maisebastian
commented
1 month ago Dear Jakob, if your question is about the naming of the different trajectory folders, it should actually not be related to LVC, but is generally true for all methods.
The naming of the folders is given in https://github.com/sharc-md/sharc/blob/9a8660e7a17dd895874bd81232d82294cfce39f5/bin/setup_traj.py#L4735C1-L4741C19 where you can see that the index is reduced by 1 if its singlets or doublets. In that way, the folders would start counting from Singlet_0 and Doublet_0, but from Triplet_1, etc. This follows an often-used convention for these states (although it might be that in some molecules the actual ground state is a triplet or another multiplicity).
So if the output.xyz says that the active state is 8, then that would be D7, assuming that you use "0 10 0" states.
In the LVC.template file, state 1 would be D0 and state 2 would be D1. There are 10 states because you requested 10 states, as in SHARC the ground state is always consistently included in the counting (even though this is somewhat unnatural for single-reference methods like TDDFT).
Best, Sebastian
jakobcasa
Dear Sebastian
I have a question about the nomenclature of the states, especially in the LVC.
In the LVC template, I have included ten excited states; by looking at the template, I can see that there is no coupling between the D1 and D0 combined with the number of states printed in the LVC template. The number of printed states is exactly ten and not eleven.
Let's assume I start with the traj at the D7 state. The state noted at the beginning of the XYZ file is state 8, which would lead to the conclusion that state 8 is populated, but the name of the folder produced by sharc tells me that I have exited the Doublet_7 state. I hope you can see my confusion; if not, I can try to explain it a bit.
Best Jakob