Mulliken population analysis is performed in AO basis. The analysis is based on the concept of labels. Each basis function is labeled by its functional type and center (and boson irrep when in SO basis, the default). The labels are given in the output. A set of primitive labels may be collected to group labels as specified by the user.
For each fermion irrep, give an orbital_string of orbitals to analyze.
Default: Analyze the occupied electronic solutions.
Give Mulliken net/overlap populations in addition to gross populations.
Define labels for use in Mulliken population analysis.
Give first the number of labels to define, then the label name (12 characters) and orbital_string for each fermion irrep.
Example:
.LABDEF
4 (total number of lines to follow)
Re s 1
Re p 2..4
Re d 5..10
Re f 11..20
... 21..35
Pitfall:
If you decide to redefine labels for the population analysis you have to reassign all of them, otherwise Dirac is unable to continue. The list of functions with the initial label definitions (default: SO-labels) can be drawn from the DIRAC output starting at the section :
GETLAB: SO-labels
Large components: 35
1 L Ag AR s 2 L Ag AR dxx 3 L Ag AR dyy 4 L Ag AR dzz etc.
Use individual small component labels.
Default: Gather all small component functions belonging to a given center and irrep, irrespective of function type.