Allow time_col type to be string

pandas_td/td.py

@@ -336,8 +336,8 @@ def to_td(frame, name, con, if_exists='fail', time_col=None, time_index=None, in
         - append: If table exists, insert data. Create if does not exist.
     time_col : string, optional
         Column name to use as "time" column for the table. Column type must be
-        integer (unixtime) or datetime. If None is given (default), then the current
-        time is used as time values.
+        integer (unixtime), datetime, or string. If None is given (default),
+        then the current time is used as time values.
     time_index : int, optional
         Level of index to use as "time" column for the table. Set 0 for a single index.
         This parameter implies index=False.
@@ -392,15 +392,13 @@ def _convert_dataframe(frame, time_col=None, time_index=None, index=None, index_
     if 'time' in frame.columns and time_col != 'time':
         raise ValueError('"time" column already exists')
     if time_col is not None:
-        col = frame[time_col]
-        if col.dtype not in (np.dtype('int64'), np.dtype('datetime64[ns]')):
-            raise TypeError('time type must be either int64 or datetime64')
+        if time_col != 'time':
+            frame.rename(columns={time_col: 'time'}, inplace=True)
+        col = frame['time']
         if col.dtype != np.dtype('int64'):
+            if col.dtype != np.dtype('datetime64[ns]'):
+                col = pd.to_datetime(col)
             frame['time'] = col.astype(np.int64, copy=True) // (10 ** 9)
-            if time_col != 'time':
-                frame.drop(time_col, axis=1, inplace=True)
-        elif time_col != 'time':
-            frame.rename(columns={time_col: 'time'}, inplace=True)
     elif time_index is not None:
         if type(time_index) is bool or not isinstance(time_index, six.integer_types):
             raise TypeError('invalid type for time_index')

pandas_td/test_td.py

@@ -399,26 +399,26 @@ def test_time_now(self):
         f2 = _convert_dataframe(f1)
         eq_(list(f2.columns), ['x', 'y', 'time'])
 
+    def test_time_col_rename(self):
+        f1 = pd.DataFrame([[0, 'a', 1], [0, 'b', 2]], columns=['unixtime', 'x', 'y'])
+        f2 = _convert_dataframe(f1, time_col='unixtime')
+        eq_(list(f2.columns), ['time', 'x', 'y'])
+
     def test_time_col_by_unixtime(self):
         f1 = pd.DataFrame([[0, 'a', 1], [0, 'b', 2]], columns=['time', 'x', 'y'])
         f2 = _convert_dataframe(f1, time_col='time')
         eq_(list(f2.columns), ['time', 'x', 'y'])
 
-    def test_time_col_by_unixtime_rename(self):
-        f1 = pd.DataFrame([[0, 'a', 1], [0, 'b', 2]], columns=['unixtime', 'x', 'y'])
-        f2 = _convert_dataframe(f1, time_col='unixtime')
-        eq_(list(f2.columns), ['time', 'x', 'y'])
-
     def test_time_col_by_datetime(self):
         f1 = pd.DataFrame([['a', 1], ['b', 2]], columns=['x', 'y'])
         f1['time'] = pd.to_datetime('2001-01-01')
         f2 = _convert_dataframe(f1, time_col='time')
         eq_(list(f2.columns), ['x', 'y', 'time'])
 
-    def test_time_col_by_datetime_rename(self):
+    def test_time_col_by_string(self):
         f1 = pd.DataFrame([['a', 1], ['b', 2]], columns=['x', 'y'])
-        f1['date'] = pd.to_datetime('2001-01-01')
-        f2 = _convert_dataframe(f1, time_col='date')
+        f1['time'] = '2001-01-01'
+        f2 = _convert_dataframe(f1, time_col='time')
         eq_(list(f2.columns), ['x', 'y', 'time'])
 
     @raises(TypeError)