Update QTSH #212
Update QTSH #212
Conversation
src/dyn/libdyn.cpp
Outdated
| @@ -100,6 +100,9 @@ void export_dyn_control_params_objects(){ | |||
| .def_readwrite("fssh3_max_steps", &dyn_control_params::fssh3_max_steps) | |||
| .def_readwrite("fssh3_err_tol", &dyn_control_params::fssh3_err_tol) | |||
|
|
|||
| ///================= FSSH3 specific ==================== | |||
| double sum = 0.0; | ||
| for(int n=0; n<nadi; n++){ | ||
| for(int k=0; k<n; k++){ | ||
| sum += ham.children[itraj]->dc1_adi[idof]->get(k, n).real() * dyn_var.dm_adi[itraj]->get(k, n).imag(); | ||
| //sum += ham.children[itraj]->dc1_adi[idof]->get(k, n).real() * dyn_var.dm_adi[itraj]->get(k, n).imag(); |
There was a problem hiding this comment.
i think, the version without T matrix should still work, but the call of the function should be after the electronic integration and the density matrix update call (so that the density matrix is consistent with the new derivative couplings)
src/dyn/dyn_ham.cpp
Outdated
| @@ -39,6 +39,7 @@ void update_Hamiltonian_variables(dyn_control_params& prms, dyn_variables& dyn_v | |||
| - 0: in response to changed q | |||
| - 1: in response to changed p | |||
| - 2: in response to changed electronic amplitudes | |||
| - 3: in response to changed p_k in QTSH | |||
There was a problem hiding this comment.
let's not complicate this function by such a pretty-methodology-specific case; this update would be of the kind "0" and "1" - because 0 - is in response to changed NAC (that depends on q) and 1 - is in response to changed p (which is changed indirectly to since p_k changes).
I'd still think on p as central variables - the kinematic p is only needed as an auxiliary variable.
| @@ -456,6 +456,14 @@ class dyn_control_params{ | |||
| double fssh3_err_tol; | |||
|
|
|||
|
|
|||
| //=========== QTSH options ========== | |||
| /** | |||
| Whether to use QTSH | |||
There was a problem hiding this comment.
add options possible and what they mean: 0 - not using QTSH; 1 - using QTSH - explain to user what is actually changed when we use QTSH (a different convention for the force to use)
src/dyn/Dynamics.cpp
Outdated
| @@ -1562,8 +1573,10 @@ void compute_dynamics(dyn_variables& dyn_var, bp::dict dyn_params, | |||
| if(prms.thermally_corrected_nbra==1 && prms.tcnbra_do_nac_scaling==1){ remove_thermal_correction(dyn_var, ham, prms); } | |||
|
|
|||
| // Update vib Hamiltonian to reflect the change of the momentum | |||
| update_Hamiltonian_variables(prms, dyn_var, ham, ham_aux, py_funct, params, 1); | |||
|
|
|||
| if(prms.use_qtsh==0){ | |||
There was a problem hiding this comment.
i don't agree with this conditional call (nested inside if block) of this function - basically, is should be always called but would do nothing for QTSH. You are right - this could be more efficient your way, but adds the complexity of the code. In this case I'd sacrifice the efficiency for the sake of clarity and simplicity
src/dyn/Energy_and_Forces.cpp
Outdated
| *dyn_vars.f = ham.forces_adi(effective_states).real(); | ||
| }// rep_force == 1 | ||
|
|
||
| // Non-classical term calculations |
There was a problem hiding this comment.
the term "non-classical" is confusing here - do you mean the total QTSH force, or only the Ehrenfest-like correction to the adiabatic forces? I think we could just use the regular forces variable to keep the entire QTSH force, if this method is used.
No description provided.