Defining solvents for transition states inconsistent with molecules about autode

[t-young31]

You are completely correct – I've been meaning to fix this, so now maybe the time! Expect to see this in v1.3.0 😄

---

tsguess.solvent = get_solvent(molecule.solvent, kind='implicit')
molecule = TransitionState(tsguess)

This doesn't work because the solvent gets inherited from the optional 'reactant' argument to TransitionState rather than the ts_guess, which I very much agree is confusing and bad!

[t-young31]

Hi @ffmulks – this has been fixed in #167. If you have time I'd very much appreciate if you could try it out and see if the behaviour matches what you'd expect – thanks 😄

---

working example:

from autode import Atom
from autode.transition_states import TransitionState, TSguess


ts_guess = TSguess(mult=2, solvent_name='water',
                   atoms=[Atom("H", 4.52368, -0.00030,   0.00065),
                          Atom("H", 5.47129,  0.00009,  -0.00023),
                          Atom("H", 6.41149,  0.00045,  -0.00109)])

ts = TransitionState(ts_guess)
assert ts.solvent is not None

ts.optimise()

[t-young31]

see if the behaviour matches what you'd expect

I'm going to hope it has because i'm going to merge #167 now to try and get a release out shortly

[ffmulks]

Weird, that works for me whereas my script in SLURM failed to do the very same thing. I suppose I may just have an environment mixup somewhere.

[t-young31]

👻