def test_solver_cpp(solver_cpp, parallel, shared_memory):
noexcept = True
if solver_cpp["entry"] == "UCT":
pytest.skip("There is a heap corruption in MCTS solver")
try:
dom = GridDomain()
solver_type = load_registered_solver(solver_cpp["entry"])
solver_args = deepcopy(solver_cpp["config"])
if "parallel" in inspect.signature(solver_type.__init__).parameters:
solver_args["parallel"] = parallel
if ("shared_memory_proxy" in inspect.signature(
solver_type.__init__).parameters and shared_memory):
solver_args["shared_memory_proxy"] = GridShmProxy()
solver_args["domain_factory"] = lambda: GridDomain()
with solver_type(**solver_args) as slv:
GridDomain.solve_with(slv)
plan, cost = get_plan(dom, slv)
except Exception as e:
print(e)
noexcept = False
assert (solver_type.check_domain(dom) and noexcept
and ((not solver_cpp["optimal"]) or parallel or
(cost == 18 and len(plan) == 18)))
def do_test_python(solver_python, result):
noexcept = True
try:
dom = GridDomain()
solver_type = load_registered_solver(solver_python['entry'])
solver_args = solver_python['config']
with solver_type(**solver_args) as slv:
GridDomain.solve_with(slv)
plan, cost = get_plan(dom, slv)
except Exception as e:
print(e)
noexcept = False
result.send(solver_type.check_domain(dom) and noexcept and \
((not solver_python['optimal']) or (cost == 18 and len(plan) == 18)))
result.close()
def test_solve_python(solver_python):
noexcept = True
try:
dom = GridDomain()
solver_type = load_registered_solver(solver_python["entry"])
solver_args = deepcopy(solver_python["config"])
if solver_python["entry"] == "StableBaseline":
solver_args["algo_class"] = PPO
with solver_type(**solver_args) as slv:
GridDomain.solve_with(slv)
plan, cost = get_plan(dom, slv)
except Exception as e:
print(e)
noexcept = False
assert (
solver_type.check_domain(dom)
and noexcept
and ((not solver_python["optimal"]) or (cost == 18 and len(plan) == 18))
)
def do_test_cpp(solver_cpp, parallel, shared_memory, result):
noexcept = True
try:
dom = GridDomain()
solver_type = load_registered_solver(solver_cpp['entry'])
solver_args = solver_cpp['config']
if 'parallel' in inspect.signature(solver_type.__init__).parameters:
solver_args['parallel'] = parallel
if 'shared_memory_proxy' in inspect.signature(solver_type.__init__).parameters and shared_memory:
solver_args['shared_memory_proxy'] = GridShmProxy()
solver_args['domain_factory'] = lambda: GridDomain()
with solver_type(**solver_args) as slv:
GridDomain.solve_with(slv)
plan, cost = get_plan(dom, slv)
except Exception as e:
print(e)
noexcept = False
result.send(solver_type.check_domain(dom) and noexcept and \
((not solver_cpp['optimal']) or parallel or (cost == 18 and len(plan) == 18)))
result.close()
{
'name': 'PPO (deep reinforcement learning)',
'entry': 'StableBaseline',
'config': {
'algo_class': PPO,
'baselines_policy': 'MlpPolicy',
'learn_config': {
'total_timesteps': 30000
},
'verbose': 1
}
}
]
# Load solvers (filtering out badly installed ones)
solvers = map(lambda s: dict(s, entry=load_registered_solver(s['entry'])),
try_solvers)
solvers = list(filter(lambda s: s['entry'] is not None, solvers))
solvers.insert(0, {
'name': 'Random Walk',
'entry': None
}) # Add Random Walk as option
# Run loop to ask user input
domain = Maze()
while True:
# Ask user input to select solver
choice = int(
input('\nChoose a solver:\n{solvers}\n'.format(solvers='\n'.join(
['0. Quit'] +
[f'{i + 1}. {s["name"]}' for i, s in enumerate(solvers)]))))
Value(
cost=sqrt(
(d.num_cols - 1 - s.x) ** 2 + (d.num_rows - 1 - s.y) ** 2
)
),
10000,
),
"parallel": True,
"debug_logs": False,
},
},
]
# Load solvers (filtering out badly installed ones)
solvers = map(
lambda s: dict(s, entry=load_registered_solver(s["entry"])), try_solvers
)
solvers = list(filter(lambda s: s["entry"] is not None, solvers))
# Run loop to ask user input
domain = MyDomain() # MyDomain(5,5)
with tqdm(total=len(solvers) * 100) as pbar:
for s in solvers:
solver_type = s["entry"]
for i in range(50):
s["config"]["shared_memory_proxy"] = None
with solver_type(**s["config"]) as solver:
MyDomain.solve_with(solver) # ,lambda:MyDomain(5,5))
rollout(
domain,
# Copyright (c) AIRBUS and its affiliates.
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
from pprint import pprint
from skdecide.utils import (
get_registered_domains,
get_registered_solvers,
load_registered_domain,
load_registered_solver,
)
if __name__ == "__main__":
print("\nAll registered domains:\n-----------------------")
pprint({d: load_registered_domain(d) for d in get_registered_domains()})
print("\nAll registered solvers:\n-----------------------")
pprint({s: load_registered_solver(s) for s in get_registered_solvers()})
# BFWS (classical planning)
{'name': 'BFWS (planning) - (num_rows * num_cols) binary encoding (1 binary variable <=> 1 cell)',
'entry': 'BFWS',
'need_domain_factory': True,
'config': {'state_features': lambda d, s: d.state_features(s),
'heuristic': lambda d, s: d.heuristic(s),
'termination_checker': lambda d, s: d.is_goal(s),
'parallel': False, 'debug_logs': False}}
]
# Load domains (filtering out badly installed ones)
domains = map(lambda d: dict(d, entry=load_registered_domain(d['entry'])), try_domains)
domains = list(filter(lambda d: d['entry'] is not None, domains))
# Load solvers (filtering out badly installed ones)
solvers = map(lambda s: dict(s, entry=load_registered_solver(s['entry'])), try_solvers)
solvers = list(filter(lambda s: s['entry'] is not None, solvers))
solvers.insert(0, {'name': 'Random Walk', 'entry': None}) # Add Random Walk as option
# Run loop to ask user input
solver_candidates = [s['entry'] for s in solvers if s['entry'] is not None]
while True:
# Ask user input to select domain
domain_choice = int(input('\nChoose a domain:\n{domains}\n'.format(
domains='\n'.join([f'{i + 1}. {d["name"]}' for i, d in enumerate(domains)]))))
selected_domain = domains[domain_choice - 1]
domain_type = selected_domain['entry']
domain = domain_type(**selected_domain['config'])
while True:
# Match solvers compatible with selected domain