Printing of orbital basis set in spherical Gaussians

[Ondrej Marsalek]

Dear all,

to go back to the problem of printing a consistent combination of
basis set information and Kohn-Sham orbitals in spherical Gaussians,
let me try to make it as specific as possible.

The function write_orb_basis_set in basis_set_types.F currently writes
out basis set information in Cartesian Gaussians based on data in
orb_basis_set (of type gto_basis_set_type), using the following loop:

       DO iset=1,orb_basis_set%nset
          DO ishell=1,orb_basis_set%nshell(iset)
             WRITE (UNIT=output_unit,FMT="(A)") ""
             DO ico=1,nco(orb_basis_set%l(ishell,iset))
                icgf = icgf + 1
                WRITE (UNIT=output_unit,&
                     FMT="(T25,I3,4X,I4,3X,A12,(T51,2F15.6))")&
                     iset,ishell,orb_basis_set%cgf_symbol(icgf),&
                     (orb_basis_set%zet(ipgf,iset),&
                     orb_basis_set%norm_cgf(icgf)*&
                     orb_basis_set%gcc(ipgf,ishell,iset),&
                     ipgf=1,orb_basis_set%npgf(iset))
             END DO
          END DO
       END DO

What I would like to get is the orbital basis set in spherical
Gaussians, to go together with the MOs printed out in spherical
Gaussians. As spherical Gaussians is how the basis set is interpreted
and there seem to be some data of this kind in gto_basis_set_type, I
would think this is not much of a problem. We are unfortunately not
able to do this currently in our group, as it is not clear how to get
the information from gto_basis_set_type without documentation of
knowledge of CP2K internals. What I can offer in return for any help
or guidance in this is a tool (which we plan to write) to convert this
information (once available) to Molden or similar format so that the
LCAO Kohn-Sham wavefunction from CP2K can be submitted to population
analysis and other processing with available tools.

Thanks a lot,
Ondrej