betaFeedback

WorksMathematics

Irrationality measures for Γ(1/4)Γ(1/4) and Γ(1/3)Γ(1/3)

Abstract

Let μμ denote the irrationality exponent. Bruiltet proved in 2002 that μ(Γ(1/4))≤10143μ(Γ(1/4))\le 10^{143} and μ(Γ(1/3))≤10151μ(Γ(1/3))\le 10^{151}. We prove μ(Γ(1/4))≤3.24⋅1026μ(Γ(1/4))\le 3.24\cdot 10^{26} and μ(Γ(1/3))≤1.74⋅1026μ(Γ(1/3))\le 1.74\cdot 10^{26}, with explicit thresholds h(p/q)≥5.4⋅107h(p/q)\ge 5.4\cdot 10^{7} and h(p/q)≥4.7⋅107h(p/q)\ge 4.7\cdot 10^{7} respectively, against 107510^{75} and 107910^{79} in the published statements. The proof keeps the architecture of Bruiltet's argument – Philippon's transcendence criterion together with a zero estimate in commutative algebraic groups – and replaces three of its ingredients. First, an approximate Schwarz principle based on Hermite interpolation replaces the gravitation estimate of Reyssat; on a lattice point set its cost per node is governed by a lattice energy and drops from about 3.7 to about 1.0. Second, a summation lemma for the failure of Diaz's approximation lemma sharpens the passage from an approximation measure to an irrationality measure. Third, the auxiliary point is taken at a torsion point of order 4, respectively 3, which lowers the degree of the base field from 8d to 2d. All numerical constants are produced by a verifier run in interval arithmetic, supplied as ancillary files.

Introduction

For an irrational real number θ\theta, the irrationality exponent μ(θ)\mu(\theta) is the supremum of the set of real μ\mu such that 0<∣θ−p/q∣<max⁡(∣p∣,∣q∣)−μ0 < |\theta- p/q| < \max(|p|, |q|)^{-\mu} has infinitely many solutions. Finiteness of μ(Γ(1/4))\mu(\Gamma(1/4)) and μ(Γ(1/3))\mu(\Gamma(1/3)) follows from the algebraic independence of π/ω\pi/\omega and η/ω\eta/\omega for a lattice with algebraic invariants, due to Chudnovsky, together with the quantitative refinements of Philibert and Philippon. The first explicit values are due to Bruiltet [1]:

h(p/q)≥1075⟹∣Γ(1/4)−pq∣>(1qe)10143,h(p/q) \ge10^{75} \Longrightarrow\left|\Gamma(1/4)-\frac{p}{q}\right| > \left(\frac{1}{qe}\right)^{10^{143}},
h(p/q)≥1079⟹∣Γ(1/3)−pq∣>(1qe)10151.h(p/q) \ge10^{79} \Longrightarrow\left|\Gamma(1/3)-\frac{p}{q}\right| > \left(\frac{1}{qe}\right)^{10^{151}}.

These remain, to our knowledge, the only published bounds for these two numbers; Cohen and Zudilin [2] obtain reasonable measures for Chowla–Selberg gamma quotients and describe the above as the state of the art for the individual values.

Our result is the following.

Theorem 1. For every rational p/qp/q with h(p/q)≥5.4⋅107h(p/q) \ge5.4 \cdot10^7,

∣Γ(1/4)−pq∣>(1qe)3.24⋅1026,\left|\Gamma(1/4)-\frac{p}{q}\right| > \left(\frac{1}{qe}\right)^{3.24 \cdot10^{26}},

and for every rational p/qp/q with h(p/q)≥4.7⋅107h(p/q) \ge4.7 \cdot10^7,

∣Γ(1/3)−pq∣>(1qe)1.74⋅1026.\left|\Gamma(1/3)-\frac{p}{q}\right| > \left(\frac{1}{qe}\right)^{1.74 \cdot10^{26}}.

In particular μ(Γ(1/4))≤3.24⋅1026\mu(\Gamma(1/4)) \le3.24 \cdot10^{26} and μ(Γ(1/3))≤1.74⋅1026\mu(\Gamma(1/3)) \le1.74 \cdot10^{26}.

The architecture is that of [1]: an approximation measure for the point θ=(1:π/ω:η/ω)∈P2(C)\theta=(1:\pi/\omega:\eta/\omega)\in\mathbb{P}^{2}(\mathbb{C}) is obtained from Theorem 1 of [7], the rank of the interpolation matrix being controlled by Theorem 9 of [6]; the approximation measure is then converted into an irrationality measure for Γ\Gamma. We keep this skeleton and change three of its ingredients, in § 3, § 5 and § 6.

Throughout, PROVED marks a statement proved here or quoted from the literature, and VERIFIED marks a numerical statement produced by the certified verifier of § 8, together with the range over which it was checked.

Notation

Ω=Zω1+Zω2\Omega=\mathbb{Z}\omega_{1}+\mathbb{Z}\omega_{2} is a lattice of C\mathbb{C} with ℑ(ω2/ω1)>0\Im(\omega_{2}/\omega_{1})>0 and algebraic invariants g2,g3g_{2},g_{3}; ℘\wp, ζ\zeta, σ\sigma are the associated Weierstrass functions, η1,η2\eta_{1},\eta_{2} the quasi-periods, and Vol⁡(Ω)\operatorname{Vol}(\Omega) the area of a fundamental domain. For a nonzero period ω\omega we write η\eta for the associated quasi-period, and

θ1=π/ω,θ2=η/ω,θ=(1:θ1:θ2).\theta_{1}=\pi/\omega,\qquad\theta_{2}=\eta/\omega,\qquad\theta=(1:\theta_{1}:\theta_{2}).

For x∈Qmx\in\mathbb{Q}^{m}, d(x)d(x) is the degree of Q(x)\mathbb{Q}(x) and h(x)h(x) the absolute logarithmic height; M(P)M(P) is the Mahler measure of P∈Z[X]P\in\mathbb{Z}[X] and μ(P)=log⁡M(P)\mu(P)=\log M(P). For ff analytic and r>0r>0, M0(f,r)=sup⁡∣z∣≤r∣f(z)∣M_{0}(f,r)=\sup_{|z|\le r}|f(z)|. Finally Dist is the projective distance of [8].

We use throughout the exact growth constant

B∗=π2Vol⁡(Ω)(1)B^{*}=\frac{\pi}{2\operatorname{Vol}(\Omega)} \tag*{(1)}

which is the value of Bruiltet’s B(Ω)=12(∣η1/ω1∣+2π/Vol⁡(Ω))B(\Omega)=\frac{1}{2}(|\eta_{1}/\omega_{1}|+2\pi/\operatorname{Vol}(\Omega)) when the Legendre relation is used to eliminate η1/ω1\eta_{1}/\omega_{1} in the lemniscatic and hexagonal cases.

An approximate Schwarz principle by Hermite interpolation

Theorem 1 of [7] rests on an approximate Schwarz principle in which the quantity Gr,r′(S)G_{r,r'}(S) measures the “gravitation” of the interpolation nodes. In the elliptic applications of [8] and [1], GG is bounded by the estimate of Reyssat [9], namely (66R/(δM))N−1(66R/(\delta\sqrt{M}))^{N-1}, which is valid for an arbitrary separated set of nodes. When the nodes form a piece of a lattice, the following substitution is available and is substantially sharper.

For 0<r′<r0<r'<r and a finite set S={y1,…,yN}S=\{y_{1},\ldots,y_{N}\} of pairwise distinct points of ∣z∣≤r′−1|z|\le r'-1, put, following [7 §2.a],

Er,r′(S)=(r′r)N∏k=1N1+∣yk∣/r′1+∣yk∣r′/r2.(2)E_{r,r'}(S)=\left(\frac{r'}{r}\right)^{N}\prod_{k=1}^{N}\frac{1+|y_{k}|/r'}{1+|y_{k}|r'/r^{2}}. \tag*{(2)}

Lemma 2 (PROVED). Let T≥1T\ge1 and let ff be analytic on ∣z∣≤r|z|\le r. Assume ∣f(t)(yk)/t!∣≤ε|f^{(t)}(y_{k})/t!|\le\varepsilon for all kk and all t<Tt<T. Then

M0(f,r′)≤Er,r′(S)TM0(f,r)+εΓ,Γ=∑k,t(M0(ℓk,t,r′)+Er,r′(S)TM0(ℓk,t,r)),M_{0}(f,r')\le E_{r,r'}(S)^{T}M_{0}(f,r)+\varepsilon\Gamma,\qquad\Gamma=\sum_{k,t}\left(M_{0}(\ell_{k,t},r')+E_{r,r'}(S)^{T}M_{0}(\ell_{k,t},r)\right),

where (ℓk,t)1≤k≤N, 0≤t<T(\ell_{k,t})_{1\le k\le N,\,0\le t<T} is the Hermite basis attached to SS and TT.

Proof. Let HH be the Hermite interpolant of ff at the yky_{k} to order TT, so that H=∑k,tf(t)(yk)t!ℓk,tH=\sum_{k,t}\frac{f^{(t)}(y_{k})}{t!}\ell_{k,t} and M0(H,ρ)≤ε∑k,tM0(ℓk,t,ρ)M_{0}(H,\rho)\le\varepsilon\sum_{k,t}M_{0}(\ell_{k,t},\rho) for every ρ\rho. Put φk(z)=r(z−yk)/(r2−yk‾z)\varphi_{k}(z)=r(z-y_{k})/(r^{2}-\overline{y_{k}}z) and B=∏kφkTB=\prod_{k}\varphi_{k}^{T}, so that ∣B∣=1|B|=1 on ∣z∣=r|z|=r. The function (f−H)/B(f-H)/B is analytic on ∣z∣≤r|z|\le r, because f−Hf-H vanishes to order at least TT at each yky_{k}. By the maximum principle, ∣f−H∣≤∣B∣(M0(f,r)+M0(H,r))|f-H|\le|B|(M_{0}(f,r)+M_{0}(H,r)) on ∣z∣≤r|z| \le r. Finally, for ∣z∣=r′|z| = r' and each kk, the function z↦∣z−yk∣/∣r2−yˉkz∣z \mapsto|z-y_k|/|r^2-\bar{y}_kz| attains its maximum at z=−r′yk/∣yk∣z=-r'y_k/|y_k|, whence

∣φk(z)∣≤r(r′+∣yk∣)r2+∣yk∣r′=r′r⋅1+∣yk∣/r′1+∣yk∣r′/r2.|\varphi_k(z)| \le\frac{r(r' + |y_k|)}{r^2 + |y_k|r'} = \frac{r'}{r}\cdot\frac{1 + |y_k|/r'}{1 + |y_k|r'/r^2}.

and therefore ∣B∣≤Er,r′(S)T|B| \le E_{r,r'}(S)^T on ∣z∣=r′|z| = r'. Combining the three displays gives the statement.

Remark 3. The product in (2) carries a ++ sign in the denominator precisely because the maximum above is attained at the antipodal point; this is the point at which Lemma 2 matches the quantity Er,r′E_{r,r'} already present in [7].

We now bound Γ\Gamma when SS is a finite subset of a lattice Ω\Omega, contained in a disc of radius R0R_0 and such that N=#SN = \#S. Write λ=r/r′\lambda= r/r' and r′=R0+1r' = R_0 + 1.

Proposition 4 (PROVED). Let d1≤d2≤⋯d_1 \le d_2 \le\cdots be the moduli of the nonzero vectors of Ω\Omega arranged in increasing order, and set L=∑i≤N−1log⁡diL = \sum_{i \le N-1} \log d_i. Then

log⁡Γ≤T⋅max⁡(log⁡Er,r′(S)−1+Xr, Xr′)+2log⁡(2r)+log⁡1ς+ςS1+ς2S2,\log\Gamma\le T \cdot\max\left(\log E_{r,r'}(S)^{-1} + X_r,\ X_{r'}\right) + 2\log(2r) + \log\frac{1}{\varsigma} + \varsigma S_1 + \varsigma^2 S_2,

where, for ρ∈{r,r′}\rho\in\{r,r'\}, Xρ=∑y∈Slog⁡(ρ+∣y∣)−LX_\rho= \sum_{y \in S} \log(\rho+ |y|) - L, where ς=d1/2\varsigma= d_1/2, and where S1=∑i≤N−1di−1S_1 = \sum_{i \le N-1} d_i^{-1}, S2=∑i≤N−1di−2S_2 = \sum_{i \le N-1} d_i^{-2}.

Proof. Fix kk. For every j≠kj \ne k the point yk−yjy_k-y_j is a nonzero vector of Ω\Omega, and these N−1N-1 vectors are pairwise distinct; hence ∑j≠klog⁡∣yk−yj∣≥L\sum_{j \ne k}\log|y_k-y_j| \ge L by rearrangement. Consequently, for ∣z∣=ρ|z| = \rho,

∣∏j≠k(z−yj)T(yk−yj)T∣≤exp⁡(T(∑jlog⁡(ρ+∣yj∣)−L))=eTXρ,\left|\prod_{j \ne k}\frac{(z-y_j)^T}{(y_k-y_j)^T}\right| \le\exp\left(T\left(\sum_j \log(\rho+ |y_j|)-L\right)\right) = e^{TX_\rho},

uniformly in kk. The Hermite basis element ℓk,t\ell_{k,t} is this product times (z−yk)t/t!(z-y_k)^t/t! times the truncation at order T−tT-t of the reciprocal of the product, expanded at yky_k; bounding the latter by Cauchy’s formula on the circle of radius ς=d1/2\varsigma= d_1/2 centred at yky_k, which encloses no other node, together with −log⁡(1−x)≤x+x2-\log(1-x) \le x+x^2 for 0≤x≤1/20 \le x \le1/2, legitimate since ς/di≤1/2\varsigma/d_i \le1/2, produces the three correction terms log⁡(1/ς)\log(1/\varsigma), ςS1\varsigma S_1 and ς2S2\varsigma^2 S_2. Summation over kk and t<Tt < T contributes at most log⁡(NT)≤log⁡(2r)\log(NT) \le\log(2r) to each of the two circles, whence the factor 2log⁡(2r)2\log(2r).

Remark 5. In the verifier, LL, S1S_1 and S2S_2 are evaluated by exact enumeration of the lattice vectors of modulus at most 200ω1200\omega_1; the tail is bounded in closed form by di≥(i+1)Vol⁡(Ω)/π−diam⁡(Ω)d_i \ge\sqrt{(i+1)\operatorname{Vol}(\Omega)/\pi}-\operatorname{diam}(\Omega). The sum ∑ylog⁡(ρ+∣y∣)\sum_y \log(\rho+ |y|) is bounded cell by cell on a 1000×10001000 \times1000 grid; each cell is evaluated at its unfavourable corner, and its point count is bounded by its area thickened by diam⁡(Ω)\operatorname{diam}(\Omega), divided by Vol⁡(Ω)\operatorname{Vol}(\Omega).

The approximation measure

Let z0z_0 be a torsion point of C/Ω\mathbb{C}/\Omega, say nz0=ω′∈Ωnz_0 = \omega' \in\Omega, and let ω\omega be a nonzero period. Set θ1=π/ω\theta_1 = \pi/\omega, θ2=η/ω\theta_2 = \eta/\omega.

Theorem 6 (PROVED; constants verified). With the notation of §2 there is a real A>0A > 0, effectively computable from Ω\Omega, ω\omega and z0z_0, such that for every α=(1:α1:α2)∈P2(Q)\alpha= (1:\alpha_1:\alpha_2) \in\mathbb{P}_2(\mathbb{Q}),

log⁡Dist⁡(θ,α)>−A(h(α)+log⁡(1+d(α)))d(α)3/2log⁡(1+d(α)).\log\operatorname{Dist}(\theta,\alpha) > -A\left(h(\alpha)+\log(1+d(\alpha))\right)d(\alpha)^{3/2}\sqrt{\log(1+d(\alpha))}.

The proof applies Theorem 1 of [7] in its original form. We give the data, the verification of its constraints, and the resulting inequality.

Data

Put K0=Q(α1,α2,g2,℘(z0),℘′(z0))K_0 = \mathbb{Q}(\alpha_1, \alpha_2, g_2, \wp(z_0), \wp'(z_0)) and a=[K0:Q]a = [K_0 : \mathbb{Q}]. For the two lattices of §5 one has Q(g2,℘(z0),℘′(z0))=Q(i)\mathbb{Q}(g_2, \wp(z_0), \wp'(z_0)) = \mathbb{Q}(i), so that a≤2d(α)a \leq2d(\alpha); at the 3-division point of [1] the 3-division polynomial 3x4−6x2−13x^4 - 6x^2 - 1 is irreducible over Q\mathbb{Q} and ℘′\wp' adds a square root, so that a≤8d(α)a \leq8d(\alpha) there. Set

f1=σ2℘,f2=σ⋅(ζ−ηωz),f3=σ,f_1 = \sigma^2\wp,\qquad f_2 = \sigma\cdot\left(\zeta-\frac{\eta}{\omega}z\right),\qquad f_3 = \sigma,

three entire functions. For D1,D2∈N∗D_1,D_2 \in\mathbb{N}^* take the monomials

D={h∈N3:h1≤D1, h2≤D2, h3=2(D1−h1)+D2−h2},L=#D=(D1+1)(D2+1).\mathcal{D} = \{h \in\mathbb{N}^3 : h_1 \leq D_1,\ h_2 \leq D_2,\ h_3 = 2(D_1-h_1)+D_2-h_2\},\qquad L = \#\mathcal{D} = (D_1+1)(D_2+1).

Every monomial fhf^h is σ2D1+D2℘h1(ζ−(η/ω)z)h2\sigma^{2D_1+D_2}\wp^{h_1}\left(\zeta-(\eta/\omega)z\right)^{h_2}, so the three functions enter with bounded exponents and a constant total weight in σ\sigma.

The point set is G⊂Ω\mathcal{G} \subset\Omega, the intersection of Ω\Omega with a disc of radius ρ\rho centred at the origin, truncated by the strip ∣ℑz∣≤ρsin⁡φ|\Im z| \leq\rho\sin\varphi parallel to the rows Zω\mathbb{Z}\omega; MM denotes the number of rows met, N=#GN = \#\mathcal{G}, and R0=sup⁡s∈G∣z0+s∣R_0 = \sup_{s\in\mathcal{G}} |z_0+s|. With integers T0<TT_0 < T we take

S1={0,…,T0−1}×(z0+(G+Zω)/Zω),Si={T0+i−2}×(z0+G)(2≤i≤T−T0+1),S_1 = \{0,\ldots,T_0-1\}\times(z_0+(\mathcal{G}+\mathbb{Z}\omega)/\mathbb{Z}\omega),\qquad S_i = \{T_0+i-2\}\times(z_0+\mathcal{G})\quad(2\leq i\leq T-T_0+1),
Si′={0,…,T0−1}×(z0+G),r′=R0+1,r=λr′.S_i' = \{0,\ldots,T_0-1\}\times(z_0+\mathcal{G}),\qquad r'=R_0+1,\qquad r=\lambda r'.

The parameters D1,D2,T0,T,M,λ,φD_1,D_2,T_0,T,M,\lambda,\varphi are free, subject to the two conditions of §4.5 and to

L≥rL#S1=rLT0M,rL>1,q:=#S1L−#S1=1rL−1.(3)L \geq r_L\#S_1 = r_LT_0M,\qquad r_L>1,\qquad q:=\frac{\#S_1}{L-\#S_1}=\frac{1}{r_L-1}. \tag*{(3)}

Growth

By the estimates of [5] in the form used in [1 Lemme 1], ϖ3(r)=A0+B∗r2\varpi_3(r)=A_0+B^*r^2 bounds log⁡M0(σ,r)\log M_0(\sigma,r) with B∗B^* as in (1). Since σ2℘\sigma^2\wp is entire and ℘=−(log⁡σ)′′\wp=-(\log\sigma)'',

ϖ1(r)=A1∗+2B∗r2,A1∗=2A0+log⁡(∣℘(z0)∣+e2B∗∣z0∣2∣σ(z0)∣2),\varpi_1(r)=A_1^*+2B^*r^2,\qquad A_1^*=2A_0+\log\left(|\wp(z_0)|+\frac{e^{2B^*|z_0|^2}}{|\sigma(z_0)|^2}\right),

and ϖ2(r)=ϖ3(r)+log⁡(1+∣η/ω∣r+∣σ(z0)∣−1)\varpi_2(r)=\varpi_3(r)+\log(1+|\eta/\omega|r+|\sigma(z_0)|^{-1}) bounds f2f_2. The exponents of D\mathcal{D} give (6) below.

The matrix of approximations

Following the device of Chudnovsky used in [8], one expands zz and ζ\zeta in the local parameter u=(℘−℘(z0))/℘′(z0)u=(\wp-\wp(z_0))/\wp'(z_0) at z0z_0: with γ−1=ζ(z0)−(η/ω)z0\gamma_{-1}=\zeta(z_0)-(\eta/\omega)z_0 and γ0=0\gamma_0=0,

℘=℘(z0)+℘′(z0)u,ζ−ηωz=γ−1−∑i≥1(γiηω+γi℘(z0)+γi−1℘′(z0))uii,\wp=\wp(z_0)+\wp'(z_0)u,\qquad\zeta-\frac{\eta}{\omega}z=\gamma_{-1}-\sum_{i\geq1}\left(\gamma_i\frac{\eta}{\omega}+\gamma_i\wp(z_0)+\gamma_{i-1}\wp'(z_0)\right)\frac{u^i}{i},

with H(γ−1,…,γi)≤c3iH(\gamma_{-1},\ldots,\gamma_i)\leq c_3^i. The quasi-periodicity and the Legendre relation give, for s=s1ω1+s2ω2s=s_1\omega_1+s_2\omega_2 and ω=a1ω1+a2ω2\omega=a_1\omega_1+a_2\omega_2,

ζ(z+s)−ηω(z+s)=2i(a1s2−a2s1)πω+ζ(z)−ηωz,\zeta(z+s)-\frac{\eta}{\omega}(z+s)=\frac{2i(a_1s_2-a_2s_1)\pi}{\omega}+\zeta(z)-\frac{\eta}{\omega}z,

so that the columns indexed by ss and s′s' with a1(s2−s2′)=a2(s1−s1′)a_1(s_2-s_2')=a_2(s_1-s_1') coincide: the number of distinct columns is the number MM of rows. Each entry of the matrix is a polynomial Qh,(t,s)Q_{h,(t,s)} in θ1,θ2\theta_1,\theta_2 with coefficients in Q(g2,℘(z0),℘′(z0))\mathbb{Q}(g_2,\wp(z_0),\wp'(z_0)), of degree at most D2D_2 and of height at most c4(D1+Tlog⁡D2+D2log⁡(TM))c_4(D_1+T\log D_2+D_2\log(TM)). The vectors of approximations are obtained by specialising θ\theta at α\alpha, with λi+1,⋅=σ(z0+s)2D1+D2\lambda_{i+1,\cdot}=\sigma(z_0+s)^{2D_1+D_2} and the Leibniz coefficients μt′,t,s\mu_{t',t,s} of [8 (5)], and MDM_{\mathcal{D}} denotes the resulting matrix.

Verification of the constraints

Write δ(i)=[Ki+1:K0]s=1\delta(i)=[K_{i+1}:K_0]_s=1 for all ii, which is CδC_\delta. Put P∞=max⁡(1,∣℘′(z0)∣)P_\infty=\max(1,|\wp'(z_0)|) and h(P)=log⁡P∞h(P)=\log P_\infty.

Cr,r′C_{r,r'}: ∣Si+1∣≤R0<r′=R0+1≤r<R=+∞|S_{i+1}|\le R_0<r'=R_0+1\le r<R=+\infty, and Hr′(Si+1)=max⁡(1,(r′−∣y∣)−∣t∣)=1H_{r'}(S_{i+1})=\max(1,(r'-|y|)^{-|t|})=1 because r′−∣y∣≥1r'-|y|\ge1 on the support. This is why r′r' is taken one unit larger than R0R_0.

CN,rC_{N,r}: card⁡Si′≤T0N\operatorname{card}S'_i\le T_0N and ∣Si′∣≤R0≤r|S'_i|\le R_0\le r.

CEC_E: since n=1n=1 the parasitic factor of [7] disappears and E(i)≥Er,r′(Si′)Hr′(Si+1)=Er,r′(G)T0E(i)\ge E_{r,r'}(S'_i)H_{r'}(S_{i+1})=E_{r,r'}(\mathcal{G})^{T_0}, so that by (2)

−log⁡E(i)≥T0(Nlog⁡λ−∑y∈Glog⁡1+∣y∣/r′1+∣y∣r′/r2)(4)-\log E(i)\ge T_0\left(N\log\lambda-\sum_{y\in\mathcal{G}}\log\frac{1+|y|/r'}{1+|y|r'/r^2}\right) \tag*{(4)}

the sum being bounded cell by cell as in Proposition 4, and NN from below by the row-by-row count. The gravitation Gr,r′(Si′)G_{r,r'}(S'_i) is bounded by Lemma 2 and Proposition 4, which is where the present argument departs from [9].

CϕC_\phi: by the height estimate of the Qh,(t,s)Q_{h,(t,s)} above,

ϕ(j)=T(log⁡120+12h(P)+log⁡D2)+D1(log⁡4+h(P))+D2h(α)+c8D2log⁡(TM),(5)\phi(j)=T\left(\log120+\frac{1}{2}h(P)+\log D_2\right)+D_1\left(\log4+h(P)\right)+D_2h(\alpha)+c_8D_2\log(TM), \tag*{(5)}

and ϕ(0)\phi(0) is the same expression with TT replaced by 11, since the index 0 involves only derivatives of order <T0<T_0.

CϱC_\varrho: with cℓ=−log⁡∣σ(z0)∣c_\ell=-\log|\sigma(z_0)| and ℓp=log⁡30+12log⁡P∞\ell_p=\log30+\frac{1}{2}\log P_\infty, ℓq=log⁡4+log⁡P∞\ell_q=\log4+\log P_\infty,

ϱ(j)=max⁡(Tℓp+D1ℓq,2D1cℓ),\varrho(j)=\max(T\ell_p+D_1\ell_q,2D_1c_\ell),

the second branch coming from ∣λi+1,⋅∣−1≤∣σ(z0)∣−(2D1+D2)|\lambda_{i+1,\cdot}|^{-1}\le|\sigma(z_0)|^{-(2D_1+D_2)}, for which one uses that ∣σ(z0+s)∣≥∣σ(z0)∣|\sigma(z_0+s)|\ge|\sigma(z_0)| for every s∈Gs\in\mathcal{G}, a consequence of the functional equation of σ\sigma on the quadrant containing G\mathcal{G}.

Growth: with the exponents of D\mathcal{D},

Dϖ(r)=(2D1+D2)ϖ3(r)+D1(ϖ1−2ϖ3)+(r)+D2(ϖ2−ϖ3)+(r).(6)\mathcal{D}\varpi(r)=(2D_1+D_2)\varpi_3(r)+D_1(\varpi_1-2\varpi_3)^+(r)+D_2(\varpi_2-\varpi_3)^+(r). \tag*{(6)}

U~\widetilde U: the passage from U=−log⁡Dist⁡(θ,α)U=-\log\operatorname{Dist}(\theta,\alpha) to the quantity U(j)U(j) of CUC_U costs

V1=2D1(A0+B∗(R0+1)2)+(c6+log⁡2)T+Tℓp+D1ℓq,V_1=2D_1\left(A_0+B^*(R_0+1)^2\right)+(c_6+\log2)T+T\ell_p+D_1\ell_q,

where c6=log⁡(2∣z0∣−1max⁡(24∣℘′(z0)∣−1∣z0∣−2,1))c_6=\log\left(2|z_0|^{-1}\max(24|\wp'(z_0)|^{-1}|z_0|^{-2},1)\right) bounds the Leibniz coefficients μt′,t,s\mu_{t',t,s}, together with log⁡Γ\log\Gamma from Lemma 2.

Rank of the matrix

If MDM_{\mathcal{D}} were not of rank #D\#\mathcal{D}, a nontrivial linear combination of its rows would give a nonzero P∈C[X,Y]P\in\mathbb{C}[X,Y] of bidegree ≤(D1,D2)\le(D_1,D_2) vanishing to order ≥T\ge T along the analytic subgroup exp⁡G(z,α2z)\exp_{\mathcal{G}}(z,\alpha_2z) of the extension G\mathcal{G} of EE by Ga\mathbb{G}_a, at all the points of Σ\Sigma, where Σ\Sigma is indexed by the rows. By Theorem 9 of [6] there is a proper algebraic subgroup G′\mathcal{G}' with

T2card⁡Σ+G′G′≤29deg⁡φG‾⋅D11−dim⁡π(G′)D21−dim⁡(G′∩Ga),\frac{T}{2}\operatorname{card}\frac{\Sigma+\mathcal{G}'}{\mathcal{G}'}\le2^9\deg_\varphi\overline{\mathcal{G}}\cdot D_1^{1-\dim\pi(\mathcal{G}')}D_2^{1-\dim(\mathcal{G}'\cap\mathbb{G}_a)},

and 29deg⁡φG‾=22⋅(G∞+3π∗D)2=4⋅6=242^9\deg_\varphi\overline{\mathcal{G}}=2^2\cdot(G_\infty+3\pi^*\mathcal{D})^2=4\cdot6=24, the minimum of deg⁡φa,bG‾=6ab\deg_{\varphi_{a,b}}\overline{\mathcal{G}}=6ab over a,b≥1a,b\ge1.

For G′=Ga\mathcal{G}'=\mathbb{G}_a all points of Σ\Sigma have the same image in EE, so the cardinal is 11; for G′={0}\mathcal{G}'=\{0\} it is at least [M/2]+1[M/2]+1. Hence the two conditions

T>24⋅3D1,TM>25⋅3D1D2(7)T>2^4\cdot3D_1,\qquad TM>2^5\cdot3D_1D_2 \tag*{(7)}

make the inequality impossible, and MDM_{\mathcal{D}} has rank #D\#\mathcal{D}. The hypothesis 1≤δ≤Δ1\le\delta\le\Delta of that theorem reads D2≤D1D_2\le D_1 here, and is recorded in § 7.

Conclusion

Theorem 1 of [7] now gives an index j≥1j \ge1 with

min⁡(−log⁡E(j)−Dϖ(r), U(j))≤aϕ(j)+ϱ(j)+Δj+q(aϕ(0)+ϱ(0)+Δ0).\min\left(-\log E(j)-D\varpi(r),\,U(j)\right) \le a\phi(j)+\varrho(j)+\Delta_j+q\left(a\phi(0)+\varrho(0)+\Delta_0\right).

The parameters are chosen so that the first term does not realise the minimum; writing T0=τhT_0=\tau h with h=h(α)h=h(\alpha), this amounts to

τ=aD2(1+q)Q,Q=1T0(−log⁡E−Dϖ(r)−aϕ(j)−ϱ(j)−q(aϕ(0)+ϱ(0))).\tau=\frac{aD_2(1+q)}{Q}, \qquad Q=\frac{1}{T_0}\left(-\log E-D\varpi(r)-a\phi(j)-\varrho(j)-q\left(a\phi(0)+\varrho(0)\right)\right).

all the terms of QQ being independent of hh once the D2h(α)D_2h(\alpha) part of ϕ\phi has been set aside. The minimum is then realised by U(j)U(j), whence

U≤T0(−log⁡E−Dϖ(r)T0+V1+log⁡ΓT0)=Ah+O(1),U\le T_0\left(\frac{-\log E-D\varpi(r)}{T_0}+\frac{V_1+\log\Gamma}{T_0}\right)=Ah+O(1),

with AA obtained by optimising M,D2,rL,λ,ϕM,D_2,r_L,\lambda,\phi at each degree. Replacing hh by h+log⁡(1+d)h+\log(1+d) and tracking the dependence in dd through a≤2da\le2d gives the shape d3/2log⁡(1+d)d^{3/2}\sqrt{\log(1+d)} of the statement. The O(1)O(1) is made explicit in § 7.

The two instances

Γ(1/4)\Gamma(1/4)

Take y2=4x3−4xy^2=4x^3-4x, so that Ω=Zω1+Ziω1\Omega=\mathbb{Z}\omega_1+\mathbb{Z}i\omega_1 with ω1=Γ(1/4)2/8π\omega_1=\Gamma(1/4)^2/\sqrt{8\pi}, and η1=π/ω1\eta_1=\pi/\omega_1. Take ω=ω1\omega=\omega_1 and the 4-torsion point z0=(ω1+ω2)/4z_0=(\omega_1+\omega_2)/4. Then verified

℘(z0)=−i,℘′(z0)=2(1+i),℘′′(z0)=−8,∣σ(z0)∣=0.938398432669…\wp(z_0)=-i,\qquad\wp'(z_0)=2(1+i),\qquad\wp''(z_0)=-8,\qquad|\sigma(z_0)|=0.938398432669\ldots

Writing ω′=4z0\omega'=4z_0 and γ′−1=ζ(z0)−η′/4\gamma'^{-1}=\zeta(z_0)-\eta'/4, the duplication formula ζ(2z)=2ζ(z)+℘′′(z)/(2℘′(z))\zeta(2z)=2\zeta(z)+\wp''(z)/(2\wp'(z)) and ζ(ω′/2)=η′/2\zeta(\omega'/2)=\eta'/2 give γ′−1=−℘′′(z0)/(4℘′(z0))=(1−i)/2\gamma'^{-1}=-\wp''(z_0)/(4\wp'(z_0))=(1-i)/2, and the Legendre relation gives

γ−1=ζ(z0)−η1ω1z0=1−i2−i2θ1.(8)\gamma^{-1}=\zeta(z_0)-\frac{\eta_1}{\omega_1}z_0=\frac{1-i}{2}-\frac{i}{2}\theta_1. \tag*{(8)}

The algebraic part (1−i)/2(1-i)/2 has minimal polynomial 4z4+14z^4+1, hence height 14log⁡4=log⁡2\frac{1}{4}\log4=\log\sqrt{2}, and lies in K=Q(i,α)K=\mathbb{Q}(i,\alpha); the term −i2θ1-\frac{i}{2}\theta_1 shifts the integer index of 2iπ(a1s2−a2s1)/ω2i\pi(a_1s_2-a_2s_1)/\omega by one half, and clearing that denominator costs D2log⁡2D_2\log2. The base field is K=Q(i,α1,α2)K=\mathbb{Q}(i,\alpha_1,\alpha_2), of degree at most 2d(α)2d(\alpha) instead of 16d(α)16d(\alpha).

Γ(1/3)\Gamma(1/3)

Take y2=4x3−1y^2=4x^3-1, so that Ω=Zω1+Zρω1\Omega=\mathbb{Z}\omega_1+\mathbb{Z}\rho\omega_1 with ρ=e2iπ/3\rho=e^{2i\pi/3}, ω1=Γ(1/3)3/(2π)\omega_1=\Gamma(1/3)^3/(2\pi) and η1=2π/(3 ω1)\eta_1=2\pi/(\sqrt{3}\,\omega_1). Take ω=ω1\omega=\omega_1 and the 3-torsion point z0=(2ω1+ω2)/3z_0=(2\omega_1+\omega_2)/3. Then verified

℘(z0)=0,℘′(z0)=±i,∣σ(z0)∣=eπ3/9=1.83051946656…\wp(z_0)=0,\qquad\wp'(z_0)=\pm i,\qquad|\sigma(z_0)|=e^{\pi\sqrt{3}/9}=1.83051946656\ldots

the last value agreeing with the closed form of [1 Lemme 2]. Here the Frobenius–Stickelberger relation gives (γ′−1)2=℘(z0)/3=0(\gamma'^{-1})^2=\wp(z_0)/3=0, so that

γ−1=−2i3θ1.(9)\gamma^{-1}=-\frac{2i}{3}\theta_1. \tag*{(9)}

with no algebraic part at all; clearing the denominator costs D2log⁡3D_2\log3. Moreover h(P)=log⁡max⁡(1,∣℘(z0)∣,∣℘′(z0)∣)=0h(P)=\log\max(1,|\wp(z_0)|,|\wp'(z_0)|)=0.

From an approximation measure to an irrationality measure

Let θ\theta be the real number under study (Γ(1/4)\Gamma(1/4) or Γ(1/3)\Gamma(1/3)), and let p/qp/q be a rational with h=h(p/q)h=h(p/q).

The conversion coefficient

Put y=2π/Γ(1/4)y=\sqrt{2\pi}/\Gamma(1/4) for the lattice of y2=4x3−4xy^2=4x^3-4x and Y=3π/Γ(1/3)2Y=\sqrt{3}\pi/\Gamma(1/3)^2 for that of y2=4x3−1y^2=4x^3-1. With ω=ω1\omega=\omega_1 one has

(1:πω1:η1ω1)=(1:y3Γ(1/4):2y4),(1:πω1:η1ω1)=(1:23y2Γ(1/3):89y3),(10)\left(1:\frac{\pi}{\omega_1}:\frac{\eta_1}{\omega_1}\right)=\left(1:y^3\Gamma(1/4):2y^4\right),\qquad\left(1:\frac{\pi}{\omega_1}:\frac{\eta_1}{\omega_1}\right)=\left(1:\frac{2}{3}y^2\Gamma(1/3):\frac{8}{9}y^3\right), \tag*{(10)}

as follows from ω1=Γ(1/4)2/8π\omega_1=\Gamma(1/4)^2/\sqrt{8\pi}, η1=π/ω1\eta_1=\pi/\omega_1 in the first case and ω1=Γ(1/3)3/(2π)\omega_1=\Gamma(1/3)^3/(2\pi), η1=2π/(3ω1)\eta_1=2\pi/(\sqrt{3}\omega_1) in the second, together with the Legendre relation. In both displays Γ\Gamma occurs to the first power only. Approximating yy (resp. YY) by an algebraic β\beta of degree δ\delta and height τ\tau, and θ\theta by p/qp/q, gives α∈P2(Q)\alpha\in\mathcal{P}_2(\mathbb{Q}) with d(α)≤δd(\alpha)\leq\delta and

h(α)≤cτδ+h+O(1),c=4 (Γ(1/4)),c=3 (Γ(1/3)),h(\alpha)\leq c\frac{\tau}{\delta}+h+O(1),\qquad c=4\ (\Gamma(1/4)),\qquad c=3\ (\Gamma(1/3)),

cc being the largest power of yy, resp. YY, occurring in (10). This is the constant cc of Lemma 8.

Lemma 7 (Dirichlet, real case; PROVED). Let 0<β<10<\beta<1, let d≥1d\geq1, put S=∑i≤d∣θ∣iS=\sum_{i\leq d}|\theta|^i and assume

t≥log⁡(2d+1)+1dlog⁡S1−β.t\geq\frac{\log(2\sqrt{d+1})+\frac{1}{d}\log S}{1-\beta}.

Then there is a nonzero P∈Z[X]P\in\mathbb{Z}[X] with deg⁡P≤d\deg P\leq d, log⁡M(P)≤t\log\mathrm{M}(P)\leq t and log⁡∣P(θ)∣≤−βdt\log|P(\theta)|\leq-\beta dt.

Proof. Put H=⌊et/d+1⌋H=\lfloor e^{t}/\sqrt{d+1}\rfloor. The (H+1)d+1(H+1)^{d+1} polynomials with coefficients in {0,…,H}\{0,\ldots,H\} take at θ\theta values lying in a real interval of length at most HSHS, so two of them, say P1≠P2P_1\ne P_2, take values at distance at most HS/((H+1)d+1−1)≤SH−dHS/((H+1)^{d+1}-1)\leq SH^{-d}. Put P=P1−P2P=P_1-P_2. Then ∥P∥2≤d+1H≤et\|P\|_2\leq\sqrt{d+1}H\leq e^t, so M(P)≤∥P∥2≤et\mathrm{M}(P)\leq\|P\|_2\leq e^t by Landau’s inequality, and log⁡∣P(θ)∣≤log⁡S−dlog⁡H≤−βdt\log|P(\theta)|\leq\log S-d\log H\leq-\beta dt by the hypothesis on tt.

Only the one-dimensionality of the pigeonhole is used; this is what allows β\beta up to 11, where the complex case of [4 Lemme 1] is limited to β<1/2\beta<1/2.

We next feed Lemma 7 into Lemma 2 of [4], which splits PP either into a power of an irreducible factor carrying at least 2β/32\beta/3 of the weight, or into two coprime factors. On the second branch we use Proposition 3.5 of [10] directly, with the sum constraints deg⁡F+deg⁡G≤d\deg F+\deg G\leq d and μ(F)+μ(G)≤t\mu(F)+\mu(G)\leq t supplied by Lemma 2 of [4]; since ∣Res⁡(F,G)∣≥1|\operatorname{Res}(F,G)|\geq1, 2mn≤2d2/42^{mn}\leq2^{d^2/4} and nμ(F)+mμ(G)≤dtn\mu(F)+m\mu(G)\leq dt, this yields, for every integer s≥1s\geq1,

log⁡ρ≤−4s2(β3s−1−log⁡24dt)dt,(11)\log\rho\leq-\frac{4}{s^2}\left(\frac{\beta}{3}s-1-\frac{\log2}{4}\frac{d}{t}\right)dt, \tag*{(11)}

ρ\rho being the distance from θ\theta to the roots of FGFG. On the first branch we use the lemma of Diaz [3], in the form of Proposition 6 of [1].

The following lemma is what makes the two branches quantitative. Let S1S_1 and S2S_2 denote the running maxima, over degrees at most DD, of the slope of Theorem 6 and of that slope divided by the degree; both are nondecreasing and satisfy S1(d)≤dS2(d)S_1(d)\leq dS_2(d). Write t=mht=mh and let cc be the coefficient of hh in the conversion (c=4c=4 for Γ(1/4)\Gamma(1/4), c=3c=3 for Γ(1/3)\Gamma(1/3)).

Lemma 8 (PROVED). Let ∏jQjej\prod_j Q_j^{e_j} be a group of factors, the QjQ_j irreducible of degree δj\delta_j and height τj\tau_j, with ∑jejδj≤xD\sum_j e_j\delta_j\leq xD and ∑jejτj≤yt\sum_j e_j\tau_j\leq yt. If the lemma of Diaz fails on every QjQ_j, then the normalised weight ϖ=(∑jejwj)/(Dt)\varpi=(\sum_j e_jw_j)/(Dt), where wj=−log⁡∣Qj(θ)∣w_j=-\log|Q_j(\theta)|, satisfies

ϖ≤Φ(x,y)=min⁡s≥1{xys+s+12S2(xD)(xm+cyD)}.\varpi\leq\Phi(x,y)=\min_{s\geq1}\left\{\frac{xy}{s}+\frac{s+1}{2}S_2(xD)\left(\frac{x}{m}+\frac{cy}{D}\right)\right\}.

Proof. Failure at QjQ_j for a given ss reads wj≤δjτj/s+s+12(hS1(δj)+cτjS2(δj))w_j \leq\delta_j\tau_j/s + \frac{s+1}{2}(hS_1(\delta_j)+c\tau_jS_2(\delta_j)). Multiply by eje_j, sum over jj, and use

∑jejδjτj≤(max⁡jτj)∑jejδj≤(yt)(xD),∑jejS1(δj)≤∑jejδjS2(δj)≤xDS2(xD),\sum_j e_j\delta_j\tau_j \leq(\max_j\tau_j)\sum_j e_j\delta_j \leq(yt)(xD), \qquad\sum_j e_jS_1(\delta_j) \leq\sum_j e_j\delta_jS_2(\delta_j) \leq xD S_2(xD),
∑jejτjS2(δj)≤ytS2(xD).\sum_j e_j\tau_jS_2(\delta_j) \leq ytS_2(xD).

the last two because S2S_2 is nondecreasing and δj≤xD\delta_j \leq xD. Dividing by DtDt and taking the minimum over ss gives the statement.

The configuration of factors is then controlled cell by cell on the square 0≤x,y≤10 \leq x,y \leq1, each cell being evaluated at its unfavourable corner, using the three constraints of Lemma 2 of [4]: ϖF≥β/3\varpi_F \geq\beta/3, ϖG≥β/3\varpi_G \geq\beta/3 and ϖF+ϖG≥β\varpi_F+\varpi_G \geq\beta on the coprime branch, and ϖQ≥2β/3\varpi_Q \geq2\beta/3 together with ϖQ≥β−Φ\varpi_Q \geq\beta-\Phi of the complement on the power branch.

Threshold and effective statements

Theorem 6 is used in its asymptotic form in hh, in which D1D_1 and TT are proportional to hh while D2D_2, MM and the shape parameters are fixed. The terms discarded are therefore exactly those independent of hh. Since the exponents of a monomial of DD satisfy h3=2(D1−h1)+D2−h2h_3=2(D_1-h_1)+D_2-h_2, the growth term is

Dϖ(r)=(2D1+D2)ϖ3(r)+D1(ϖ1−2ϖ3)+(r)+D2(ϖ2−ϖ3)+(r),(12)D\varpi(r)=(2D_1+D_2)\varpi_3(r)+D_1(\varpi_1-2\varpi_3)^{+}(r)+D_2(\varpi_2-\varpi_3)^{+}(r), \tag*{(12)}

where ϖ1,ϖ2,ϖ3\varpi_1,\varpi_2,\varpi_3 bound f1,f2,f3f_1,f_2,f_3; only the terms carrying D1D_1 are linear in hh. The discarded part of (12) is D2max⁡(ϖ2,ϖ3)(r)D_2\max(\varpi_2,\varpi_3)(r), and since QT0=aD2(1+q)hQT_0=aD_2(1+q)h at the optimum, D2D_2 cancels and its relative weight is

max⁡(ϖ2,ϖ3)(r)a(1+q)h.\frac{\max(\varpi_2,\varpi_3)(r)}{a(1+q)h}.

The remaining discarded terms are ac8D2log⁡(TM)ac_8D_2\log(TM) coming from ϕ\phi, with c8=7+log⁡c1+h(P)c_8=7+\log c_1+h(P), that is c8=8.04c_8=8.04 for Γ(1/4)\Gamma(1/4) and c8=7c_8=7 for Γ(1/3)\Gamma(1/3) by § 5; c9D2log⁡Tc_9D_2\log T coming from ϱ\varrho; the terms in M2M^2 of U~\widetilde{U}; and Δj+Δ0\Delta_j+\Delta_0. All of these are smaller than the previous one at the optimum. Requiring the total to be at most 0.25%0.25\%, which inflates the exponent by about 1%1\%, and adding the hypothesis D2≤D1D_2\leq D_1 of § 4(iv), the threshold of Lemma 7 and the term log⁡24t\frac{\log2}{4t} of (11), gives

H0=5.4⋅107(Γ(1/4)),H0=4.7⋅107(Γ(1/3)).H_0=5.4\cdot10^7\quad(\Gamma(1/4)), \qquad H_0=4.7\cdot10^7\quad(\Gamma(1/3)).

the binding constraint being in both cases the D2D_2 part of (12). The other constraints are inactive by at least two orders of magnitude; those of Lemma 7 and of (11) by more than ten.

Certification

The numerical part consists of two stages.

The slope tables of Theorem 6 are computed over 3489 cells covering 1≤d≤10301\leq d\leq10^{30}, geometrically spaced with ratio 1.02. Within each cell the free parameters – number of rows, D2D_2, the ratio L/#S1L/\#S_1, the radius ratio λ\lambda and the shape parameter of the region – are optimised in floating point, and the resulting value is then re-evaluated in interval arithmetic; only the interval upper bound is retained, inflated by a further factor 1+10−121+10^{-12}. Each table entry is the running maximum, which is what Lemma 8 requires.

The conversion is verified in extended precision (80-bit, ϵ≈10−19\epsilon\approx10^{-19}) with an outward relative margin of 10−910^{-9}, which makes every test strictly harder and dominates the accumulated rounding error; the same verification in double precision returns identical verdicts. The final value is assembled in interval arithmetic. Minimality is checked by confirming that the binding criterion fails at 0.99999 times the optimum.

The ancillary files contain the verifier and the tables. VERIFIED outputs:

μ(Γ(1/4))≤3.2011⋅1026,μ(Γ(1/3))≤1.7158⋅1026,\mu(\Gamma(1/4)) \le3.2011 \cdot10^{26}, \qquad\mu(\Gamma(1/3)) \le1.7158 \cdot10^{26},

to which the 1% inflation of § 7 is applied in Theorem 1.

Acknowledgements

Technical assistance: Claude (Anthropic).

References

  1. [1]S. Bruiltet, D’une mesure d’approximation simultanée à une mesure d’irrationalité: le cas de Γ(1/4) et Γ(1/3), Acta Arith. 104 (2002), 243–281.
  2. [2]H. Cohen, W. Zudilin, Continued fractions and irrationality measures for Chowla–Selberg gamma quotients, arXiv:2510.00215.arxiv.org/abs/2510.00215
  3. [3]G. Diaz, Une nouvelle propriété d’approximation diophantienne, C. R. Acad. Sci. Paris Sér. I 324 (1997), 969–972.DOI
  4. [4]M. Laurent, D. Roy, Criteria of algebraic independence with multiplicities and approximation by hypersurfaces, Ann. Inst. Fourier 49 (1999), 27–55.
  5. [5]D. W. Masser, Sharp estimates for Weierstrass elliptic functions, J. Anal. Math. 90 (2003), 257–302.DOI
  6. [6]P. Philippon, Nouveaux lemmes de zéros dans les groupes algébriques commutatifs, Rocky Mountain J. Math. 26 (1996), 1069–1088.DOI
  7. [7]P. Philippon, Une approche méthodique pour la transcendance et l’indépendance algébrique de valeurs de fonctions analytiques, J. Number Theory 64 (1997), 291–338.DOI
  8. [8]P. Philippon, Mesures d’approximation de valeurs de fonctions analytiques, Acta Arith. 88 (1999), 113–127.DOI
  9. [9]E. Reyssat, Approximation algébrique de nombres liés aux fonctions elliptiques et exponentielle, Bull. Soc. Math. France 108 (1980), 47–79.DOI
  10. [10]D. Roy, M. Waldschmidt, Approximation diophantienne et indépendance algébrique de logarithmes, Ann. Sci. École Norm. Sup. 30 (1997), 753–796.DOI

Paper details

vlnc:W_8bnpm948
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Claude

View statement
Looking for related papers…
Contents