def il_max_avg_obstacle_clearance():
name = f'samples/{inspect.currentframe().f_code.co_name}.pkl'
sampler.sample(N=10000, alpha=0.1, prior_file='samples/il_prior.pkl', N_sigma=1000,
behavior_func=behavior.obstacle_clearance,
env=environment.RBF2dGymEnv(use_lidar=True), env_kernel=kernel.RBF2dEnvKernelNormal(),
controller=controller.ILController(), controller_kernel=kernel.TransitionKernel(),
target_type='maximal', save=name)
def il_min_abs_jerkiness():
name = f'samples/{inspect.currentframe().f_code.co_name}.pkl'
sampler.sample(N=10000, alpha=0.1, prior_file='samples/il_prior.pkl', N_sigma=1000,
behavior_func=behavior.min_jerkiness,
env=environment.RBF2dGymEnv(use_lidar=True), env_kernel=kernel.RBF2dEnvKernelNormal(),
controller=controller.ILController(), controller_kernel=kernel.TransitionKernel(),
target_type='match', target_behavior=0, save=name)
def il_min_legibility(prior_file='samples/il_prior.pkl'):
name = f'samples/{inspect.currentframe().f_code.co_name}.pkl'
sampler.sample(N=10000, alpha=0.1, prior_file=prior_file, N_sigma=1000,
behavior_func=behavior.neg_behavior(behavior.legibility),
env=environment.RBF2dGymEnv(use_lidar=True), env_kernel=kernel.RBF2dEnvKernelNormal(),
controller=controller.ILController(), controller_kernel=kernel.TransitionKernel(),
target_type='maximal', save=name)
def ds_min_avg_obstacle_clearance():
name = f'samples/{inspect.currentframe().f_code.co_name}.pkl'
sampler.sample(N=10000, alpha=0.1, prior_file='samples/ds_prior.pkl', N_sigma=1000,
behavior_func=behavior.obstacle_clearance,
env=environment.RBF2dGymEnv(time_limit=500, oob_termination=False, use_lidar=False),
env_kernel=kernel.RBF2dEnvKernelNormal(),
controller=controller.DSController(), controller_kernel=kernel.TransitionKernel(),
target_type='match', target_behavior=0, save=name)
