[BUG] Ray executor run inv_mapper raises ValueError: assignment destination is read-only

[fyrestone]

Describe the bug A clear and concise description of what the bug is.

__________________________ test_label_encoder[int64] ___________________________

setup = <mars.deploy.oscar.session.SyncSession object at 0x337a3c7f0>
values = array([2, 1, 3, 1, 3]), classes = array([1, 2, 3])
unknown = array([4])

    @pytest.mark.parametrize(
        "values, classes, unknown",
        [
            (
                np.array([2, 1, 3, 1, 3], dtype="int64"),
                np.array([1, 2, 3], dtype="int64"),
                np.array([4], dtype="int64"),
            ),
            (
                np.array(["b", "a", "c", "a", "c"], dtype=object),
                np.array(["a", "b", "c"], dtype=object),
                np.array(["d"], dtype=object),
            ),
            (
                np.array(["b", "a", "c", "a", "c"]),
                np.array(["a", "b", "c"]),
                np.array(["d"]),
            ),
        ],
        ids=["int64", "object", "str"],
    )
    def test_label_encoder(setup, values, classes, unknown):
        # Test LabelEncoder's transform, fit_transform and
        # inverse_transform methods
        values_t = mt.tensor(values)
    
        le = LabelEncoder()
        le.fit(values_t)
        assert_array_equal(le.classes_.fetch(), classes)
        assert_array_equal(le.transform(values_t).fetch(), [1, 0, 2, 0, 2])
        assert_array_equal(le.inverse_transform(mt.tensor([1, 0, 2, 0, 2])).fetch(), values)
    
        le = LabelEncoder()
>       ret = le.fit_transform(values)

mars/learn/preprocessing/tests/test_label.py:300: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
mars/learn/preprocessing/_label.py:122: in fit_transform
    self.classes_, y = execute_tileable(
mars/deploy/oscar/session.py:1888: in execute
    return session.execute(
mars/deploy/oscar/session.py:1682: in execute
    execution_info: ExecutionInfo = fut.result(
../../.pyenv/versions/3.8.13/lib/python3.8/concurrent/futures/_base.py:444: in result
    return self.__get_result()
../../.pyenv/versions/3.8.13/lib/python3.8/concurrent/futures/_base.py:389: in __get_result
    raise self._exception
mars/deploy/oscar/session.py:1868: in _execute
    await execution_info
../../.pyenv/versions/3.8.13/lib/python3.8/asyncio/tasks.py:695: in _wrap_awaitable
    return (yield from awaitable.__await__())
mars/deploy/oscar/session.py:105: in wait
    return await self._aio_task
mars/deploy/oscar/session.py:953: in _run_in_background
    raise task_result.error.with_traceback(task_result.traceback)
mars/services/task/supervisor/processor.py:372: in run
    await self._process_stage_chunk_graph(*stage_args)
mars/services/task/supervisor/processor.py:250: in _process_stage_chunk_graph
    chunk_to_result = await self._executor.execute_subtask_graph(
mars/services/task/execution/ray/executor.py:551: in execute_subtask_graph
    meta_list = await asyncio.gather(*output_meta_object_refs)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

awaitable = ObjectRef(b2d1bf24c5f98f84ffffffffffffffffffffffff0100000001000000)

    @types.coroutine
    def _wrap_awaitable(awaitable):
        """Helper for asyncio.ensure_future().
    
        Wraps awaitable (an object with __await__) into a coroutine
        that will later be wrapped in a Task by ensure_future().
        """
>       return (yield from awaitable.__await__())
E       ray.exceptions.RayTaskError(ValueError): ray::execute_subtask() (pid=97485, ip=127.0.0.1)
E         File "/home/admin/mars/mars/services/task/execution/ray/executor.py", line 185, in execute_subtask
E           execute(context, chunk.op)
E         File "/home/admin/mars/mars/core/operand/core.py", line 491, in execute
E           result = executor(results, op)
E         File "/home/admin/mars/mars/core/custom_log.py", line 94, in wrap
E           return func(cls, ctx, op)
E         File "/home/admin/mars/mars/utils.py", line 1160, in wrapped
E           return func(cls, ctx, op)
E         File "/home/admin/mars/mars/tensor/base/map_chunk.py", line 170, in execute
E           ctx[op.outputs[0].key] = op.func(in_data, *args, **kwargs)
E         File "/home/admin/mars/mars/learn/utils/_encode.py", line 72, in inv_mapper
E           c[c > idx] = idx
E       ValueError: assignment destination is read-only

../../.pyenv/versions/3.8.13/lib/python3.8/asyncio/tasks.py:695: RayTaskError(ValueError)

To Reproduce To help us reproducing this bug, please provide information below:

1. Your Python version
2. The version of Mars you use
3. Versions of crucial packages, such as numpy, scipy and pandas
4. Full stack of the error.
5. Minimized code to reproduce the error.

Expected behavior A clear and concise description of what you expected to happen.

Additional context Add any other context about the problem here.