Bug when simulating an action leading to a diverging powerflow with LightSim

[marota]

Environment

• Grid2op version: 1.6.0
• System: osx
• Additional system information

Bug description

Bug when simulating an action leading to a diverging powerflow with LightSim.

How to reproduce

Code snippet

import grid2op
# Backend class to use
from lightsim2grid.LightSimBackend import LightSimBackend
BACKEND = LightSimBackend
#from grid2op.Backend import PandaPowerBackend
#BACKEND = PandaPowerBackend

env_path='/YourPath/L2RPN_icaps2021/input_data_val'

env = grid2op.make(env_path,backend=BACKEND(),difficulty="competition")

final_dict={'generators_id': [(19, 1)],
 'loads_id': [(30, 2)],
 'lines_or_id': [(58, 1)],
 'lines_ex_id': [(46, 1), (47, 2)]}
action=env.action_space({"set_bus": final_dict})

obs=env.get_obs()
obs.simulate(action,time_step=0)

Current output

File "grid2op/Observation/CompleteObservation.py", line 206, in update
self._shunt_p[:] = sh_p
ValueError: could not broadcast input array from shape (0) into shape (6)

When doing env.step(action) your get the following:

(<grid2op.Space.GridObjects.CompleteObservation_input_data_val at 0x7f815e84aa60>,
-1.0,
True,
{'disc_lines': array([-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, 1, 1, -1, -1, -1, -1, -1, -1, -1, 1, 1, 1,
1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0,
0, -1, -1, -1, -1, -1, -1, -1], dtype=int32),
'is_illegal': False,
'is_ambiguous': False,
'is_dispatching_illegal': False,
'is_illegal_reco': False,
'reason_alarm_illegal': None,
'opponent_attack_line': None,
'opponent_attack_sub': None,
'opponent_attack_duration': 0,
'exception': [Grid2OpException DivergingPowerFlow DivergingPowerFlow('divergence of powerflow (non connected grid)')],
'rewards': {}})

Or doing the same simulate but with pandapower you get:

(<grid2op.Space.GridObjects.CompleteObservation_input_data_val at 0x7fd3ab019eb0>,
-1.0,
True,
{'disc_lines': array([-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, 1, 1, -1, -1, -1, -1, -1, -1, -1, 1, 1, 1,
1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0,
0, -1, -1, -1, -1, -1, -1, -1], dtype=int32),
'is_illegal': False,
'is_ambiguous': False,
'is_dispatching_illegal': False,
'is_illegal_reco': False,
'reason_alarm_illegal': None,
'opponent_attack_line': None,
'opponent_attack_sub': None,
'opponent_attack_duration': 0,
'exception': [Grid2OpException DivergingPowerFlow DivergingPowerFlow('powerflow diverged with error :"Power Flow nr did not converge after 10 iterations!"')],
'rewards': {}})

Expected output

Would expect the same kind of output as pandapower

[BDonnot]

Should be fixed in the latest grid2op release (version 1.6.0.post1)