README WFN1_H2
==============

This test case demonstrates that the WFN1 method maintains degeneracy even when
degenerate determinants are mixed. The case in question is H2 at large bond
distances in a minimal basis RHF wavefunction. In this case the determinant
with both bonding orbitals occupied is degenerate with the one that has both
anti-bonding orbitals occupied. In principle if we mix these two states
(without introducing electron correlation) then the mixed state should be
degenerate with these two determinants as well. In the conventional Hartree-
Fock method the mixed state will be too high in energy as spurious electron
self-interaction effect emerge. In the WFN1 method these self-interaction
artifacts are rigorously eliminated. 

This calculation performs a conventional Hartree-Fock calculation to generate
the initial RHF wavefunction. Then a calculation within the WFN1 method on
the RHF wavefunction is performed (the correlation functions will be optimized) 
(WFN1 T=0). Subsequently a calculation to generate the mixed state is performed.
A finite temperature is imposed to drive the mixing (WFN1 T>0). This also
introduces an entropy related energy term that needs to be subtracted to
generate the mixed state energy.

The results are:

RHF     : -0.545993021469
WFN1 T=0: -0.5459930215
WFN1 T>0: -0.5459930215   (Tot en -0.8232014916 plus entropy 0.2772084701)

Clearly the WFN1 method maintains the degeneracy of two determinants when
generating a state that mixes them.
