Documentation for GAUSBAS 15 May 2001                        A.F, J.D.

########################################################################

GTOBAS is a program for fitting Gaussian-type orbitals (GTOs) to
continuum functions. The Exponents of GTOs are optimized using the
method of Nestmann and Peyerimhoff [1]. The numerical routines are
from Numerical Recipes [2]. The numerical continuum functions, which
provide the input for the program, must be generated from an external
module (e.g. NUMBAS).

For a more detailed description of the program see: Faure et al.,
Computer Physics Communications 144 (2002) 224–241.

[1] Nestmann B. M. and Peyerimhoff S. D., J. Phys. B. (1990) L773
[2] Press W. H., Teukolski S. A., Vetterling W. T. and Flannery
     B. P., Numerical Recipes in Fortran 77, Cambridge University Press
     (1986-1992).

########################################################################


The code is set up to run in 'batch' mode, ie all input is from the
namelist /fit/ on the standard input. 

The variables input via namelist /fit/ are 
(name   ::  description (default value)

     iguess  :: initial selection of exponents (2), 
                =0 exponents provided by the user, 
                =1 random selection, 
                =2 selection from an appropriate function
     expo    :: list of initial exponents (needed if iguess=0)
     noexp   :: number of exponents to be used (needed if iguess=1,2). 
                Must be larger or equal to the number of functions. 
     nplot   :: no functions (0) / all functions (1) on unit luplot (0)
     lunumb  :: unit number for numerical functions input (13 for NUMBAS) 
     luplot  :: unit number for plotable functions output (if used) (17)
     iswmol3 :: unit number for final exponents in SWMOL3 format (18)
     ftol    :: convergence criterion usid in POWELL (1.d-9)
     iprint  :: print flag (0) 
                =1 all iteration data, 
                =2 plus mesh. 
     rdlow   :: lower limit of the random selection (0.01) 
                (used if iguess=1)
     rdup    :: upper limit of the random selection (0.49) 
                (used if iguess=1) 
     beta    :: even-tempered beta coefficient (0.016)
                (used if iguess=2)
     gamma   :: even-tempered gamma coefficient (1.39)
                (used if iguess=2)

Examples (both examples require running the program NUMCBAS first in order to
generate the n umerical functions, Bessel or Coulomb, that are fitted by GTOBAS.
________


1) Ionic targets : Coulomb functions (model potential = point charge) 
   *************

cat > basis.data << EOF
 &INPUT
  title='Ionic target',
  lval=0,
  ecmax=5.0D0,
  rlim=12.0D0,
  charge=1.0D0,
 /
EOF
./numbas < basis.data

cat > gtobas.data << EOF
 &FIT
 noexp=12,
 iguess=2,
 nplot=1,     
 iprint=0,
 /  
EOF
./gtobas < gtobas.data


2) Neutral targets : Bessel functions (model potential = zero). 
   ***************

cat > basis.data << EOF
 &INPUT
  title='Neutral target',
  lval=0,
  ecmax=4.0D0,
  rlim=12.0D0,
  chrage=0.0D0,
 /
EOF
./numbas < basis.data

cat > gtobas.data << EOF
 &FIT
 expo=6.0,3.0,0.7,0.3,0.1,0.07,0.05,0.02,0.01,
 iguess=0,
 nplot=1,                          
 iprint=0,
 /
EOF
./gtobas < gtobas.data 
