum $g(\chi_{21})$

The Kronecker page identifies the real primitive character of discriminant $21$ by $\chi_{21}(n)=\left(\frac{21}{n}\right)$, with conductor $|D|=21$ (Kronecker symbol). Because the odd part of $21$ is $1 \pmod 4$, that page also gives $\left(\frac{\cdot}{21}\right)=\left(\frac{21}{\cdot}\right)$. Its factorization into odd Legendre symbols is the Jacobi definition (Jacobi symbol): $$\chi_{21}(n)=\left(\frac{21}{n}\right)=\left(\frac{3}{n}\right)\left(\frac{7}{n}\right)$$ $$g(\chi_{21}) = \sum_{n=0}^{20} \chi_{21}(n), e^{2\pi i n/21}$$

  1. The equality $\lvert g(\chi)\rvert=\sqrt{N}$ requires a primitive character, equivalently conductor equal to modulus (Gauss sum and Dirichlet character). Being square-free does not by itself prove primitivity. The conductor statement above supplies that hypothesis, so $\lvert g(\chi_{21})\rvert=\sqrt{21}$. For coprime moduli, the Gauss-sum page includes the phase $\chi(N')\chi'(N)$. Here $\chi_3(7)\chi_7(3)=-1$: $$g(\chi_{21}) = \chi_{3}(7)\chi_{7}(3),g(\chi_{3})g(\chi_{7}) = -g(\chi_{3})g(\chi_{7}).$$
  2. Modulus square of primitive real Dirichlet characters modulo prime $p$: $$\lvert g(\chi_{p})\rvert^{2} = p \implies \lvert g(\chi_{3})\rvert^{2} = 3, \quad \lvert g(\chi_{7})\rvert^{2} = 7$$
  3. Product of moduli: $$\lvert g(\chi_{21})\rvert^{2} = 3 \times 7 = 21 \implies \lvert g(\chi_{21})\rvert = \sqrt{21}$$ $$\Lambda(6) = \frac{1}{\lvert g(\chi_{21})\rvert} = \frac{1}{\sqrt{21}}$$

3.2. Dirichlet Class Number Formula

For the real quadratic field $\mathbb{Q}(\sqrt{21})$:

  • Fundamental discriminant $d = 21$ ($21 \equiv 1 \pmod 4$)
  • Fundamental unit $\varepsilon_{21} = (5 + \sqrt{21})/2$, of norm $+1$, since $5^{2} - 21\cdot 1^{2} = 4$. The continued fraction of $\sqrt{21}$ is $[4;\overline{1,1,2,1,1,8}]$, so the repeating block has length $6$ and the period is $r=6$. Counting from the initial term, the sixth convergent, written $h_{5}/k_{5}$, is $55/12$, and $55^{2}-21\cdot 12^{2}=1$. Because $r$ is even, the fundamental-solution rule selects convergent index $r-1=5$, which is the same fraction (continued fraction). The solution is the cube of the unit: $\varepsilon_{21}^{3}=55+12\sqrt{21}$ (Pell's equation). The class-number formula uses $\ln \varepsilon_{21}$, not $\ln(55 + 12\sqrt{21})$, which is three times as large (class number formula and fundamental unit).
  • Class number $h(21) = 1$ (class-number-one fields). In the formula below, $\chi_{21}(m)=\left(\frac{21}{m}\right)$ is the same character as in section 3.1 (class number formula).

The analytic Dirichlet class number formula is: $$h(d) = \frac{\sqrt{d}}{2\ln \varepsilon_{d}} L(1, \chi_{d})$$ Substituting $h(21) = 1$: $$1 = \frac{\sqrt{21}}{2\ln \left(\frac{5+\sqrt{21}}{2}\right)} L(1, \chi_{21}) \implies \frac{1}{\sqrt{21}} = \frac{L(1, \chi_{21})}{2\ln \left(\frac{5+\sqrt{21}}{2}\right)} = \Lambda(6)$$

3.3. Cusp Form Space Dimension $\dim S_{24}(\Gamma_{0}(6))$

Dimension formula for cusp forms of even weight $k \ge 4$ on congruence subgroup $\Gamma_{0}(N)$ (Stein, Proposition 6.1, citing Diamond and Shurman): $$\dim S_{k}(\Gamma_{0}(N)) = (k-1)(g-1) + \left\lfloor \frac{k}{4} \right\rfloor \nu_{2} + \left\lfloor \frac{k}{3} \right\rfloor \nu_{3} + \left(\frac{k}{2} - 1\right) \nu_{\infty}$$

Geometric parameters of modular curve $X_{0}(6)$ for level $N = 6$:

  1. Modular index: $\mu = [\mathrm{SL}_{2}(\mathbb{Z}) : \Gamma_{0}(6)] = 6 \left(1 + \frac{1}{2}\right)\left(1 + \frac{1}{3}\right) = 12$
  2. Elliptic points: Stein sets $\mu_{0,2}(N)=0$ when $4 \mid N$, and otherwise $\mu_{0,2}(N)=\prod_{p \mid N}\left(1+\left(\frac{-4}{p}\right)\right)$. Since $4 \nmid 6$, the product is required, and $\left(\frac{-4}{2}\right)=0$ makes $\nu_{2}=0$. Stein sets $\mu_{0,3}(N)=0$ when $2 \mid N$, so $\nu_{3}=0$ (Stein, definitions preceding Proposition 6.1)
  3. Cusps: $\nu_{\infty} = \sum_{d \mid 6} \phi(\gcd(d, 6/d)) = \phi(1) + \phi(1) + \phi(1) + \phi(1) = 4$
  4. Riemann surface genus $g$: $$g = 1 + \frac{\mu}{12} - \frac{\nu_{2}}{4} - \frac{\nu_{3}}{3} - \frac{\nu_{\infty}}{2} = 1 + \frac{12}{12} - 0 - 0 - \frac{4}{2} = 0$$
  5. Evaluation at weight $k = 24$: $$\dim S_{24}(\Gamma_{0}(6)) = (24-1)(0-1) + 0 + 0 + \left(\frac{24}{2} - 1\right) \times 4 = -23 + (11 \times 4) = -23 + 44 = 21$$ $$\Lambda(6) = \frac{1}{\sqrt{\dim S_{24}(\Gamma_{0}(6))}} = \frac{1}{\sqrt{21}}$$

3.4. Ratio Identity of Euler Zeta Values $\zeta(6)/\zeta(4)$

Euler formula for even integer values of the Riemann zeta function (particular values): $$\zeta(2k) = \frac{(-1)^{k-1} (2\pi)^{2k} B_{2k}}{2(2k)!}$$

  1. For $k = 2$ ($B_{4} = -1/30$): $$\zeta(4) = \frac{(2\pi)^{4} (1/30)}{2(24)} = \frac{16\pi^{4}}{1440} = \frac{\pi^{4}}{90}$$
  2. For $k = 3$ ($B_{6} = 1/42$): $$\zeta(6) = \frac{(2\pi)^{6} (1/42)}{2(720)} = \frac{64\pi^{6}}{60480} = \frac{\pi^{6}}{945}$$
  3. Ratio of zeta values: $$\frac{\zeta(6)}{\zeta(4)} = \frac{\pi^{6} / 945}{\pi^{4} / 90} = \pi^{2} \cdot \frac{90}{945} = \frac{2\pi^{2}}{21}$$
  4. Substitution of $\Lambda(6)^{2} = 1/21$: $$\frac{\zeta(6)}{\zeta(4)} = 2\pi^{2} \Lambda(6)^{2} \implies \Lambda(6) = \sqrt{\frac{\zeta(6)}{2\pi^{2} \zeta(4)}} = \frac{1}{\sqrt{21}}$$

4. Applied Mathematical and Statistical Derivations

4.1. Artificial Intelligence and Machine Learning

  1. Transformer Attention Scaling (Vaswani et al. 2017) $$A(Q, K) = \mathrm{softmax}\left(\frac{QK^{T}}{\sqrt{d_{k}}}\right)$$ For key projection dimension $d_{k} = 21$: $$\text{Scale} = \frac{1}{\sqrt{d_{k}}} = \frac{1}{\sqrt{21}} = \Lambda(6)$$
  2. Xavier and Glorot Weight Initialization (Glorot and Bengio 2010) The cited initialization is uniform on $[-\sqrt{6/(n_{in}+n_{out})}, \sqrt{6/(n_{in}+n_{out})}]$. The endpoint equals $\Lambda(6)$ when $n_{in}+n_{out} = 126$. The Gaussian shorthand $\sigma = \sqrt{1/n_{in}}$ is a later variant, not the formula in that paper.
  3. He (Kaiming Normal - ReLU) Weight Initialization (He et al. 2015) $$\sigma = \sqrt{\frac{2}{n_{l}}}$$ For fan-in $n_{l} = 42 = 2 \cdot T(6)$: $$\sigma = \sqrt{\frac{2}{42}} = \sqrt{\frac{1}{21}} = \Lambda(6)$$
  4. Diffusion Noise Schedule (Ho, Jain, and Abbeel 2020) $$\sigma_{t} = \sqrt{1 - \bar{\alpha}{t}}$$ The value $\bar{\alpha}_{t} = 20/21$ is chosen here. It is not a schedule value derived in that paper: $$\sigma{t} = \sqrt{1 - \frac{20}{21}} = \sqrt{\frac{1}{21}} = \Lambda(6)$$
  5. Echo State input weights (Jaeger 2001 and corrected report) and sample-mean standard error For independent observations, the standard error page gives $\sigma_{\bar x}=\sigma/\sqrt{n}$. The factor $1/\sqrt{n}$ is that standard error only when the population standard deviation is also $1$. It is not the definition of Rademacher complexity. Jaeger's echo-state example sets input weights to $+1$ or $-1$ with equal probability. $1/\sqrt{n}$ is not an input scale prescribed by that report: $$\text{SE}(\bar{X}) = \frac{1}{\sqrt{n}}$$ For $n = 21$ independent zero-mean unit-variance variables: $$\text{SE} = \frac{1}{\sqrt{21}} = \Lambda(6)$$

4.2. Statistics and Probability (Natural Exponential Families with Quadratic Variance Function)

  1. Coefficient of Variation ($\mathrm{CV}$) $$\mathrm{CV} = \frac{\sigma}{\mu}$$
    • Poisson distribution (Poisson distribution, $\lambda = 21$): $\mu = 21$ and $\sigma^{2} = 21$, so $$\mathrm{CV} = \frac{\sqrt{21}}{21} = \frac{1}{\sqrt{21}} = \Lambda(6).$$
    • Skewness of Poisson distribution ($\lambda = 21$): $\gamma_{1} = \frac{1}{\sqrt{\lambda}} = \frac{1}{\sqrt{21}} = \Lambda(6)$
    • Gamma distribution (gamma distribution, $\alpha = 21, \beta = 1$): $\mu = 21$ and $\sigma^{2} = 21$, so $$\mathrm{CV} = \frac{\sqrt{21}}{21} = \frac{1}{\sqrt{21}} = \Lambda(6).$$
    • Chi-Square distribution (chi-squared distribution, $k = 42$): $\mu = 42$ and $\sigma^{2} = 84$, so $$\mathrm{CV} = \frac{\sqrt{84}}{42} = \frac{2\sqrt{21}}{42} = \frac{1}{\sqrt{21}} = \Lambda(6).$$
    • Binomial distribution (binomial distribution, $n = 21, p = 0.5$): $\mu = 10.5$ and $\sigma^{2} = 5.25$, so $$\mathrm{CV} = \frac{\sqrt{5.25}}{10.5} = \frac{1}{\sqrt{21}} = \Lambda(6).$$
    • Erlang distribution (Erlang distribution, $k = 21, \lambda$): $\mu = k/\lambda$ and $\sigma = \sqrt{k}/\lambda$, so $$\mathrm{CV} = \frac{\sqrt{21}/\lambda}{21/\lambda} = \frac{1}{\sqrt{21}} = \Lambda(6).$$
  2. Information Geometry Metrics (Poisson Family)
    • Fisher information for one $\text{Poisson}(\lambda)$ observation (Fisher information): $I(\lambda) = \frac{1}{\lambda}$, so $$\sqrt{I(21)} = \frac{1}{\sqrt{21}} = \Lambda(6).$$ The metric component is $I(\lambda)$, while $\sqrt{I(\lambda)},d\lambda$ is the line element.
    • Jeffreys Prior (Jeffreys prior): $\pi(\lambda) \propto \sqrt{g(\lambda)} = \frac{1}{\sqrt{\lambda}}$, so $$\pi(21) \propto \frac{1}{\sqrt{21}} = \Lambda(6).$$

4.3. Physics and Relativistic Mechanics

  1. Schwarzschild time-dilation factor (gravitational time dilation) For a static exterior observer, proper time and distant Schwarzschild coordinate time satisfy $t_{0}=t_{f}\sqrt{1-r_{s}/r}$. The redshift measured from infinity is the reciprocal, $1+z = 1/\sqrt{1-r_{s}/r}$, so $z$ itself is not that factor (gravitational redshift). At the chosen radius $r = \frac{21}{20} r_{s}$: $$\sqrt{1 - \frac{r_{s}}{r}} = \sqrt{1 - \frac{20}{21}} = \sqrt{\frac{1}{21}} = \Lambda(6).$$ In units $c=r_{s}=1$, writing $\Phi=-1/(2r)=-10/21$ makes $\sqrt{1+2\Phi}=\Lambda(6)$ by algebra. That equality does not make $\Phi$ the Newtonian-limit potential: $r=1.05r_{s}$ is not a weak field. It is not a measured interval near a physical black hole, and it lies inside the photon sphere. The static formula does not describe a circular orbit, whose factor is $\sqrt{1-\frac{3}{2}r_{s}/r}$.
  2. Reciprocal of the Lorentz factor (Lorentz factor) The page defines $\gamma = 1/\sqrt{1-v^{2}/c^{2}}$. The quantity below is $1/\gamma$, not $\gamma$: $$\gamma^{-1} = \sqrt{1 - \frac{v^{2}}{c^{2}}}$$ For velocity $v = \sqrt{\frac{20}{21}} c$: $$\gamma^{-1} = \sqrt{1 - \frac{20}{21}} = \sqrt{\frac{1}{21}} = \Lambda(6)$$
  3. Harmonic Oscillation Period (harmonic oscillator) For an undamped harmonic oscillator $\ddot{x} + \omega^{2} x = 0$ with chosen stiffness-to-mass ratio $\omega^{2} = k/m = 21$: $$T = \frac{2\pi}{\omega} = \frac{2\pi}{\sqrt{21}} = 2\pi \Lambda(6)$$
  4. Molecular Vibrational Degrees of Freedom (molecular vibration) For non-linear molecules with $N$ atoms, degrees of freedom are $3N - 6$. For $N = 9$: $$3(9) - 6 = 27 - 6 = 21$$
  5. Adiabatic index (heat capacity ratio) $$\gamma = 1 + \frac{2}{f}$$ For effective internal degrees of freedom $f = 2\sqrt{21}$: $$\gamma = 1 + \frac{2}{2\sqrt{21}} = 1 + \frac{1}{\sqrt{21}} = 1 + \Lambda(6)$$

4.4. Engineering, Information Theory, and Kinetics

  1. Second-Order System Percentage Overshoot (overshoot)

    Damping ratio formula:

    $$%OS = \exp\left(-\frac{\pi \zeta}{\sqrt{1 - \zeta^{2}}}\right) \times 100%$$ For damping ratio $\zeta = \Lambda(6) = \frac{1}{\sqrt{21}}$: $$\zeta^{2} = \frac{1}{21} \implies 1 - \zeta^{2} = \frac{20}{21} \implies \sqrt{1 - \zeta^{2}} = \frac{\sqrt{20}}{\sqrt{21}}$$ $$\frac{\zeta}{\sqrt{1-\zeta^{2}}} = \frac{1/\sqrt{21}}{\sqrt{20}/\sqrt{21}} = \frac{1}{\sqrt{20}}$$ $$%OS = \exp\left(-\frac{\pi}{\sqrt{20}}\right) \times 100% = 49.5354568\ldots%,$$ displayed as $49.54%$. The intermediate value $\exp(-\pi/\sqrt{20})=0.495354568\ldots$ is not $0.70248$.

  2. Quality Factor (Q factor) $$Q = \frac{1}{2\zeta}$$ For $\zeta = \Lambda(6) = \frac{1}{\sqrt{21}}$: $$Q = \frac{1}{2(1/\sqrt{21})} = \frac{\sqrt{21}}{2} \approx 2.2913$$

  3. Fixed-length alphabet ceiling (prefix code) and Hash Table Load Factor (hash table)

    • A fixed-length code of $k$ bits encodes at most $2^{k}$ symbols, so $D = \lceil \log_{2} N \rceil$ is that ceiling, not the Huffman weighted path length (Huffman coding). For $N = 21$ symbols: $$\log_{2}(21) \approx 4.3923 \implies \lceil 4.3923 \rceil = 5\text{ bits}$$
    • Hash Table Load Factor: $\alpha = n/m$. Choosing $m = 21$ buckets and $n = \sqrt{21}$ stored elements gives $\alpha = \Lambda(6)$. A real table has an integer element count: $$\alpha = \frac{\sqrt{21}}{21} = \frac{1}{\sqrt{21}} = \Lambda(6)$$
  4. First-Order Reaction Rate and Half-Life (rate equation) $$k = \frac{\ln 2}{t_{1/2}}$$ For half-life $t_{1/2} = \Lambda(6) = \frac{1}{\sqrt{21}}$: $$k = \frac{\ln 2}{1/\sqrt{21}} = \ln 2 \cdot \sqrt{21} \approx 3.1764$$

  5. Michaelis-Menten Enzyme Kinetics (Michaelis and Menten 1913, English translation) and Bioavailability

    • Normalized reaction velocity: $$\frac{v}{V_{\max}} = \frac{[S]}{K_{m} + [S]} = \frac{[S]/K_{m}}{1 + [S]/K_{m}}$$ Setting the substrate ratio to $[S]/K_{m} = 1/\sqrt{21}$ does not make the velocity ratio equal $\Lambda(6)$: $$\frac{v}{V_{\max}} = \frac{1/\sqrt{21}}{1 + 1/\sqrt{21}} = \frac{1}{\sqrt{21} + 1} \approx 0.179129.$$
    • Absolute bioavailability (bioavailability) is the dose-normalized ratio of extravascular to intravenous area under the concentration curve. Setting that fraction to $\Lambda(6)$ gives $F = 1/\sqrt{21} \approx 21.82%$. This is an input choice, not a measured drug value.
    • Bazett rate correction (Bazett 1920), in dimensionally consistent form: $$\mathrm{QTc} = \frac{\mathrm{QT}}{\sqrt{\mathrm{RR}/1,\mathrm{s}}}.$$ Choosing $\mathrm{RR} = 1/\sqrt{21},\mathrm{s}$ is an input substitution. It is not a measured cardiac interval, and a bare $\mathrm{QT}/\sqrt{\mathrm{RR}}$ is dimensionally inconsistent.
  6. Clinical Trial Sample Size per Group (Cohen, Statistical Power Analysis, 2nd ed.) $$n = \frac{2(z_{\alpha} + z_{\beta})^{2}}{d^{2}}$$ For Cohen's effect size $d = \Lambda(6) = \frac{1}{\sqrt{21}}$, it follows that $d^{2} = \frac{1}{21}$: $$n = 2 \cdot 21 (z_{\alpha} + z_{\beta})^{2} = 42(z_{\alpha} + z_{\beta})^{2}$$

  7. Kelly Criterion Fraction (Kelly criterion and Kelly 1956) The cited page's binary formula is $f = p/l - q/g$. Kelly's 1956 paper does not write that formula. Its even-money calculation gives the same fraction, $\ell = 2q - 1$, after swapping its win and loss labels. For even money, $g = l = 1$, so $$f = p - q = 2p - 1.$$ Setting that edge to $\Lambda(6) = 1/\sqrt{21}$ gives $$p = \frac{1 + \Lambda(6)}{2} = \frac{1 + 1/\sqrt{21}}{2} \approx 0.609109.$$ The odds form $(bp - q)/b$ is this case only when the whole stake is lost and the net gain on a win is $b$. It is not the formula displayed by either cited source.

  8. Transmission-Line Reflection and Impedance Ratio (reflection coefficient) The cited page displays the load reflection coefficient $\Gamma=(Z_{L}-Z_{0})/(Z_{L}+Z_{0})$, not an acoustic pressure-amplitude formula. Its linked acoustic section instead defines $R$ as the ratio of reflected to incident intensity, written there as $R=p_{\mathrm{reflected}}/p_{\mathrm{incident}}$, and then uses $\alpha=1-R^{2}$. For a real ratio $z=Z_{L}/Z_{0}$, $$\Gamma = \frac{z - 1}{z + 1}.$$ For $\Gamma = \Lambda(6) = \frac{1}{\sqrt{21}}$: $$z = \frac{1 + \Gamma}{1 - \Gamma} = \frac{\sqrt{21} + 1}{\sqrt{21} - 1} = \frac{(\sqrt{21}+1)^{2}}{20} = \frac{22 + 2\sqrt{21}}{20} = \frac{11 + \sqrt{21}}{10} \approx 1.558258$$

  9. Planetary, Geophysics, and Neutrino Metrics

    • Kepler's third law in solar units (Kepler's laws), where the primary has one solar mass: $T^{2} = a^{3}$ with $T$ in years and $a$ in AU. For orbital period $T = 21$ yr: $$a = 21^{2/3} = \sqrt[3]{441} \approx 7.61166\text{ AU}$$
    • Gutenberg-Richter law (Gutenberg-Richter law) with chosen baseline parameters $a = 5.0$, $b = 1.0$, and chosen cumulative event rate $N = 1/\sqrt{21}$: $$M = \frac{a - \log_{10}(1/\sqrt{21})}{b} = 5.0 + \log_{10}(\sqrt{21}) = 5.0 + 0.6611 = 5.6611 \approx 5.66$$
    • Effective neutrino number: Akita and Yamaguchi 2020 calculate $N_{\mathrm{eff}} = 3.044$, with numerical and mixing error at most $0.0005$. The numerical proximity $\ln(21) \approx 3.04452$ is not an identity and is not the definition of $N_{\mathrm{eff}}$.
  10. Game Theory and Economics (Cournot Equilibrium and First-Price Auction)

  • Cournot Oligopoly (Cournot 1838 and linear identical-firm case Marker): for inverse demand $p = a-bQ$ and common marginal cost $c$, each equilibrium output is $q_i = (a-c)/((n+1)b)$. For $n = 20$, $q_i/((a-c)/b) = 1/21$.
  • First-Price Sealed-Bid Auction (first-price sealed-bid auction): the page states the symmetric Bayesian Nash equilibrium $b(v) = \frac{n-1}{n} v$ for valuations that are i.i.d. uniform on $[0,1]$. Its general symmetric BNE is $E[y_i \mid y_i<v_i]$. For $n = 21$ bidders, the uniform strategy is $b(v) = \frac{20}{21} v$.
  • Herd Immunity Threshold (herd immunity): $p_c = 1 - \frac{1}{R_{0}}$ under homogeneous mixing, solid immunity, no immune escape, and no nonhuman vector. For $R_{0} = 21$, that threshold is $\frac{20}{21}$.
  • Autoregressive Time Series AR(1) (autoregressive model): for $|\phi|<1$, stationary variance $\mathrm{Var}(X_{t}) = \frac{\sigma_{\varepsilon}^{2}}{1 - \phi^{2}}$. For $\phi = \frac{1}{\sqrt{21}}$, $\mathrm{Var}(X_{t}) = \frac{21}{20} \sigma_{\varepsilon}^{2}$.