与 ``group_by`` 变换类似,但使用默认的key/value提取函数对元素分组 Args: **options: 可配置参数 Returns: PTable: 分组结果 >>> _pipeline.parallelize([("A", 4), ("A", 3), ("B", 2), ("A", 1)]).group_by_key().get() {"A": [4, 3, 1], "B": [2]}
(self, **options)
source not stored for this graph (policy: none)
no outgoing calls