Fix DataFrame.from_records error message

Make `create_block_manager_from_arrays` error message use the correct dimensions.  This is different from the situation in `create_block_manager_from_blocks` because in that method the blocks are concatenated along their first dimension, whereas in this method the arrays are concatenated along a new dimension.

Also fix `test_from_records_to_records`, which specified a misleading error message when the shape of the input data did not match the shape implied by the indices.  In particular, the error message omitted the offending dimension entirely!

Author: ischwabacher

pandas/core/internals.py

@@ -3507,7 +3507,7 @@ def create_block_manager_from_arrays(arrays, names, axes):
         mgr._consolidate_inplace()
         return mgr
     except (ValueError) as e:
-        construction_error(len(arrays), arrays[0].shape[1:], axes, e)
+        construction_error(len(arrays), arrays[0].shape, axes, e)
 
 
 def form_blocks(arrays, names, axes):

pandas/tests/test_frame.py

@@ -4032,7 +4032,7 @@ def test_from_records_to_records(self):
         tm.assert_frame_equal(DataFrame.from_records(arr2), DataFrame(arr2))
 
         # wrong length
-        msg = r'Shape of passed values is \(3,\), indices imply \(3, 1\)'
+        msg = r'Shape of passed values is \(3, 2\), indices imply \(3, 1\)'
         with assertRaisesRegexp(ValueError, msg):
             DataFrame.from_records(arr, index=index[:-1])