def test_simple_randomized_search():
"""Test a simple reandomized search. Start and goal should be
chosen to be a random, valid GridState because that is
default behavior when they are not provided. As long as the
state space is not disconnected, a path should be returned
and it must include at least the start ang goal
"""
astar = GenericAstar(env, heuristic=OctileGridHeuristic())
path = astar.get_path()
# simplest possible path must include start and goal
# each should have been chosen for expansion from the open list
assert astar.nodes_expanded >= 2
assert len(path) >= 2
def test_simple_shortest_search():
"""Test a search where the start and goal are neighbors. Ensure
correct path and open list sizes. Assumes start is in an
'open area' where all neighbors are valid.
"""
start = GridState(18, 24, valid=True)
goal = GridState(19, 24, valid=True)
astar = GenericAstar(env,
heuristic=OctileGridHeuristic(),
start=start,
goal=goal)
astar.get_path()
print(astar.open)
assert len(astar.path) == 2 # start, goal
assert astar.nodes_expanded == 2 # start, goal
assert len(astar.open) == 7 # start + 8 neighbors - start - goal
def test_undefined_goal():
"""Test all 4 map boundaries <0, >width, <0, >height.
if unchecked, negatives would index backwards into grid.env[][]
"""
bad_goal = GridState(-1, env.get_random().y)
with pytest.raises(StateDoesNotExistError):
GenericAstar(env, heuristic=OctileGridHeuristic(), goal=bad_goal)
bad_goal = GridState(env.get_random().x, -1)
with pytest.raises(StateDoesNotExistError):
GenericAstar(env, heuristic=OctileGridHeuristic(), goal=bad_goal)
bad_goal = GridState(env.width + 1, env.get_random().y)
with pytest.raises(StateDoesNotExistError):
GenericAstar(env, heuristic=OctileGridHeuristic(), goal=bad_goal)
bad_goal = GridState(env.get_random().x, env._height + 1)
with pytest.raises(StateDoesNotExistError):
GenericAstar(env, heuristic=OctileGridHeuristic(), goal=bad_goal)
def speed_test(env: Environment,
search_1: Search.__class__,
search_2: Search.__class__,
start: State = None,
goal: State = None,
heuristic=None) -> Tuple[float, float]:
if start is None:
start = env.get_random(valid=True)
if goal is None:
goal = env.get_random(valid=True)
generic_astar = GenericAstar(env,
heuristic=OctileGridHeuristic(),
start=SearchNode(start),
goal=SearchNode(goal))
gridopt_astar = GridOptimizedAstar(env,
heuristic=OctileGridHeuristic(),
start=SearchNode(start),
goal=SearchNode(goal))
if verbose:
print('\n-----Beginning Speeedtest-----\n')
print(f'environment: {str(env)}')
print(f'start: {str(start)}')
print(f'goal: {str(goal)}')
print(f'heuristic: {str(OctileGridHeuristic())}')
print('\nrunning GenericAstar...')
generic_t1 = time.time()
generic_astar.get_path()
generic_t2 = time.time()
generic_time = generic_t2 - generic_t1
if verbose:
print(f'completed in: ' + str(round(generic_time, 2)) + 's')
print('\nrunning GridOptimizedAstar...')
gridopt_t1 = time.time()
gridopt_astar.get_path()
gridopt_t2 = time.time()
gridopt_time = gridopt_t2 - gridopt_t1
if verbose:
print(f'completed in: ' + str(round(gridopt_time, 2)) + 's')
print(
f"\n{'GenericAstar' if generic_time < gridopt_time else 'GridOptimizedAstar'} was faster!"
)
if not verbose:
print(generic_time + ' ' + gridopt_time)
def get_random_search(
environment: Environment = env,
heuristic: Heuristic = OctileGridHeuristic()
) -> GenericAstar:
start = environment.get_random(valid=True)
goal = environment.get_random(valid=True)
gastar = GenericAstar(environment, heuristic, start=start, goal=goal)
return gastar
def speed_diff():
filepath = os.path.join(
os.path.dirname(os.path.dirname(os.path.abspath(__file__))) +
'/data/maps/large/brc202d.map')
environment = OctileGrid(filepath)
start = environment.get_random(valid=True)
goal = GridState(474, 426)
generic_astar = GenericAstar(environment,
heuristic=OctileGridHeuristic(),
start=SearchNode(start),
goal=SearchNode(goal),
verbose=True)
gridopt_astar = GridOptimizedAstar(environment,
heuristic=OctileGridHeuristic(),
start=SearchNode(start),
goal=SearchNode(goal))
print('\n-----Beginning Speeedtest-----\n')
print(f'environment: {str(environment)}')
print(f'start: {str(start)}')
print(f'goal: {str(goal)}')
print(f'heuristic: {str(OctileGridHeuristic())}')
print('\nrunning GenericAstar...')
generic_t1 = time.time()
generic_astar.get_path()
generic_t2 = time.time()
generic_time = generic_t2 - generic_t1
print(f'completed in: ' + str(round(generic_time, 2)) + 's')
print('\nrunning GridOptimizedAstar...')
gridopt_t1 = time.time()
gridopt_astar.get_path()
gridopt_t2 = time.time()
gridopt_time = gridopt_t2 - gridopt_t1
print(f'completed in: ' + str(round(gridopt_time, 2)) + 's')
print(
f"\n{'GenericAstar' if generic_time < gridopt_time else 'GridOptimizedAstar'} was faster!"
)
def test_invalid_goal():
"""Try to create search with goal node that is not passable"""
bad_goal = env.get_random(valid=False)
with pytest.raises(StateNotValidError):
GenericAstar(env, heuristic=OctileGridHeuristic(), goal=bad_goal)
def pressed_start(self) -> None:
if self.start is None or self.goal is None:
raise Exception('Search started without start / goal')
search = 'gas'
if self.app_frame.combo1.get() == 'Generic Dijkstra':
search = 'gd'
elif self.app_frame.combo1.get() == 'Grid Optimized Dijkstra':
search = 'god'
elif self.app_frame.combo1.get() == 'Grid Optimized A*':
search = 'goas'
elif self.app_frame.combo1.get() == 'BFS':
search = 'bfs'
elif self.app_frame.combo1.get() == 'DFS':
search = 'dfs'
heuristic = OctileGridHeuristic()
if self.app_frame.combo2.get() == 'Manhattan Distance':
heuristic = ManhattanGridHeuristic()
elif self.app_frame.combo2.get() == 'Euclidean Distance':
heuristic = EuclideanGridHeuristic()
if not self.env.is_valid(self.start.state):
raise StateNotValidError(self.start.state)
else:
self.start.state._valid = True
if not self.env.is_valid(self.goal.state):
raise StateNotValidError(self.goal.state)
else:
self.goal.state._valid = True
if search == 'gas':
self.search = GenericAstar(self.env,
start=self.start,
goal=self.goal,
heuristic=heuristic,
verbose=self.verbose)
elif search == 'gd':
self.search = GenericDijkstra(self.env,
start=self.start,
goal=self.goal,
verbose=self.verbose)
elif search == 'god':
self.search = GridOptimizedDijkstra(self.env,
start=self.start,
goal=self.goal,
verbose=self.verbose)
elif search == 'goas':
self.search = GridOptimizedAstar(self.env,
start=self.start,
goal=self.goal,
heuristic=heuristic,
verbose=self.verbose)
elif search == 'bfs':
self.search = GenericBFS(self.env,
start=self.start,
goal=self.goal,
verbose=self.verbose)
elif search == 'dfs':
self.search = GenericDFS(self.env,
start=self.start,
goal=self.goal,
verbose=self.verbose)
self.draw_initial()
self.after(1000, self.draw_step)