3 7 75
  Here you have to specify  distributions, which will be filled during the
 Monte Carlo sessions. The distributions show  you  dependence of 
 cross section on one or two physical observables. In the last  case the 
 user has to fill  all 6 columns. If only first free columns are filled
 it is treated as a request for normal one-parameter distribution.
 
  The 'Min' and 'Max' colunms  have to define range of changing of  tested
 observable. Some numbers are expected in these colunms. The 'Parameter'
 colunm specify the observable parameter. 
  
 For collision processes there are 3 special 'observables' which can not
 be measured in practics, but could clarify phase space measure:
   E1,E2 - energies of first and second  incoming particles;
   M12   - CMS energy of incoming particles.

 Other observables can be defined by the following formulae
     KeyCharacter( name1, ....) or  
     KeyCharacter^( name1, ....) or KeyCharacter_( name1, ....)
 where name1, name2, name3 present names of outgoing particles involved 
 in the process. For example, M(u,D) corresponds to total mass of 
 {u,D} quark pair.  The  available KeyCharacters are presented below:

  A  - Angle  in degree unit.  
  C  - Cosine of angle   
  J  - Jet cone angle    
  E  - Energy of the particle set
  M  - Mass of the particle set 
  P  - Cosine of angle between first particle in the list and 
       direction of boosting of all particle set in the rest frame of 
       particles set. 
  T  - Transverse momentum P_t  of the particle set 
  S  - Squared mass
  Y  - Rapidity of particle set.
  N  - Pseudo-rapidity of particle set.
  Z  - transverce energy
  U  - user defined function 

In general 'A','C'  have  2 arguments. But in case of collision  
these functions with one argument designate angle between
the given particle and first (in the list) incoming particle.

One can use here names of particles which are involved in others 
subprocesses. But for given Monte Carlo session CalcHEP fills only 
distributions related to the current subprocess.
 
Sometimes the list of outgoing particles contans identical ones
like  in the case of "e,E->A,A" process. For such subprocess 
E(A) distribution  sums  cross sections for both gammas. 
If there are identical particles in the output  list of  subprocess then
all possibilities to realize the functions will be tested and summed.
The '^'/'_'  modifiers following to KeyCharacter force the function 
to choose only the highest/lowest value correspondingly. 

One can type 'Jet' instead of particles name. CalcHEP  interprets 
it as one of 5 quarks of gluon. Summation over all possible 
substitution will be realized.

 User defined distributions also can be applied. They are specified by 
 'U<character_set>'. The list of particles is not expected after this
 name. CalcHEP will call  
          usrfun("<character_set>")
 For example 
   'Uabc'   leads to call usrfun("abs")
 Originally CalcHEP contains a dummy version of 'usrfun' which leads to 
 stop of n_calchep execution. An object file or library with a code of 
 usrfun has to be written in table 'Libraries' which is a part of model
 specication. As well one can use the "Edit linker" option of CalcHEP 
 symbolic session. An approach to particle momenta and their names 
 needed to write 'usrfun' are explained in CalcHEP manual. 
   
 To see the  plots use "Display Distributions" menu function.
 After end of each Monte Carlo session calchep save generated 
 distributions in 'result/distr_#' file where # designates session number.
 One can display them later on by the 
       $CALCHEP/bin/show_distr  result/distr_# 
 function.
 As well as 
        $CALCHEP/bin/sum_distr   distr_1 distr_2 ... >sum
 is intended for summation of distributions generated in different 
 subprocesses.
 
