added tranposed attribute, used in _read_array

the transposed attribute keeps track of whether the data has been 
transposed, allowing it to be correctly read by _read_array.
Some minor rearrangement of _read_array was necessary to reduce
duplicate code.

Author: John-Colvin

pandas/io/pytables.py

@@ -704,6 +704,7 @@ def _write_array(self, group, key, value):
                                               value.shape,
                                               filters=self.filters)
                 ca[:] = value
+                getattr(group, key)._v_attrs.transposed = True
                 return
 
         if value.dtype.type == np.object_:
@@ -722,6 +723,8 @@ def _write_array(self, group, key, value):
                 getattr(group, key)._v_attrs.shape = value.shape
             else:
                 self.handle.createArray(group, key, value)
+                
+        getattr(group, key)._v_attrs.transposed = True
 
     def _write_table(self, group, items=None, index=None, columns=None,
                      values=None, append=False, compression=None):
@@ -961,23 +964,29 @@ def _read_array(group, key):
     import tables
     node = getattr(group, key)
     data = node[:]
+    attrs = node._v_attrs
+
+    transposed = getattr(attrs, 'transposed', False)
 
     if isinstance(node, tables.VLArray):
-        return data[0]
+        ret = data[0]
     else:
-        attrs = node._v_attrs
-
         dtype = getattr(attrs, 'value_type', None)
         shape = getattr(attrs, 'shape', None)
 
         if shape is not None:
             # length 0 axis
-            return np.empty(shape, dtype=dtype)
+            ret = np.empty(shape, dtype=dtype)
 
         if dtype == 'datetime64':
-            return np.array(data, dtype='M8[ns]')
-        return data
-
+            ret = np.array(data, dtype='M8[ns]')
+        ret = data
+    
+    if transposed == True:
+        return ret.T
+    else:
+        return ret
+    
 def _unconvert_index(data, kind):
     if kind == 'datetime64':
         index = DatetimeIndex(data)