Mahler measures of the body-centred cubic lattice family

edit · history · discussion · files · long url · graph theory Mahler measure statistical mechanics spanning forests integral special values number theory
Numbers
$k$ 
$M(k)=m(Q_k)$
8:
1.990191418271940771710519085433364992945360003470834199902047571563144838824765844743331831627484518
9:
2.135101201802818589374850176686903602983073045506476248592719228399715831413899489151700537967494445
10:
2.255142142310768420913396585621522023955826241644495851332577607138732800999338195184661068760148344
11:
2.360101391113999141797527662855175725089826872265276114390418117506333197685881401748967981785769050
12:
2.453941905201338002594885379297690009960119673442614667035861907159514622704156008056977952628670148
13:
2.539045788994498838160861190343529453281497925683761019812536019238470166856458776197626667645585006
14:
2.617030336231851464907961572547776439611572916141908522574777975980838693443598131714231910665900684
15:
2.689068565581389129520279221870177323472328018221299729601806343828136390286912381574913284170503016
16:
2.756048554162093571750665616011339507588759257339744258135960802413771259796861674549563543705218464
17:
2.818663657340571610701680162147205884796496696182618483143417125710724865796013192300620637877305219
18:
2.877468069306622919600710328539593474381996849081654956305619300255604627024617388636115961073917471
19:
2.932913124803214202253711808472192852918753585671200793961290974066231021890299626347839465159635621
20:
2.985372091093149058560420933372524294041843741619708314622885032467643658099517880830813563602742797
21:
3.035157695251908346414049198453871198917977637143531802114564301878486171009961353687869307658283746
22:
3.082534866445229944485486259547530738246572722895442191440550732490612882097432996207907104821943595
23:
3.127730216311558718656093933855902284261137000762885315268521233348585299726630475342542759876435614
24:
3.170939231750245776778685468090864535280462941694890444199562525739096482256168514345323914340949254
25:
3.212331824582242345685375728405819073743416326058257896355480526268682933309199083661933138888420887
26:
3.252056676569148931715965814989702054160416016797728109404927298859930220170059827606922684062020517
27:
3.290244685414650949527087441775333675465582726341398318738598385268289550213563288033796713365950440
28:
3.327011729270134449679230596549416019784554191810397562309568579550783510614491939971894436988338201
29:
3.362460907427095459387994974176580942552388289024258037850871200561844542423724780422343920470343410
30:
3.396684373373100711671676709675418464423045804381537748765912826377400812056165487745541562734517993
31:
3.429764847058356428196158588610725732278840328296355910303520032175116350392096273906683848646801932
32:
3.461776872147687412099849828162660439544047347852054885028968126006164626567671111669512242321566764
Definition
For real $k\geq8$, let $Q_k(X,Y,Z)=k-(X+X^{-1})(Y+Y^{-1})(Z+Z^{-1})$. Its ordinary logarithmic Mahler measure [3] is $M(k)=m(Q_k)=\pi^{-3}\int_{[0,\pi]^3}\log(k-8\cos\theta_1\cos\theta_2\cos\theta_3)\,d\theta_1d\theta_2d\theta_3$, with natural logarithms and Haar probability measure [1].
Parameters
$k$
—   constant coefficient ($k\geq8$)
Formulas
(1)
$\displaystyle M(k)=\log k-\frac12\sum_{n=1}^{\infty}\frac{a_n}{n}\left(\frac{64}{k^2}\right)^n,\qquad a_n=\left(\frac{\binom{2n}{n}}{4^n}\right)^3.$ The series converges for every $k\geq8$; the attached generator uses geometric tails for $k>8$ and a transformed series at the edge.
(2)
$\displaystyle M(k)=\log k-\frac4{k^2}\,{}_5F_4\!\left(1,1,\frac32,\frac32,\frac32;2,2,2,2;\frac{64}{k^2}\right)$ [1], equations (20) and (31).
(3)
$\displaystyle\begin{aligned}kM^\prime(k)&=W(k)=\frac4{\pi^2}K(\alpha(k))^2,\\ \alpha(k)&=\frac{1-\sqrt{1-64/k^2}}2,\qquad k>8.\end{aligned}$ Here $W(k)=\pi^{-3}\int_{[0,\pi]^3}(1-8\cos\theta_1\cos\theta_2\cos\theta_3/k)^{-1}\,d^3\theta$, and the complete elliptic integral uses the parameter convention $K(u)=\int_0^{\pi/2}(1-u\sin^2 v)^{-1/2}dv$. All square roots are positive real. Also $M(k)-\log k\to0$ as $k\to\infty$ [1].
(4)
$\displaystyle M(k)=\log k-\int_0^{\alpha(k)}\left(\frac4{\pi^2}K(u)^2-1\right)\frac{1-2u}{2u(1-u)}\,du.$ Here $\alpha(k)$ and $K(u)$ are defined in the derivative formula; the integrand has the removable limit $1/4$ at $u=0$ and this identity includes $k=8$. It follows by the substitution $64/k^2=4u(1-u)$ in the logarithmic-moment integral.
(5)
$\displaystyle M(k)=\frac1{\pi^2}\int_0^\pi\int_0^\pi\log\!\left(\frac{k+\sqrt{k^2-64\cos^2 s\cos^2 t}}2\right)\,ds\,dt.$ This follows by integrating out the third torus angle; the square root is nonnegative.
Comments
(6)
The BCC graph has the eight steps $(\pm1,\pm1,\pm1)$ from the origin; its vertices in these coordinates have all three coordinates of the same parity. Its adjacency operator $A$ has Fourier symbol $8\cos\theta_1\cos\theta_2\cos\theta_3$ and its combinatorial Laplacian is $\Delta=8I-A$. Normalization is per vertex. For $k>8$, $M(k)$ is the infinite-volume logarithmic determinant per vertex of $\Delta+(k-8)I$ over growing compatible periodic quotients.
(7)
The integral has its ordinary convergent meaning at $k=8$: the logarithmic singularities at the isolated torus zeros are integrable. No regularization is taken. The value $M(8)$ is the BCC spanning-tree entropy [1], equation (32); its finite-volume interpretation uses a reduced determinant, since the full zero-mass determinant vanishes.
(8)
For positive mass $\mu=k-8$, the finite-graph determinant $\det(\Delta+\mu I)$ is the spanning rooted-forest partition function with weight $\mu$ per component, obtained by scaling the matrix-forest theorem [2]. At $k=9$ every rooted forest has unit weight. The lattice limit follows from the displayed Fourier symbol.
(9)
In [1], equations (18)–(20) and (31), the conversions are $M(k)=T_3^{\mathrm{bcc}}(8/k)=p(k^2)=\tfrac12 f_2(k^2)$. The plus sign in the source pencil is equivalent by $X\mapsto-X$, which preserves Haar measure. Negative coefficients have the same measure by torus symmetry; the present domain selects $k\geq8$. Values inside the spectral band are outside this table.
(10)
In the intermediate series of [1], equation (31), the printed power $z^l$ must read $z^{2l}$: odd cosine moments vanish. The integral and the final hypergeometric expression use the even powers.
Programs
(P1)
Python
import mpmath as mp

def M(k, working=150):
    # Elliptic K uses parameter u. Quadrature here is heuristic.
    with mp.workdps(working):
        k = mp.mpf(k)
        if not mp.isfinite(k) or k < 8:
            raise ValueError("A finite real k >= 8 is required.")
        alpha = (1-mp.sqrt(1-64/k**2))/2
        def integrand(u):
            if u == 0:
                return mp.mpf(1)/4
            return ((2*mp.ellipk(u)/mp.pi)**2-1)*(1-2*u)/(2*u*(1-u))
        answer = mp.log(k)-mp.quad(integrand, [0, alpha/2, alpha])
        return mp.nstr(answer, working-20)

# This reference selection is the integer range stored in the table.
for k in range(8, 33):
    print(k, M(k))
# Compare higher working precision before trusting new quadrature values.
# The attached generator instead uses explicit, rigorous series-tail bounds.
References
[1]
Anthony J. Guttmann and Mathew D. Rogers, Spanning tree generating functions and Mahler measures, Journal of Physics A 45 (2012), 494001. (arXiv) (doi)
[2]
Pavel Chebotarev and Elena Shamis, Matrix-Forest Theorems, arXiv version 2 (2023), section 3, Theorem 5. (arXiv)
Links
Similar tables
Entropy constants of lattice models —   holds spanning-tree entropies for other named lattices; $M(8)$ is the body-centred cubic counterpart
Laplacian polynomials of connected graphs —   holds finite-graph polynomials $\lambda(G,x)=\det(xI-\Delta)$; $(-1)^{|V(G)|}\lambda(G,-\mu)=\det(\Delta+\mu I)$ gives the finite determinants underlying $M(8+\mu)$
Watson integrals of the cubic lattices —   holds the BCC Green endpoint $W(8)$ in its row $(A^*,3)$; it equals the right limit of $kM^\prime(k)$
Mahler measures of the simple-cubic Laurent family —   uses the additive six-term simple-cubic kernel, whereas this pencil uses the product giving eight BCC steps
Body-centred cubic lattice Green function —   holds the normalized diagonal Green function $W(k)$ of the same eight-step BCC operator; $kM^\prime(k)=W(k)$ for $k>8$, and $M(k)-\log k\to0$ as $k\to\infty$ fixes the integration constant
Data properties
Entries are of type: real number
Table is complete: no (it holds every exact integer coefficient $8\leq k\leq32$; $8$ is the intrinsic spectral edge and $9$ gives the unweighted rooted-forest determinant; $10\leq k\leq32$ is exploratory reference coverage of primitive integral pencils, equivalently graph masses $2$ through $24$, reaching four times the spectral edge; the cutoff $32$ is a noncanonical low-height reference extent independent of numeral notation; no completeness over the real domain is claimed)
How they were obtained:

The generator stores 100 significant digits from Arb enclosures with explicit omitted-tail bounds. For k>8 it sums the positive central-binomial moment series and bounds the remaining measure correction by x^(N+1)/(2(N+1)(1-x)), where x=64/k^2. The endpoint uses an absolutely convergent series in the elliptic parameter 1/2. Its exact dyadic coefficient convolution and tail proof are attached in numerical-evidence.md. Exact rational inputs, ball logarithms and propagated rounding errors are included; the weakest computed ball retained 406 accuracy bits, over 122 decimal digits.

more

A separate implementation using mpmath elliptic K and quadrature at 140 and 185 working decimal digits agrees with every one of the 25 recorded intervals. Direct two-angle integration of the original Laurent kernel also checks k=8,9,10,12,16,24,32 to roughly twelve decimal places. Nonzero known-answer and deliberately incorrect factor, power and elliptic-convention controls are retained. These quadrature checks are heuristic corroboration; the explicit series enclosures support the proven precision. The source's endpoint decimal differs at its last printed place and is used only as a limited external check, not as a 100-digit oracle. Executable checks and measured results are attached.