MatthewRHermes / mrh

MRH's research code
Other
18 stars 30 forks source link

LASSIS choice of Sz in optimizing charge-hop fragment basis functions #125

Open MatthewRHermes opened 3 days ago

MatthewRHermes commented 3 days ago

When optimizing the CI vectors of fragment basis functions, even though we use spin shuffle operations to guarantee that the final model space is spin-adapted, we still have to choose in what Sz sector for each fragment we optimize the CI vectors, because the fragments' spins interact with each other through a 1-electron operator. Currently, the choice is extremely arbitrary and difficult to even describe. A more systematic way of choosing the Sz space needs to be made in order for the method to be reproducible.