Generalised Version : JADE or DURHAM or GENEVA or INVMAS algorithm

The arguments and usage are identical to that for QJMMCL above , except for the extra input arguments : Input arguments:

SCHEME

CHARACTER*2 : 'E' or 'P' or 'E₀' (see above)

VERSN

CHARACTER*6 : 'JADE' or 'DURHAM' or 'GENEVA'
or 'INVMAS' to select the corresponding algorithm

YCUT

- if > 0. : Cut on the scaled invariant mass of 2 tracks. Pairs of tracks are merged if their scaled invariant mass is smaller than YCUT.
- if < 0. : fixed number of jet asked for, i.e. you can request that the event is clustered until a certain number of jets is found. Example: setting YCUT=-4. will force the clustering to get 4 jets and only 4 jets as output.

Note that 'NORMAL' and 'BETTER' are alternatives for 'JADE' and 'DURHAM' for historical reasons.

If the 'INVMAS' algorithm is selected, the 'E' scheme is automatically switched on, whatever has been put in the input argument SCHEME.

QJMMCL and QDMMCL both call this routine , which itself calls the ALEPHLIB routine FJMMCL , which actually does the jet finding.

If you have called QGJMMC with a negative value of YCUT to obtain a fixed number of output jets, this routine returns a vector of all Yijs computed, and the number of steps needed to get 4 jets.

Example: In your user routine you define a vector:

      Real*4  MyYvalues(1000)
      Integer Nsteps
      Real EVIS,Y34,Y45,YCUT,EVIS
      Character*2 Scheme
      Character*6 VERSN

   .... then at some stage you call the clustering ...

      call QJOPTR('EF',' ')
      EVIS=0.
      YCUT=-4.
      SCHEME = 'E'
      VERSN  = 'DURHAM'
      call QGJMMC(NJETS,'myjets',KRECO,YCUT,EVIS,SCHEME,VERSN)
      IF (NJETS.LE.0) RETURN  ! Error

   ... and afterwards you call ...

      call GETYIJ(MyYvalues,Nsteps)

   ... this returns a vector of all yijs computed and
       the number of steps needed to get to 4 jets. Therefore

      Y34 = MyYvalues(Nsteps)
      Y45 = MyYvalues(Nsteps-1)