我认为您需要

I think you need numpy.in1d:

data[np.in1d(data.date, dateref)] 

示例:

import numpy as np
import datetime as datetime

test = np.array([datetime.date(2007, 4, 10), 
                 datetime.date(2007, 5, 10), 
                 datetime.date(2007, 6, 10)])

states = [datetime.date(2007, 4, 10), 
          datetime.date(2007, 5, 10)]

mask = np.in1d(test, states)
print mask
[ True  True False]

print test[mask]
[datetime.date(2007, 4, 10) datetime.date(2007, 5, 10)]