Make the bosonic solvers have parity with the fermions in how multiple baths and their lists of parameters are handled. We left the Bosonic one more general in a way, but in hindsight I think how we do it in the fermionic case is much easier for most cases.
In combining common exponents, we use a somewhat arbitrary closeness measure. E.g., in one case, I had to manually change it to get it to do what I want. can we do this is a different way? Maybe it is ok as it is now for most situations..
Add options to the ODE solvers to return the full hierarchy for every time step. A simple boolean choice will probably suffice for the moment. I did this on my local versions, but didn't commit to it for some reason.
Features:
Spectrum solvers. Still chasing the elusive kondo resonance, then will add this. At the moment users can do it themselves with RHSmat and the steady-state solver, but a built in version would be nice.
RHS construction in parallel. This will help the python examples at least. General speed-up of the Python construction would be good in the long-term.
Dynamic truncation of ADOs.
Fermionic hermiticity relation optimization (this is a priority).
Just a quick TODO list:
Usability
Make the bosonic solvers have parity with the fermions in how multiple baths and their lists of parameters are handled. We left the Bosonic one more general in a way, but in hindsight I think how we do it in the fermionic case is much easier for most cases.
In combining common exponents, we use a somewhat arbitrary closeness measure. E.g., in one case, I had to manually change it to get it to do what I want. can we do this is a different way? Maybe it is ok as it is now for most situations..
Add options to the ODE solvers to return the full hierarchy for every time step. A simple boolean choice will probably suffice for the moment. I did this on my local versions, but didn't commit to it for some reason.
Features:
Spectrum solvers. Still chasing the elusive kondo resonance, then will add this. At the moment users can do it themselves with RHSmat and the steady-state solver, but a built in version would be nice.
RHS construction in parallel. This will help the python examples at least. General speed-up of the Python construction would be good in the long-term.
Dynamic truncation of ADOs.
Fermionic hermiticity relation optimization (this is a priority).