Talk:Multireference configuration interaction

For me choosing the configuration for the CI-space from some references is done always by taking singly, doubly, ... excited configurations relative to all reference. I do not know what is meant by "chosen by the program according to some perturbation theoretical ansatz according to a threshold provided by the user". Maybe this can be explained or taken into a new paragraph, describing this different/special method.
 * Yes of course. The extrapolation to the 0 threshold value is also possible. The analysis of the convergence give usually hints on wether the references are well chosen.  —Preceding unsigned comment added by 79.233.247.103 (talk) 11:37, 3 March 2009 (UTC)

CSFs are not necessarily determinants
The third sentence start with 'The higher excited determinants (also called configuration state functions (CSFs) or shortly configurations)...'. However, a CSF is a linear combination of determinants, and not always one determinant. The linear coefficients are called spin coupling coefficients. A CSF is equal to one determinant if and only if all orbitals are doubly occupied in the respective configuration. Libavius (talk) 17:08, 3 February 2020 (UTC)