这是我的代码：
from IPython.parallel import Client
from sklearn.datasets import load_digits
def mytask(data, labels, id):
# ...
pass
engines = Client()
bview = engines.load_balanced_view()
bview.block = False
digits = load_digits()
X, y = digits.data, digits.target
job = bview.apply(mytask, X, y, 1)
while not job.ready(): # line 242
time.sleep(2)
print job.result
偶尔使用相同的输入我的代码会失败：
Traceback (most recent call last):
File "task.py", line 242, in <module>
while not job.ready():
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/asyncresult.py", line 111, in ready
self.wait(0)
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/asyncresult.py", line 121, in wait
self._ready = self._client.wait(self.msg_ids, timeout)
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/client.py", line 844, in wait
self.spin()
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/client.py", line 799, in spin
self._flush_results(self._task_socket)
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/client.py", line 692, in _flush_results
handler(msg)
File "/usr/lib/python2.7/dist-packages/IPython/parallel/client/client.py", line 657, in _handle_apply_reply
self.results[msg_id] = util.unserialize_object(msg['buffers'])[0]
File "/usr/lib/python2.7/dist-packages/IPython/parallel/util.py", line 262, in unserialize_object
return uncanSequence(map(unserialize, sobj)), bufs
File "/usr/lib/python2.7/dist-packages/IPython/utils/newserialized.py", line 177, in unserialize
return UnSerializeIt(serialized).getObject()
File "/usr/lib/python2.7/dist-packages/IPython/utils/newserialized.py", line 161, in getObject
result = numpy.frombuffer(buf, dtype = self.serialized.metadata['dtype'])
ValueError: offset must be non-negative and smaller than buffer lenth (0)
这似乎与我的代码无关。我不确定会出现什么问题。