Pandas 在 DatetimeIndex TypeError 上合并：“NoneType”类型的对象没有 len()

Question

举个非常简单的例子：

import pandas as pd
import numpy as np
import datetime

base = datetime.datetime(2016, 10, 1)
date_list = [base - datetime.timedelta(days=x) for x in range(0, 100)]

df1 = pd.DataFrame(np.random.randint(0,100,size=(100, 4)), columns=list('ABCD'), index = date_list)
df2 = pd.DataFrame(np.random.randint(0,100,size=(100, 4)), columns=list('ABCD'), index = date_list)

pd.merge(df1, df2, how = 'outer', left_on = True)

返回错误，TypeError: 'NoneType' 类型的对象没有 len()。如果我想在索引上合并这两个 DataFrame，它是相同的 DatetimeIndex，我是否错过了合并应该如何工作？

我正在运行 Python 2.7.12、Pandas 0.18.1 和 Numpy 1.11.1

完整的回溯是：

TypeError                                 Traceback (most recent call last)
<ipython-input-1-3174c0ff542d> in <module>()
      9 df2 = pd.DataFrame(np.random.randint(0,100,size=(100, 4)), columns=list('ABCD'), index = date_list)
     10
---> 11 pd.merge(df1, df2, how = 'outer', left_on = True)

/Users/user/anaconda/lib/python2.7/site-packages/pandas/tools/merge.pyc in merge(left, right, how, on, left_on, right_on, left_index, right_index, sort, suffixes, copy, indicator)
     36                          right_on=right_on, left_index=left_index,
     37                          right_index=right_index, sort=sort, suffixes=suffixes,
---> 38                          copy=copy, indicator=indicator)
     39     return op.get_result()
     40 if __debug__:

/Users/user/anaconda/lib/python2.7/site-packages/pandas/tools/merge.pyc in __init__(self, left, right, how, on, left_on, right_on, axis, left_index, right_index, sort, suffixes, copy, indicator)
    208         (self.left_join_keys,
    209          self.right_join_keys,
--> 210          self.join_names) = self._get_merge_keys()
    211
    212     def get_result(self):

/Users/user/anaconda/lib/python2.7/site-packages/pandas/tools/merge.pyc in _get_merge_keys(self)
    405         left_keys, right_keys
    406         """
--> 407         self._validate_specification()
    408
    409         left_keys = []

/Users/user/anaconda/lib/python2.7/site-packages/pandas/tools/merge.pyc in _validate_specification(self)
    521                                      'of levels in the index of "left"')
    522                 self.left_on = [None] * n
--> 523         if len(self.right_on) != len(self.left_on):
    524             raise ValueError("len(right_on) must equal len(left_on)")
    525

TypeError: object of type 'NoneType' has no len()

Answer 1

在documentation 中声明：“left_on”可以是“标签或列表，或类似数组”当您传递“True”时，会出现错误。 如果您只是省略“left_on”，它似乎工作正常。

我是不是误解了这个问题？

也许你确实想这样做：

pd.merge(df1, df2, how = 'outer', left_index = True, right_index=True)

这会导致

    A_x B_x C_x D_x A_y B_y C_y D_y
2016-10-01  99  9   89  27  2   10  63  44
2016-09-30  42  74  58  87  33  56  83  72
2016-09-29  89  41  89  94  75  66  74  17
2016-09-28  53  42  4   83  84  48  2   36
2016-09-27  81  97  1   14  86  27  49  53