c This file is part of the EERAD3 NNLO event generator. c Copyright (C) 2025 the authors. c This program is free software: you can redistribute it and/or c modify it under the terms of the GNU General Public License as c published by the Free Software Foundation, either version 3 of c the License, or any later version. See COPYING for details. c This file contains subroutines calculating event-shape c observables. C The code has been ported to fit EVENT2 variable definitions. c----------------------------------------------------------------------- c C and D parameters. c----------------------------------------------------------------------- subroutine getCD(Cpar,Dpar,P,N) implicit real*8(a-h,o-z) double precision P(4,7) t11=theta(1,1,P,N) t12=theta(1,2,P,N) t13=theta(1,3,P,N) t22=theta(2,2,P,N) t23=theta(2,3,P,N) t33=theta(3,3,P,N) Dpar=t11*t22*t33-t11*t23**2-t22*t13**2-t33*t12**2+2d0*t12*t13*t23 Dpar=27d0*Dpar Cpar=-t12**2-t13**2-t23**2+t11*t22+t11*t33+t22*t33 Cpar=3d0*Cpar return end ************************************************************************ function theta(ii,ij,P,N) implicit real*8(a-h,o-z) double precision P(4,7) top=0d0 bot=0d0 do i=1,N top=top+P(ii,i)*P(ij,i)/P(4,i) bot=bot+P(4,i) enddo theta=top/bot return end c----------------------------------------------------------------------- c Sphericity observables. c----------------------------------------------------------------------- c Calculate S, A, and P parameters. subroutine getSAP(Spar,Apar,Planar,P,N) implicit real*8(a-h,o-z) double precision P(4,7) parameter(pi=3.141592653589793238d0) t11=phi(1,1,P,N) t12=phi(1,2,P,N) t13=phi(1,3,P,N) t22=phi(2,2,P,N) t23=phi(2,3,P,N) t33=phi(3,3,P,N) x=t11+t22+t33 y=t11*t22+t11*t33+t22*t33-t12**2-t13**2-t23**2 z=t11*t22*t33-t11*t23**2-t22*t13**2-t33*t12**2+2d0*t12*t13*t23 C=3d0*y/x**2 D=27d0*z/x**3 sqrt1mc=sqrt(1d0-C) ctheta=(D-3d0*C+2d0)/2d0/(1d0-C)/sqrt1mc if (ctheta.gt.1d0) ctheta=1d0 if (ctheta.lt.-1d0) ctheta=-1d0 theta=acos(ctheta) ct=cos(theta/3d0) st=sin(theta/3d0) rt3=dsqrt(3d0) s1=1d0/3d0+2d0/3d0*sqrt1mc*ct s2=1d0/3d0-1d0/3d0*sqrt1mc*(1d0*ct+rt3*st) s3=1d0/3d0-1d0/3d0*sqrt1mc*(1d0*ct-rt3*st) s1=2d0/3d0*sqrt1mc*cos(theta/3d0)+1d0/3d0 s2=2d0/3d0*sqrt1mc*cos((theta+2d0*pi)/3d0)+1d0/3d0 s3=2d0/3d0*sqrt1mc*cos((theta+4d0*pi)/3d0)+1d0/3d0 e3=dmin1(s1,s2,s3) if(e3.lt.0d0)e3=0d0 e1=dmax1(s1,s2,s3) e2=1d0-e1-e3 Spar=3d0/2d0*(e2+e3) Apar=3d0/2d0*e3 Planar=e2-e3 return end ************************************************************************ function phi(ii,ij,P,N) implicit real*8(a-h,o-z) double precision P(4,7) phi=0d0 do i=1,N phi=phi+P(ii,i)*P(ij,i) enddo return end c----------------------------------------------------------------------- c Thrust and friends. c----------------------------------------------------------------------- c Compute thrust T and all observables related to it. c (MH, BW, BT, TMinor, TMajor, FCx, etc) subroutine getT(Tpar,Tmajor,Tminor,Opar,em2h,em2l,em2d, . bmax,bmin,bsum,bdiff,fc0,fc1,fc2,fc3,bks0,bks1,bks2,bks3, . P,N) implicit real*8(a-h,o-z) double precision P(4,7) double precision pp(5,15),ta(3),tmja(3),tmna(3) do i=1,15 do j=1,5 pp(j,i)=0d0 enddo enddo np=0 do j=1,N np=np+1 do i=1,4 pp(i,np)=P(i,j) enddo pp(5,np)=pp(4,np) enddo if(N.eq.4)then do k=2,N np=np+1 do i=1,4 pp(i,np)=P(i,1)+P(i,k) enddo pp(5,np)=sqrt(pp(1,np)**2+pp(2,np)**2+pp(3,np)**2) enddo endif c Find thrust axis and calculate thrust. ithrust=0 Tpar=0d0 do i=1,15 t=2d0*pp(5,i) if(t.gt.Tpar) then Tpar=t ithrust=i endif enddo c Thrust axis (unit vector). do j=1,3 ta(j) = 2d0*pp(j,ithrust)/Tpar enddo c Find major axis (maximizes momentum perp to thrust axis). imajor=0 Tmajor=0d0 Tminor=0d0 Opar=0d0 do i=1,15 if (i.ne.ithrust)then tmp = pp(1,i)*ta(1)+pp(2,i)*ta(2)+pp(3,i)*ta(3) t = 2d0*sqrt(abs(pp(5,i)**2-tmp**2)) if (t.gt.Tmajor)then Tmajor = t imajor = i ppta = tmp endif endif enddo if (Tmajor.gt.0)then do i=1,3 tmja(i) = 2d0*(pp(i,imajor) - ta(i)*ppta)/Tmajor enddo c Find minor axis. tmna(1) = ta(2)*tmja(3)-ta(3)*tmja(2) tmna(2) = ta(3)*tmja(1)-ta(1)*tmja(3) tmna(3) = ta(1)*tmja(2)-ta(2)*tmja(1) Tminor=0d0 do i=1,N Tminor = Tminor + . abs(pp(1,i)*tmna(1)+pp(2,i)*tmna(2)+pp(3,i)*tmna(3)) enddo c Oblateness. Opar = Tmajor-Tminor endif c Hemisphere masses. em2h = pp(4,ithrust)**2-pp(5,ithrust)**2 em2l = (1d0-pp(4,ithrust))**2-pp(5,ithrust)**2 if (em2l.gt.em2h)then tmp = em2l em2l = em2h em2h = tmp endif em2d = em2h-em2l c Hemisphere broadening. bplus = 0d0 bminus = 0d0 bot = 0d0 do i=1,N bot = bot+pp(4,i) if (i.ne.ithrust)then tmp1 = pp(1,i)*ta(1)+pp(2,i)*ta(2)+pp(3,i)*ta(3) tmp2 = sqrt(abs(pp(4,i)**2-tmp1**2)) if (tmp1.gt.0d0)then bplus = bplus+tmp2 elseif (tmp1.lt.0d0)then bminus = bminus+tmp2 endif endif enddo bplus = bplus/bot/2d0 bminus = bminus/bot/2d0 bmax = dmax1(bplus,bminus) bmin = dmin1(bplus,bminus) bsum = bmax+bmin bdiff = bmax-bmin c Fractional energy correlations. Etot = 0d0 FC0 = 0d0 FC1 = 0d0 FC2 = 0d0 FC3 = 0d0 do i=1,N Etot = Etot + P(4,i) do j=1,i-1 qit = pp(1,i)*ta(1)+pp(2,i)*ta(2)+pp(3,i)*ta(3) qjt = pp(1,j)*ta(1)+pp(2,j)*ta(2)+pp(3,j)*ta(3) if (qit*qjt.le.0) cycle pipj = pp(1,i)*pp(1,j)+pp(2,i)*pp(2,j)+pp(3,i)*pp(3,j) pixpj = pp(2,i)*pp(3,j)-pp(3,i)*pp(2,j) . + pp(3,i)*pp(1,j)-pp(1,i)*pp(3,j) . + pp(1,i)*pp(2,j)-pp(2,i)*pp(1,j) Ei = pp(4,i) Ej = pp(4,j) pi = dsqrt(pp(1,i)**2+pp(2,i)**2+pp(3,i)**2) pj = dsqrt(pp(1,j)**2+pp(2,j)**2+pp(3,j)**2) omcTh = 1d0-dabs(pipj/pi/pj) sTh = dabs(pixpj/pi/pj) if (omcTh.lt.0d0) cycle FC0 = FC0 + 2d0*Ei*Ej*omcTh FC1 = FC1 + 2d0*Ei*Ej*sTh**0.5d0*omcTh**0.5d0 FC2 = FC2 + 2d0*Ei*Ej*sTh FC3 = FC3 + 2d0*Ei*Ej*sTh**1.5d0*omcTh**(-0.5d0) enddo enddo FC0 = FC0/Etot**2 FC1 = FC1/Etot**2 FC2 = FC2/Etot**2 FC3 = FC3/Etot**2 c BKS observables. qsum = 0d0 BKS0 = 0d0 BKS1 = 0d0 BKS2 = 0d0 BKS3 = 0d0 do i=1,N Ei = pp(4,i) qi = dsqrt(pp(1,i)**2+pp(2,i)**2+pp(3,i)**2) qit = pp(1,i)*ta(1)+pp(2,i)*ta(2)+pp(3,i)*ta(3) qixt = pp(2,i)*ta(3)-pp(3,i)*ta(2) . + pp(3,i)*ta(1)-pp(1,i)*ta(3) . + pp(1,i)*ta(2)-pp(2,i)*ta(1) omcTh = 1d0-dabs(qit/qi) sTh = dabs(qixt/qi) qsum = qsum + qi if (omcTh.lt.0d0) cycle BKS0 = BKS0 + Ei*omcTh BKS1 = BKS1 + Ei*sTh**0.5d0*omcTh**0.5d0 BKS2 = BKS2 + Ei*sTh BKS3 = BKS3 + Ei*sTh**1.5d0*omcTh**(-0.5d0) enddo BKS0 = BKS0/qsum BKS1 = BKS1/qsum BKS2 = BKS2/qsum BKS3 = BKS3/qsum return end c----------------------------------------------------------------------- c Energy-energy correlation. c----------------------------------------------------------------------- c All outputs are per unit event weight (multiply by the weight). c For each pair of partons i>j (npair=N(N-1)/2 pairs) returns c cosang(k) = cos(chi_ij) c eec(k) = 2 (1-cos^2) E_i E_j / Q^2 (EVENT2 convention) c and the sums over pairs c bs(k,l) = sum eec * sin^(l-1) * cos^(k-1) (moments) c eectot = sum eec c eecwin = sum eec/(2 coslim) for |cos| < coslim c P must be normalised to Q=1 (CM frame). subroutine getEEC(cosang,eec,npair,bs,eecwin,eectot,P,N) implicit real*8(a-h,o-z) double precision P(4,7),cosang(6),eec(6),bs(2,6) parameter (coslim=0.1d0) do k=1,2 do l=1,6 bs(k,l)=0d0 enddo enddo eecwin=0d0 eectot=0d0 npair=0 do i=2,N do j=1,i-1 npair=npair+1 pij=P(4,i)*P(4,j)-P(3,i)*P(3,j) . -P(2,i)*P(2,j)-P(1,i)*P(1,j) c=1d0-pij/(P(4,i)*P(4,j)) if (abs(c).ge.1d0) c=sign(1d0-1d-12,c) cosang(npair)=c eec(npair)=2d0*(1d0-c*c)*P(4,i)*P(4,j) eectot=eectot+eec(npair) if (abs(c).lt.coslim) eecwin=eecwin+eec(npair)/(2d0*coslim) sn=sqrt(1d0-c**2) do k=1,2 do l=1,6 bs(k,l)=bs(k,l)+eec(npair)*sn**(l-1)*c**(k-1) enddo enddo enddo enddo return end