Merge branch 'tst_excel_duplicated_cols' of https://github.com/jmcnamara/pandas into jmcnamara-tst_excel_duplicated_cols

Author: jreback

doc/source/release.rst

@@ -615,6 +615,8 @@ Bug Fixes
     the original ordering (:issue:`4621`).
   - Fixed ``Period`` with a business date freq to always roll-forward if on a
     non-business date. (:issue:`5203`)
+  - Fixed bug in Excel writers where frames with duplicate column names weren't
+    written correctly. (:issue:`5235`)
 
 pandas 0.12.0
 -------------

pandas/core/format.py

@@ -1371,8 +1371,8 @@ def _format_regular_rows(self):
             for idx, idxval in enumerate(index_values):
                 yield ExcelCell(self.rowcounter + idx, 0, idxval, header_style)
 
-        for colidx, colname in enumerate(self.columns):
-            series = self.df[colname]
+        for colidx in range(len(self.columns)):
+            series = self.df.iloc[:, colidx]
             for i, val in enumerate(series):
                 yield ExcelCell(self.rowcounter + i, colidx + coloffset, val)
 
@@ -1408,8 +1408,8 @@ def _format_hierarchical_rows(self):
                                     indexcolval, header_style)
                 gcolidx += 1
 
-        for colidx, colname in enumerate(self.columns):
-            series = self.df[colname]
+        for colidx in range(len(self.columns)):
+            series = self.df.iloc[:, colidx]
             for i, val in enumerate(series):
                 yield ExcelCell(self.rowcounter + i, gcolidx + colidx, val)
 

pandas/io/tests/test_excel.py

@@ -874,6 +874,24 @@ def roundtrip(df, header=True, parser_hdr=0):
         self.assertEqual(res.shape, (1, 2))
         self.assertTrue(res.ix[0, 0] is not np.nan)
 
+    def test_duplicated_columns(self):
+        # Test for issue #5235.
+        _skip_if_no_xlrd()
+        ext = self.ext
+        path = '__tmp_to_excel_duplicated_columns__.' + ext
+
+        with ensure_clean(path) as path:
+            write_frame = DataFrame([[1, 2, 3], [1, 2, 3], [1, 2, 3]])
+            colnames = ['A', 'B', 'B']
+
+            write_frame.columns = colnames
+            write_frame.to_excel(path, 'test1')
+
+            read_frame = read_excel(path, 'test1').astype(np.int64)
+            read_frame.columns = colnames
+
+            tm.assert_frame_equal(write_frame, read_frame)
+
 
 class OpenpyxlTests(ExcelWriterBase, unittest.TestCase):
     ext = 'xlsx'