我正在尝试合并一些由于某些大文本文件而获得的CSV文件。格式是这样的：
Key, Value
002 C000000,NY
002 D010000,11788
003 000000,N
004 000000,Y
005 000000,N
006 000000,N
007C0100,5
007C0200,XTF ADVISORS TRUST - ETF 2030 PORTFOLIO
007C0300,Y
0240000,N
025D0001,0
025D0002,0
025D0003,0
025D0004,0
025D0005,0
025D0006,0
025D0007,0
025D0008,0
028A0100,10
028A0200,0
028A0300,0
028A0400,5
028B0100,28
028B0200,0
028B0300,0
028B0400,54
028C0100,1
028C0200,0
028C0300,0
028C0400,6
028D0100,20
028D0200,0
028D0300,0
我有一个hacky方法，我不确定问题是否与我的方法（内存？）或不是，但这里是将所有这些与键，值在同一行中合并的代码：
def mergeCsv(filename1,seriesnum):
firstfile = pd.read_csv(filename1 + "_1.csv")
for i in xrange(2,int(seriesnum)):
eachfile = pd.read_csv(filename1 + "_" + str(i) + ".csv")
merged = firstfile.merge(eachfile, on='Key', left_index=False, how='inner') #also tried how='left' and how='outer'
merged.to_csv("result.csv", index=False, na_rep='NA')
firstfile = pd.read_csv("result.csv")
6 csv'加入后，我收到以下错误（在这种情况下有16个文件）
File "./parser_nsar.py", line 165, in <module>
mergeCsv(hardcodename, 16)
File "./parser_nsar.py", line 112, in mergeCsv
merged = firstfile.merge(eachfile, on='Key', left_index=False, how='inner')
File "/Library/Python/2.7/site-packages/pandas/core/frame.py", line 3632, in merge
suffixes=suffixes, copy=copy)
File "/Library/Python/2.7/site-packages/pandas/tools/merge.py", line 40, in merge
return op.get_result()
File "/Library/Python/2.7/site-packages/pandas/tools/merge.py", line 189, in get_result
ldata, rdata = self._get_merge_data()
File "/Library/Python/2.7/site-packages/pandas/tools/merge.py", line 284, in _get_merge_data
copydata=False)
File "/Library/Python/2.7/site-packages/pandas/core/internals.py", line 3439, in _maybe_rename_join
to_rename = self.items.intersection(other.items)
File "/Library/Python/2.7/site-packages/pandas/core/index.py", line 962, in intersection
indexer = self.get_indexer(other.values)
File "/Library/Python/2.7/site-packages/pandas/core/index.py", line 1120, in get_indexer
raise InvalidIndexError('Reindexing only valid with uniquely'
pandas.core.index.InvalidIndexError: Reindexing only valid with uniquely valued Index objects
所以到目前为止我有这个结果（这正是我想要的，但对于所有的csvs）：
006 000000,N,N,N,N,N,N
007C0100,1,2,3,4,5,6
007C0200,XTF ADVISORS TRUST - ETF 2010 PORTFOLIO,XTF ADVISORS TRUST - ETF 2015 PORTFOLIO,XTF ADVISORS TRUST - ETF 2020 PORTFOLIO,XTF ADVISORS TRUST - ETF 2025 PORTFOLIO,XTF ADVISORS TRUST - ETF 2030 PORTFOLIO,XTF ADVISORS TRUST - ETF 2040 PLUS PORTFOLIO
007C0300,Y,Y,Y,Y,Y,Y
0240000,N,N,N,N,N,N
025D0001,0,0,0,0,0,0
025D0002,0,0,0,0,0,0
025D0003,0,0,0,0,0,0
025D0004,0,0,0,0,0,0
025D0005,0,0,0,0,0,0
025D0006,0,0,0,0,0,0
025D0007,0,0,0,0,0,0
025D0008,0,0,0,0,0,0
028A0100,0,472,4,49,10,30
028A0200,0,0,0,0,0,0
028A0300,0,0,0,0,0,0
028A0400,1,3,2,10,5,5
028B0100,1,302,196,107,28,93
028B0200,0,0,0,0,0,0
028B0300,0,0,0,0,0,0
028B0400,10,16,22,80,54,10
028C0100,27,0,23,10,1,23
028C0200,0,0,0,0,0,0
028C0300,0,0,0,0,0,0
028C0400,1,27,180,171,6,60