Source code for sbp_env.samplers.prmSampler

from overrides import overrides

from ..samplers.randomPolicySampler import RandomPolicySampler
from ..utils import planner_registry


[docs]class PRMSampler(RandomPolicySampler): r"""This sampler is basically the same as :class:`samplers.randomPolicySampler.RandomPolicySampler`, excepts we always overrides the goal bias epsilon :math:`\epsilon=0` as PRM could not utilise the benefit of goal bias. """ def __init__(self, **kwargs): super().__init__(**kwargs)
[docs] @overrides def init(self, **kwargs): """The delayed **initialisation** method""" kwargs["goalBias"] = 0 super().init(**kwargs)
# start register sampler_id = "prm_sampler" planner_registry.register_sampler( sampler_id, sampler_class=PRMSampler, ) # finish register