Selecting multiple columns from DataFrame with duplicate column labels failure.

[lodagro]

In [25]: df = pandas.DataFrame(np.random.randn(4,4), columns=list('AABC'))

In [26]: df
Out[26]: 
          A         A         B         C
0 -0.174905  0.332522  1.134984 -0.201270
1  1.730445  0.382556 -0.607761  1.221815
2  0.513049  0.196231 -1.746732 -0.252282
3 -0.297577 -1.000121 -0.090442 -2.129467

In [27]: df.ix[:,['A', 'B']]
Out[27]: 
          A         A         B
0 -0.174905  0.332522  1.134984
1  1.730445  0.382556 -0.607761
2  0.513049  0.196231 -1.746732
3 -0.297577 -1.000121 -0.090442

In [28]: df[['A', 'B']]
---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
...
Exception: Reindexing only valid with uniquely valued Index objects

In [29]: df[['B', 'C']]
---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
...
Exception: Reindexing only valid with uniquely valued Index objects

[wesm]

It won't reorder, but it will at least work

[jankatins]

I still get this error with a long dataframe (pandas.version == '0.10.0'). Renaming the column makes it work. Unfortuantelly I can't produce a small reproduceable example: using the above code it simple works :-(

In [85]: journals = journals[["sid", "title",  "SNIP2_2009" , "SJR2_2009", "SNIP2_2010", "SJR2_2010", "SNIP2_2011", "SJR2_2011"]]
---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
<ipython-input-85-7a986a34fc84> in <module>()
----> 1 journals = journals[["sid", "title",  "SNIP2_2009" , "SJR2_2009", "SNIP2_2010", "SJR2_2010", "SNIP2_2011", "SJR2_2011"]]

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in __getitem__(self, key)
   1932             if com._is_bool_indexer(key):
   1933                 key = np.asarray(key, dtype=bool)
-> 1934             return self._getitem_array(key)
   1935         elif isinstance(self.columns, MultiIndex):
   1936             return self._getitem_multilevel(key)

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in _getitem_array(self, key)
   1968                         raise KeyError("No column(s) named: %s" %
   1969                                        com.pprint_thing(k))
-> 1970                 return self.take(mask.nonzero()[0], axis=1)
   1971 
   1972     def _slice(self, slobj, axis=0):

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in take(self, indices, axis)
   2850             else:
   2851                 new_columns = self.columns.take(indices)
-> 2852                 return self.reindex(columns=new_columns)
   2853         else:
   2854             new_values = com.take_2d(self.values,

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in reindex(self, index, columns, method, level, fill_value, limit, copy)
   2505         if columns is not None:
   2506             frame = frame._reindex_columns(columns, copy, level,
-> 2507                                            fill_value, limit)
   2508 
   2509         if index is not None:

C:\portabel\Python27\lib\site-packages\pandas\core\frame.pyc in _reindex_columns(self, new_columns, copy, level, fill_value, limit)
   2592                          limit=None):
   2593         new_columns, indexer = self.columns.reindex(new_columns, level=level,
-> 2594                                                     limit=limit)
   2595         return self._reindex_with_indexers(None, None, new_columns, indexer,
   2596                                            copy, fill_value)

C:\portabel\Python27\lib\site-packages\pandas\core\index.pyc in reindex(self, target, method, level, limit)
    873             else:
    874                 indexer = self.get_indexer(target, method=method,
--> 875                                            limit=limit)
    876         return target, indexer
    877 

C:\portabel\Python27\lib\site-packages\pandas\core\index.pyc in get_indexer(self, target, method, limit)
    792 
    793         if not self.is_unique:
--> 794             raise Exception('Reindexing only valid with uniquely valued Index '
    795                             'objects')
    796 

Exception: Reindexing only valid with uniquely valued Index objects

[changhiskhan]

looks like your data is mixed type but the original issue is not.

[changhiskhan]

@wesm I moved this to 0.10.2

Not enough time for 0.10.1

[jreback]

pushing to 0.12 when can deal with this in block manager, see #3012

[jreback]

this was actually fixed in 0.9.0 and test for it in test_frames...so closing...UOD (unless otherwise directed)