readme

The most recent package is QMPACK version 1.01

Please work with a Linux machine. No test has been done with Windows machines.

The following is the procedure to use this Fortran program package.

1. LAPACK and BLAS subroutine packages

  are necessary. These are usually installed in default for almost all recent       
  Linux machines. If not, please download these subroutine packages from the    
  following homepage (free of charge).

LAPACK homepage

  The intel MKL (math kernel library) (not free of charge) is most desirable.

2. ARPACK subroutine package

  Please download this from the following homepage (free of charge).
  Do not forget to download patch.tar.gz also.
  These are not actually necessary, if you only use fssG.f. In such a case,
  please take care of the Fortran compiler by yourself.

ARPACK homepage

  Please make sure first, by using some SAMPLE drivers, if the ARPACK  
  subroutines are properly installed. In particular, it is necessary
  to set up with the ARPACK, LAPACK and BLAS autocall subroutine libraries
  by properly modifying ARmake.inc, for example,

  ARPACKLIB  = $(home)/arpack.a
  LAPACKLIB =  /home/user02/fujiwara/kudmsp/LAPACK/lapack_LINUX.a
  BLASLIB = /home/user02/fujiwara/kudmsp/LAPACK/blas_LINUX.a
  ALIBS =  $(ARPACKLIB) $(LAPACKLIB) $(BLASLIB) 

3. QMPACK

  should be put in the same directory as LAPACK and ARPACK exist.
  Otherwise, you have to modify the top line of the makefile:
  include ../ARPACK/ARmake.inc

4. make clean and make test

  yield channel information contained in the arrays in the common sentence
  /COMJUN/  The results of the Born amplitudes are given in the arrays
  in /COMBRN/ The main program main.f of the executable program test explains
  where the necessary Born amplitudes are stored. 

  If you set  
  IYTMX=2
  IYTMN=-2
  at the top of main.f (or fssG.f etc.), calculations are made for
  all the B8 B8 channels.
  If you need Sigma+ p (or Sigma N (I=3/2) ), modify IQTMN(2)=1 and IQTMXA(2)
  in trans.h as
  C   IQTMNA(2)=1
  C   IQTMXA(2)=1 ! only Lambda p (make valid for HYPTRN use)
  (For using hypt.f, the isospin I=3/2 channel should be suppressed.)

5. make clean or make fssG

  yields phase shifts and cross sections of fss2 or FSS.
  The selection of the model can be done by specifying 
  MODEL=1 (FSS) or MODEL=2 (fss2) at the top of the programs.
  ICRSON=1 yields cross sections, but also make JTMAX=8 just below to include
  a sufficient number of partial waves.
  If you take off the comment sign of CALL SUBOBS(NE, IPRINT), contained in
  the subroutine CRSPLT (in brnsub.f), almost all of simple spin observables
  can be calculated (but with a large amount of output). 
  If you set ICRSON＝0 and JTMAX=4 for IYTMX=2 and IYTMN=2, the experimental
  NN phase shifts (Arndt SP99) will be given just below the calculated values,      
  so that you can make sure the agreement between them.
  The selection of two-baryon systems can be controlled by the maximum (IYTMX)  
  and the minimum (IYTMN) values of the total hypercharge Y (from 2 to -2);
  for example, IYTMX=2 and IYTMN=-2 gives a full output.

6. Other sample programs

  are also executable by
  make clean
  make program
  whre program is fssG, triton or hypt etc.
  triton is the triton Faddeev calculation、and hypt is the hypertriton
  Faddeev calculation. The selection of the momentum mesh points can
  be done by rewriting triton.h, hypt.h, etc. Please choose
  n_1-n_2-n_3=5-5-5 or 10-10-5. If you choose 10-10-5, you should get the 
  same value as the one we published, but it is usually very time consuming.
  (This is the default setting.) The selection of the model is the same as 
  fssG; namely, MODEL=1 (FSS) or MODEL=2 (fss2). Please modify this at the top
  of triton.f or hypt.f.

If you have any troubles, please feel free to contact us.

Y. Fujiwara

phone: +81-75-753-3875

fax: +81-75-753-3886

e-mail: fujiwara@ruby.scphys.kyoto-u.ac.jp

revised on 5 August, 2004, for version 1.01

readme

manual

figures in particle basis

download

publications

link

memo

recent(10)

edit