Chiral Potts logarithmic integral and Mahler measure

edit · history · discussion · files · short url · Mahler measure statistical mechanics Potts model integral physics special values number theory
Numbers
$t$
$\nu$ 
$A_\nu(t)$
2
$J(t)$:
0.2530456905667143128553792873726296466860288288059323997259725244138002291344267469954246134358164086
2
$\mu(t)=m(P_t)$:
2.332487232246550241107075651747159350912529231886698162088012552893981095043510893813014594425072471
17/8
$J(t)$:
0.4547831516883800373248310893546381549690612188686173799137127117411874050655708628599901492444947422
17/8
$\mu(t)=m(P_t)$:
2.534224693368215965576527453729167859195561621949383142275752740221368270974655009677580130233750805
9/4
$J(t)$:
0.5700461598093963737459094306047453303130833595010805221602651060582142437654915877478479566590412330
9/4
$\mu(t)=m(P_t)$:
2.649487701489232301997605794979275034539583762581846284522305134538395109674575734565437937648297296
19/8
$J(t)$:
0.6627086724724962661014194915236357718161624128496225618333159705095871261046899020827520268656831947
19/8
$\mu(t)=m(P_t)$:
2.742150214152332194353115855898165476042662815930388324195355998989767992013774048900342007854939257
5/2
$J(t)$:
0.7423481440691796280926979138355758656762266651694937884858519472959812300236288624205776510415632068
5/2
$\mu(t)=m(P_t)$:
2.821789685749015556344394278210105569902727068250259550847891975776162095932713009238167632030819269
21/8
$J(t)$:
0.8130742352139213166206602598625632069022663163810356775752207110963914767283451191282197443360657025
21/8
$\mu(t)=m(P_t)$:
2.892515776893757244872356624237092911128766719461801439937260739576572342637429265945809725325321765
11/4
$J(t)$:
0.8771462195125857112621347209083758708882095963522830015104823473328925532361954546867380795355498019
11/4
$\mu(t)=m(P_t)$:
2.956587761192421639513831085282905575114709999433048763872522375813073419145279601504328060524805865
23/8
$J(t)$:
0.9359797232769296149019184889270512190219189810588763566707446209305107328227224029452080187596955155
23/8
$\mu(t)=m(P_t)$:
3.015421264956765543153614853301580923248419384139642119032784649410691598731806549762797999748951578
3
$J(t)$:
0.9905379714247187056357123607383809063691285343239080562456848440852259839767649859203799999849412445
3
$\mu(t)=m(P_t)$:
3.069979513104554633887408725112910610595628937404673818607724872565406849885849132737969980974197307
25/8
$J(t)$:
1.041514847483881929973961580850937788601871249162035892930625574606370668063521731901147499687735564
25/8
$\mu(t)=m(P_t)$:
3.120956389163717858225657945225467492828371652242801655292665603086551533972605878718737480676991627
13/4
$J(t)$:
1.089431620326489690417909365156501403465369122192240284483876302202040879876653695611585752405629484
13/4
$\mu(t)=m(P_t)$:
3.168873162006325618669605729531031107691869525273006046845916330682221745785737842429175733394885547
27/8
$J(t)$:
1.134692687642516787954261603007718328627608719953997571905104958463667683066752545614222631360014400
27/8
$\mu(t)=m(P_t)$:
3.214134229322352716205957967382248032854109123034763334267144986943848548975836692431812612349270462
7/2
$J(t)$:
1.177619872603694328038377672589831406885412808169102358997496541914075630638545677908470902795756498
7/2
$\mu(t)=m(P_t)$:
3.257061414283530256290074036964361111111913211249868121359536570394256496547629824726060883785012560
29/8
$J(t)$:
1.218474634219134700529449825121849091997830376426568173614074054849235609080399448871112244589103372
29/8
$\mu(t)=m(P_t)$:
3.297916175898970628781146189496378796224330779507333935976114083329416474989483595688702225578359434
15/4
$J(t)$:
1.257473056626470013191048941672354681943065433053194860750183959457893920299505490877874319707363042
15/4
$\mu(t)=m(P_t)$:
3.336914598306305941442745306046884386169565836133960623112223987938074786208589637695464300696619104
31/8
$J(t)$:
1.294796314807387985491916035132992152671423719577062872963891902928479263992520777092043615930839986
31/8
$\mu(t)=m(P_t)$:
3.374237856487223913743612399507521856897924122657828635325931931408660129901604923909633596920096048
4
$J(t)$:
1.330598193152717889065371641365149402945785128660723126070010857750875045879531985144677091308363087
4
$\mu(t)=m(P_t)$:
3.410039734832553817317068005739679107172285531741488888432050886231055911788616131962267072297619150
33/8
$J(t)$:
1.365010619390612045545730521793111678329104940820863876617015850178091772890075587798301346164871807
33/8
$\mu(t)=m(P_t)$:
3.444452161070447973797426886167641382555605343901629638979055878658272638799159734615891327154127870
17/4
$J(t)$:
1.398147823741262560252016685245760268205577770268963260718898354036909551988967349381791415978802798
17/4
$\mu(t)=m(P_t)$:
3.477589365421098488503713049620289972432078173349729023080938382517090417898051496199381396968058861
35/8
$J(t)$:
1.430109522196458959227702243118172157613328126152960862000787521336852514115667498572566252262965611
35/8
$\mu(t)=m(P_t)$:
3.509551063876294887479398607492701861839828529233726624362827549817033380024751645390156233252221674
9/2
$J(t)$:
1.460983392130974479374890867398366991323638525517047312495740213559599116350609581750735787011987509
9/2
$\mu(t)=m(P_t)$:
3.540424933810810407626587231772896695550138928597813074857780242039779982259693728568325768001243572
37/8
$J(t)$:
1.490847024991737287399428627035241432440755367900985111133022618220695365513580923095016838517768264
37/8
$\mu(t)=m(P_t)$:
3.570288566671573215651124991409771136667255770981750873495062646700876231422665069912606819507024327
19/4
$J(t)$:
1.519769486068898419870353066782724607185758482348417670814941704439471824604467322447807163365029896
19/4
$\mu(t)=m(P_t)$:
3.599211027748734348122049431157254311412258885429183433176981732919652690513551469265397144354285959
39/8
$J(t)$:
1.547812574588231723992852717606705312849982210837789193052344578009374683829795423688162707357054224
39/8
$\mu(t)=m(P_t)$:
3.627254116268067652244549081981235017076482613918554955414384606489555549738879570505752688346310287
5
$J(t)$:
1.575031852146238365707678377496415320816234818564913370834309482918916292998966927978118836150256848
5
$\mu(t)=m(P_t)$:
3.654473393826074293959374741870945025042735221645679133196349511399097158908051074795708817139512911
41/8
$J(t)$:
1.601477489880325156969720631446058738984527854376355262195636545071827899831468861575014318149845357
41/8
$\mu(t)=m(P_t)$:
3.680919031560161085221416995820588443211028257457121024557676573552008765740553008392604299139101420
21/4
$J(t)$:
1.627194972229773217347905635419004455016352619659858121802696311458088088725954032915564303532673047
21/4
$\mu(t)=m(P_t)$:
3.706636513909609145599601999793534159242853022740623884164736339938268954635038179733154284521929110
43/8
$J(t)$:
1.652225686088236862044889082702400099417957378366303336530209830961397649679004611165831721693039382
43/8
$\mu(t)=m(P_t)$:
3.731667227768072790296585447076929803644457781447069098892249859441578515588088757983421702682295445
11/2
$J(t)$:
1.676607417514895274582953092472605076443571192100508663294997057768688941486145717852057993596507829
11/2
$\mu(t)=m(P_t)$:
3.756048959194731202834649456847134780670071595181274425657037086248869807395229864669647974585763892
45/8
$J(t)$:
1.700374773248128894030420342927418351003762840937239696744831534893906215096574102905200885878029789
45/8
$\mu(t)=m(P_t)$:
3.779816314927964822282116707301948055230263244018005459106871563374087081005658249722790866867285851
23/4
$J(t)$:
1.723559540567853394294544269888922467700388987327398702768569056977997005887430423965127764901483211
23/4
$\mu(t)=m(P_t)$:
3.803001082247689322546240634263452171926889390408164465130609085458177871796514570782717745890739273
47/8
$J(t)$:
1.746190996244619091870255451930261370011911616719782956075579533616927744232924631104776661119707053
47/8
$\mu(t)=m(P_t)$:
3.825632537924455020121951816304791074238412019800548718437619562097108610142008777922366642108963115
6
$J(t)$:
1.768296173159369166508217690979102290602516703542057118248269522845621838091981783601310130928505649
6
$\mu(t)=m(P_t)$:
3.847737714839205094759914055353631994829017106622822880610309551325802704001065930418900111917761711
49/8
$J(t)$:
1.789900091509421363997394938753266622766984190697539582646042192050321012146044292543151395299590753
49/8
$\mu(t)=m(P_t)$:
3.869341633189257292249091303127796326993484593778305345008082220530501878055128439360741376288846816
25/4
$J(t)$:
1.811025960212789675302387178696192874206563503407423114081801926658627565628629614568477177681627939
25/4
$\mu(t)=m(P_t)$:
3.890467501892625603554083543070722578433063906488188876443841955138808431537713761386067158670884002
51/8
$J(t)$:
1.831695353096202750585827232520915070022814518923993593828369673778888602768901106102477098686719547
51/8
$\mu(t)=m(P_t)$:
3.911136894776038678837523596895444774249314922004759356190409702259069468677985252920067079675975610
13/2
$J(t)$:
1.851928363637142677940006883172602844282288129606121352788727023570259653500430893148729398895336838
13/2
$\mu(t)=m(P_t)$:
3.931369905316978606191703247547132548508788532686887115150767052050440519409515039966319379884592901
53/8
$J(t)$:
1.871743741378594332898291266563562132406688859522604333404528599654405297237051293922669969344835196
53/8
$\mu(t)=m(P_t)$:
3.951185283058430261149987630938091836633189262603370095766568628134586163146135440740259950334091258
27/4
$J(t)$:
1.891159012610674903480860156287128628058433470262826616474821478874455384710815700070976308587253741
27/4
$\mu(t)=m(P_t)$:
3.970600554290510831732556520661658332284933873343592378836861507354636250619899846888566289576509803
55/8
$J(t)$:
1.910190587488420564981314828844982867136201924277072251590789934487519581078237422813519513510664964
55/8
$\mu(t)=m(P_t)$:
3.989632129168256493233011193219512571362702327357838013952829962967700446987321569631109494499921027
7
$J(t)$:
1.928853855408753447330790751869737215678025386377721487517822118971973891240906615936880598051899232
7
$\mu(t)=m(P_t)$:
4.008295397088589375582487116244266919904525789458487249879862147452154757149990762754470579041155295
57/8
$J(t)$:
1.947163270185884124171103330711717426413793827056154366501558913411899752457356632733484121359133991
57/8
$\mu(t)=m(P_t)$:
4.026604811865720052422799695086247130640294230136920128863598941892080618366440779551074102348390054
29/4
$J(t)$:
1.965132426330600978268812846099113705261331120556272271280819323867344361129929353914198704021396431
29/4
$\mu(t)=m(P_t)$:
4.044573968010436906520509210473643409487831523637038033642859352347525227039013500731788685010652493
59/8
$J(t)$:
1.982774127545283765491166299079714159371970840966949370255441990818154820701080108393036404644609676
59/8
$\mu(t)=m(P_t)$:
4.062215669225119693742862663454243863598471244047715132617482019298335686610164255210626385633865738
15/2
$J(t)$:
2.000100448385375132768179021587553882072904337962992202153681599987311007341606262661244905037490183
15/2
$\mu(t)=m(P_t)$:
4.079541990065211061019875385962083586299404741043757964515721628467491873250690409478834886026746246
61/8
$J(t)$:
2.017122789903379782022448090493892701898530399302404074504748206742946202314880060350187846356422237
61/8
$\mu(t)=m(P_t)$:
4.096564331583215710274144454868422406125030802383169836866788235223127068223964207167777827345678300
31/4
$J(t)$:
2.033851929978408711169114224661717321779671352525329846532996542542947042926078495464443438898202818
31/4
$\mu(t)=m(P_t)$:
4.113293471658244639420810589036247026006171755606095608895036571023127908835162642282033419887458880
63/8
$J(t)$:
2.050298068938984957371254719771797989924381809349661564139803625286445763413523904534466242308785206
63/8
$\mu(t)=m(P_t)$:
4.129739610618820885622951084146327694150882212430427326501843653766626629322608051352056223298041269
8
$J(t)$:
2.066470871006173909443635600362295555468323578718196379664699798145954866481547569814375859283652406
8
$\mu(t)=m(P_t)$:
4.145912412686009837695331964736825259694823981798962142026739826626135732390631716631965840272908469
65/8
$J(t)$:
2.082379502015587787570514979331761792093876362309844964822649002124595611654342825492732879812033793
65/8
$\mu(t)=m(P_t)$:
4.161821043695423715822211343706291496320376765390610727184689030604776477563426972310322860801289856
33/4
$J(t)$:
2.098032663818404183350079254902172856655170878814066791803225511524176877302641473696378321841512727
33/4
$\mu(t)=m(P_t)$:
4.177474205498240111601775619276702560881671281894832554165265540004357743211725620513968302830768789
67/8
$J(t)$:
2.113438625711571764277287915251325568025138033890979243842090532537345303733806815019363329301922735
67/8
$\mu(t)=m(P_t)$:
4.192880167391407692528984279625855272251638436971745006204130561017526169642890961836953310291178797
17/2
$J(t)$:
2.128605253204491641199682093494360674567474365634598456031883265606306846905767024193934117289048258
17/2
$\mu(t)=m(P_t)$:
4.208046794884327569451378457868890378793974768715364218393923294086487712814851171011524098278304321
69/8
$J(t)$:
2.143540034392540903328307820936776527955940834975817358509593884819919824080426234480469409595408754
69/8
$\mu(t)=m(P_t)$:
4.222981576072376831580004185311306232182441238056583120871633913300100689989510381298059390584664816
35/4
$J(t)$:
2.158250104175920554000175498123408661251921696453996357158394192594617037532878881309354062514162265
35/4
$\mu(t)=m(P_t)$:
4.237691645855756482251871862497938365478422099534762119520434221074797903441963028126944043503418328
71/8
$J(t)$:
2.172742266534696499170921093807123357435494091526031409230188698662992557970783595563110823711218900
71/8
$\mu(t)=m(P_t)$:
4.252183808214532427422617458181653061661994494606797171592228727143173423879867742380700804700474963
9
$J(t)$:
2.187023015046920856626997009012796353248394806883558253008165298541838836478756265034945252713640170
9
$\mu(t)=m(P_t)$:
4.266464556726756784878693373387326057474895209964324015370205327022019702387840411852535233702896233
73/8
$J(t)$:
2.201098551815838267381055703046761510988306942522964952161510785179612059964786428038583688544182053
73/8
$\mu(t)=m(P_t)$:
4.280540093495674195632752067421291215214807345603730714523550813659792925873870574856173669533438116
37/4
$J(t)$:
2.214974804953950961587248222203820038835305647259045024090409073273163635817868911861096181399173842
37/4
$\mu(t)=m(P_t)$:
4.294416346633786889838944586578349743061806050339810786452449101753344501726953058678686162388429905
75/8
$J(t)$:
2.228657444755760987117134501805352058474122400643418027257624778843794336142594871479529902601141320
75/8
$\mu(t)=m(P_t)$:
4.308098986435596915368830866179881762700622803724183789619664807323975202051679018297119883590397383
19/2
$J(t)$:
2.242151898677010940226262014424107891101334034283980870268278231664735026589192808922088450633644407
19/2
$\mu(t)=m(P_t)$:
4.321593440356846868477958378798637595327834437364746632630318260144915892498276955739678431622900470
77/8
$J(t)$:
2.255463365225937193236900650862676307545858124478904259629455431641907848344845772111607683363400070
77/8
$\mu(t)=m(P_t)$:
4.334904906905773121488597015237206011772358527559670021991495460122088714253929918929197664352656132
39/4
$J(t)$:
2.268596826861204035145243959737009636083697709405159863295672956821446951248067776486370696991370362
39/4
$\mu(t)=m(P_t)$:
4.348038368541039963396940324111539340310198112485925625657712985301627817157151923303960677980626425
79/8
$J(t)$:
2.281557061981609272693175231789979465089924855446642059412467414780123608525193900651134677450227277
79/8
$\mu(t)=m(P_t)$:
4.360998603661445200944871596164509169316425258527407821774507443260304474434278047468724658439483340
10
$J(t)$:
2.294348656084176032684659683466758600393068317976184491663938325842578775248264814624705770845449964
10
$\mu(t)=m(P_t)$:
4.373790197764011960936356047841288304619568721056950254025978354322759641157348961442295751834706027
14
$J(t)$:
2.634883211307943144242882981507941449576073720411053621494212172997218012674936633954416149484041586
14
$\mu(t)=m(P_t)$:
4.714324752987779072494579345882471153802574123491819383856252201477398878584020780772006130473297649
322
$J(t)$:
5.774543709132565187160051799472315459193967847434696638750526200488867889762408230224082483109886383
322
$\mu(t)=m(P_t)$:
7.853985250812401115411748163846845163420468250515462401112566228969048755671492377041672464099142445
Definition
For $c_j=\cos\theta_j$ and $h=c_1+c_2+c_3-c_1c_2c_3$, put $J(t)=\pi^{-3}\int_{[0,\pi]^3}\log(t-h)\,d\theta_1d\theta_2d\theta_3$ for real $t\geq2$ [1]. The two tabulated conventions are $A_{\mathrm{potts}}(t)=J(t)$ and $A_{\mathrm{mahler}}(t)=\mu(t)=J(t)+\log8$, the logarithmic Mahler measure of the Laurent polynomial $P_t$ defined in Comments.
Parameters
$t$
—   spectral parameter ($t\geq2$)
$\nu$
—   normalization
Formulas
(1)
$J(2)=\dfrac{8}{\pi}L(2,\chi_{-4})-3\log2,\qquad \mu(2)=\dfrac{8}{\pi}L(2,\chi_{-4})$, where $L(2,\chi_{-4})$ is Catalan's constant and $\chi_{-4}$ is the primitive quadratic character of conductor $4$ [1], equation (2).
(2)
For $t>2$, define $G(t)=J\prime(t)=\pi^{-3}\int_{[0,\pi]^3}(t-h)^{-1}\,d\theta_1d\theta_2d\theta_3$. If $B(z)=(1-z)^{-1/4}{}_2F_1(1/8,5/8;1;z)^2=\sum_{r\geq0}b_rz^r$, with real principal branches for $0\leq z<1$, then $G(t)=B(4/t^2)/t$ [1], equation (4), and $J(t)=\log t-\tfrac12\sum_{r\geq1}b_r(4/t^2)^r/r$. The constant is fixed by $J(t)-\log t\to0$ as $t\to\infty$.
(3)
For $t>2$, let $u=2/(t+\sqrt{t^2-4})$ and $\alpha=(1-\sqrt{1-u^2})/2$, using positive square roots. Then $J\prime(t)=8\sqrt{\alpha(1-\alpha)}K(\alpha)^2/[\pi^2(1-2\alpha)]$, where $K$ uses the Legendre parameter, not its square root [1], equation (9).
Comments
(4)
All logarithms are natural and the angles are in radians. The integration measure has total mass one. The kernel $h$ takes values in $[-2,2]$, so $t-h$ is positive for $t>2$. At $t=2$ the logarithmic integral is convergent in the ordinary improper sense, with no principal value or subtraction.
(5)
For independent Laurent variables $X,Y,Z$, write $A_X=X+X^{-1}$, $A_Y=Y+Y^{-1}$, $A_Z=Z+Z^{-1}$ and $P_t(X,Y,Z)=8t-4(A_X+A_Y+A_Z)+A_XA_YA_Z$. Here $m(P)=\int_{[0,1]^3}\log|P(e^{i*2\pi u},e^{i*2\pi v},e^{i*2\pi w})|\,du\,dv\,dw$ uses Haar probability measure. On the torus $P_t=8(t-h)$, which gives the additive conversion [1]. Multiplying $P_t$ by $XYZ$ clears negative exponents without changing $m(P_t)$.
(6)
The value $J(2)$ is one term in the free-energy formula for the chiral $sl(n)$ Potts model [2]; the model order $n$ is not a parameter of this integral. The same source gives an arithmetic evaluation at $t=5/2$ and names further identities at $14$ and $322$ without displaying them.
(7)
When $m_0=8t$ is an integer, $P_t$ is a primitive integral Laurent polynomial with fixed nonconstant coefficients. The finite coefficient prefix in this table is exploratory reference coverage; it is not a list of arithmetic special-value theorems for every input.
Programs
(P1)
Python
import mpmath as mp
mp.mp.dps = 130

def J(t):
    t = mp.mpf(t)
    if t == 2:
        return 8*mp.catalan/mp.pi - 3*mp.log(2)
    if t < 2:
        raise ValueError("real exterior domain: t >= 2")
    def integrand(z):
        if z == 0:
            return mp.mpf(13)/64
        B = (1-z)**(-mp.mpf(1)/4) * mp.hyp2f1(mp.mpf(1)/8, mp.mpf(5)/8, 1, z)**2
        return (B-1)/(2*z)
    return mp.log(t) - mp.quad(integrand, [0, 4/t**2])

# Another value in each convention, outside the stored coefficient prefix:
t = mp.mpf(81)/8
print(J(t))
print(J(t) + mp.log(8))
References
[1]
A. J. Guttmann and M. D. Rogers, An integral arising from the chiral sl(n) Potts model, Journal of Physics A: Mathematical and Theoretical 46 (2013), 045202. (arXiv) (doi)
Links
Similar tables
Values of Dirichlet L-functions at positive integers —   The conductor-$4$ quadratic value at $s=2$ gives the spectral endpoint; the conductor-$3$ value occurs in the source's evaluation at $t=5/2$.
Complete elliptic integral of the first kind —   The derivative of $J$ is an algebraic prefactor times $K(\alpha)^2$, with the parameter transformation stated in Formulas.
Mahler measures of the simple-cubic Laurent family —   A different three-variable Laurent polynomial family, with only the six linear monomials and a constant.
Mahler measures of the Bertin Q family —   A different three-variable Laurent polynomial family with twelve nonconstant monomials; it does not use the signed linear-plus-triple-product kernel here.
Chiral Potts lattice Green function —   The resolvent of the same signed cosine kernel satisfies $J\prime(t)=\mu\prime(t)=G(t)$ for $t>2$. The condition $J(t)-\log t\to0$ fixes the additive constant; $J(2)$ is finite while $G(t)$ diverges as $t\downarrow2$.
Entropy constants of lattice models —   The square-lattice dimer entropy per site is $h_{\mathrm{sq,dimer}}=L(2,\chi_{-4})/\pi$. Thus the endpoint values here are $J(2)=8h_{\mathrm{sq,dimer}}-3\log2$ and $\mu(2)=8h_{\mathrm{sq,dimer}}$; $J(2)$ is one term in the chiral Potts free energy.
Data properties
Entries are of type: real number
Table is complete: no (it contains both normalizations at every $t=m_0/8$ for integers $16\leq m_0\leq80$, and at $t=14,322$, giving 67 arguments and 134 entries. The endpoint $2$ is intrinsic; $5/2,14,322$ are source-named arithmetic arguments. The other 63 arguments form an exploratory complete coefficient prefix of the fixed primitive integral Laurent family. Its noncanonical cutoff covers the edge to five times the edge parameter and balances useful breadth with near-edge convergence cost)
How they were obtained:

The attached generator uses Arb balls at 464 bits for 100 significant digits. It convolves the positive coefficients of (1-z)^(-1/4) and two copies of 2F1(1/8,5/8;1;z). Each factor coefficient is at most 1, so b_r is at most binomial(r+2,2). An explicit polynomial-geometric sum bounds the omitted tail of the integrated series. At the nearest exterior input 17/8, 2464 retained terms give a tail bound below 9e-127. Rational inputs, convolutions, logarithms and the log(8) conversion are enclosed. The endpoint uses the cited Catalan formula. See numerical-evidence.md and checks.py for the actual checks and their scope.

more

All 134 serialized intervals were checked against a separate implementation of the source's hypergeometric formula (with a Hurwitz-zeta endpoint check); all 66 exterior derivatives were checked through the elliptic transformation. Direct original-kernel quadrature at 2,17/8,5/2,3,5,10,14,322 is a sampled double-precision check, not certification of 100 digits. The arithmetic identity at 5/2 was independently checked by an enclosing eta-product Mellin integral with explicit product and integration tails. Exact kernel moments and deliberately wrong signs, normalizations and elliptic parameters validate the checks. See the attached numerical evidence for methods, shared assumptions, error bounds and exact scope.