Projector implementation
@all I suggest to implement a projector version of the BSS algorithm. From what I see, it should be rather easy to do. We essentially will have to add a function to the Hamiltonian module providing the Slater determinant, modify the udv_state (reset and matmult) and adapt the calculation of the Greens function. I believe that all updating routines remain untouched as they only require the equal time Greens function and don't care whether it is a finite or zero temperature one. I volunteer to prepare the udv_state for this by adding a N_p member and assuming rectangular shape. @fassaad I think we should discuss whether I am overlooking something. Also, I have an idea how to adapt cgr1.f90, how to modify cgr2_2.f90 is not clear to me yet.