integer nba,nphase,nka,npmax integer iden,igamow,idirconv double precision v_source integer iicheap,ident integer ireduce parameter (nba=1,nphase=10000,nka=25,npmax=10000) parameter (iden=1,igamow=0,idirconv=1,v_source=0.d0) parameter (iicheap=0,ident=1) parameter (ireduce=2) c c 'nba' refers to the number of impact parameters to be used. c Keeping 'nba' no larger than it has to be will reduce the c memory needed. nphase refers to the number of phase space c points needed for each impact parameter. Again keep this to c a minimum to reduce the memory needed. c c 'nka' is the number of momenta for which the correlation function c will be calculated. The program should prompt you if these c parameters have not been chosen sufficiently large. c c Choose 'iden=2', if you wish to construct the denominator only using c events where 2 or more particles were measured. I DO NOT c recommend doing this but several groups are stuck on it for hardware c reasons. Choose 'npmax' large enough, approximately 25000 or more c to make a good sampling of single-particle spectra, but only if using c the 'iden=2' option. c c It also is not the usual definition of the correlation function. c If one wishes to construct the denom. using all possible events, c choose 'iden=1'. If 'iden=1', you can set npmax arbitrarily small c to same memory as none of those arrays are used. c c If one likes Gamow corrected correlations choose igamow=1, c otherwise choose 'igamow=0'. c c Choose idirconv to be an integer 1-5, depending on the convention c you wish to use for defining the direction and frame for the relative c momentum. See additional explanation/code, in the subroutine c "decompose". c --------------------------------------------------------------- c Choose 'idirconv=1' for convention where 'kx,ky,kz' are defined first c in frame where 'pz1+pz2=0', then boosted outwards to the frame of the c pair. (in this conv. 'k^2=Qinv', and the z direction is parallel c to the beam axis) c c Choose 'idirconv=2 for conv. where one boosts to frame of the pair, c but not the outwards boost. (in this conv. the z-direction is parallel c to the beam axis) c c Choose 'idirconv=3' for conv. where one boosts along the z-axis to c a frame specified by 'v_source=constant' (in this conv. the z-direction c is along the beam axis) c c Choose 'idirconv=4' for conv. where one boosts along the z-axis to c a frame specified by 'v_source=constant', and chooses the x-axis c parallel to the total momentum (in this conv. the z-axis is not c necessarily parallel to the beam axis, and 'k^2' does not equal 'Qinv^2') c c Choose 'idirconv=5' for conv. where one does the same as in convention c 4, only one does one more boost parallel to 'ptot', such that c 'k^2=Qinv^2'. c ---------------------------------------------------------------- c c If 'idirconv' = 3,4 or 5 above, then you need to specify the velocity c of the source's frame, 'v_source'. Note that if you know the rapidity c of the source, 'y_source', then 'v_source=tanh(y_source)'. c c Make iicheap=1 if you are using Breit-Wigner form for wave-function. c Otherwise, make it zero. There are Gamow-Teller forms for (K+,K-) c (p,pi+) (d,alpha) (p,alpha). Each of these forms accounts for only one c resonant peak. Look in the subroutine "cheapset" for more info. c c If "ident" =1 then we are using identical particles c or at least the same phase space pt.s for both, and we c do not need to set arrays for both types. c If the particles are not identical, comment out the c equivalence statement in the main program and set "ident"=0. c c Choose 'ireduce=1' if you wish to use the reduced relative c momenta to plot '{ (m1*p2-m2*p1)/(m1+m2), = .5*(p2-p1) if c m1=m2 }' or choose 'ireduce=2' if you prefer '{p2-p1}' as the c variable to plot. All HE people use the =2 option for pi-pi or K-K.