def __run(params, single_run, fixed_votes, token):
print(f"Running {token}")
start_time = time.time()
estimator = Estimator(params)
output = estimator.run(single_run, fixed_votes)
r.set(token, output.to_json(orient='records'))
r.set(f"{token}_status", 'DONE')
total_time = time.time() - start_time
print(f"{token} DONE. Time: {total_time} seconds")
class StaticEstimator(Estimator):
"""docstring for ClassName"""
def __init__(self):
super(StaticEstimator, self).__init__()
def update(self):
obArray = obstacleArray()
obArray.names.append("quad_x")
ellipsoidMsg = Ellipsoid()
# fill fake ellipsoid
ellipsoidMsg.center = [1, -1, 2] + rand.random(3) * .1
ellipsoidMsg.shape = np.eye(3).flatten().tolist()
# push back
obArray.ellipsoids.append(ellipsoidMsg)
obArray.names.append("quad_y")
ellipsoidMsg = Ellipsoid()
# fill fake ellipsoid
ellipsoidMsg.center = [1.5, 0, .7] + rand.random(3) * .1
ellipsoidMsg.shape = np.eye(3).flatten().tolist()
# push back
obArray.ellipsoids.append(ellipsoidMsg)
return obArray
if __name__ == '__main__':
Estimator = StaticEstimator()
rospy.loginfo("Estimator: Initalization complete")
while not rospy.is_shutdown():
Estimator.run()
theta - overall proportion of positive items
filters_num - number of filters
filters_select - selectivity of filters (probability of applying a filter)
filters_dif - difficulty of filters
iter_num - number of iterations for averaging results
---------------------
FP == False Exclusion
FN == False Inclusion
'''
if __name__ == '__main__':
params = {
'filters_num': 4,
'items_num': 1000,
'baseround_items': 20,
'items_per_worker': 10,
'votes_per_item': 3,
'filters_select': [0.14, 0.14, 0.28, 0.42],
'filters_dif': [1., 1., 1.1, 0.9],
'worker_tests': 5,
'lr': 5,
'stop_score': 100,
'iter_num': 50,
'z': 0.3,
'theta': 0.3
}
estimator = Estimator(params)
output = estimator.run(True)
output.to_csv('output.csv', index=False)