Steady-state Solver

Once a (Bosonic or Fermionic) problem has been defined, one can solve for the steady-state dynamics. By default this is done by direct SPLU decomposition (combined with the standard imposition of normalization on the system density operator).

Typical usage is:

HEOMMats = BosonicHEOMSolver(Hsys, Q2, ckAR, ckAI, vkAR, vkAI, NC, options=options)
#get steady state
rho_ss,full_ss=resultHEOM.steady_state()

rho_ss is a standard QuTiP quantum object for the system state alone. full_ss includes that and every auxiliary density operator encoding the environment, but in numpy array format (not converted to quantum objects). To determine which ADO is which, one can use the enr_state_dictionaries() built into QuTiP. This is also used to generate the ordering of all the HEOM operators in our code. An example of how this is done is presented in example notebook 4b, where it is used to get the electronic current from the ADOs.

We will supplement this with further examples for the Bosonic case in the future.

Additional options for steady_state() include use_mkl = True which will use MKL parallelization for the problem, if available. BoFiN_fast also includes an approx=True option for use of an iterative lgmres approach.