Merge pull request #21 from dkillick/nd_sample_point

Add convert support for nd arrays

Author: bjlittle

nc_time_axis/__init__.py

@@ -226,17 +226,24 @@ def default_units(cls, sample_point, axis):
         Computes some units for the given data point.
 
         """
-        try:
-            # Try getting the first item. Otherwise we just use this item.
-            sample_point = sample_point[0]
-        except (TypeError, IndexError):
-            pass
-
-        if not hasattr(sample_point, 'calendar'):
-            msg = 'Expecting netcdftimes with an extra "calendar" attribute.'
-            raise ValueError(msg)
-
-        return sample_point.calendar, cls.standard_unit
+        if hasattr(sample_point, '__iter__'):
+            # Deal with nD `sample_point` arrays.
+            if isinstance(sample_point, np.ndarray):
+                sample_point = sample_point.reshape(-1)
+            calendars = np.array([point.calendar for point in sample_point])
+            if np.all(calendars[0] == calendars):
+                calendar = calendars[0]
+            else:
+                raise ValueError('Calendar units are not all equal.')
+        else:
+            # Deal with a single `sample_point` value.
+            if not hasattr(sample_point, 'calendar'):
+                msg = ('Expecting netcdftimes with an extra '
+                       '"calendar" attribute.')
+                raise ValueError(msg)
+            else:
+                calendar = sample_point.calendar
+        return calendar, cls.standard_unit
 
     @classmethod
     def convert(cls, value, unit, axis):
@@ -245,11 +252,13 @@ def convert(cls, value, unit, axis):
         with :func:`netcdftime.utime().date2num`.
 
         """
+        shape = None
         if isinstance(value, np.ndarray):
             # Don't do anything with numeric types.
             if value.dtype != np.object:
                 return value
-
+            shape = value.shape
+            value = value.reshape(-1)
             first_value = value[0]
         else:
             # Don't do anything with numeric types.
@@ -270,7 +279,11 @@ def convert(cls, value, unit, axis):
         if isinstance(value, CalendarDateTime):
             value = [value]
 
-        return ut.date2num([v.datetime for v in value])
+        result = ut.date2num([v.datetime for v in value])
+        if shape is not None:
+            result = result.reshape(shape)
+
+        return result
 
 
 # Automatically register NetCDFTimeConverter with matplotlib.unit's converter

nc_time_axis/tests/unit/test_NetCDFTimeConverter.py

@@ -24,20 +24,55 @@ def test_axis_default_limits(self):
 
 
 class Test_default_units(unittest.TestCase):
-    def test_360_day_calendar(self):
+    def test_360_day_calendar_point(self):
+        calendar = '360_day'
+        unit = 'days since 2000-01-01'
+        val = CalendarDateTime(netcdftime.datetime(2014, 8, 12), calendar)
+        result = NetCDFTimeConverter().default_units(val, None)
+        self.assertEqual(result, (calendar, unit))
+
+    def test_360_day_calendar_list(self):
         calendar = '360_day'
         unit = 'days since 2000-01-01'
         val = [CalendarDateTime(netcdftime.datetime(2014, 8, 12), calendar)]
         result = NetCDFTimeConverter().default_units(val, None)
         self.assertEqual(result, (calendar, unit))
 
+    def test_360_day_calendar_nd(self):
+        # Test the case where the input is an nd-array.
+        calendar = '360_day'
+        unit = 'days since 2000-01-01'
+        val = np.array([[CalendarDateTime(netcdftime.datetime(2014, 8, 12),
+                                          calendar)],
+                       [CalendarDateTime(netcdftime.datetime(2014, 8, 13),
+                                         calendar)]])
+        result = NetCDFTimeConverter().default_units(val, None)
+        self.assertEqual(result, (calendar, unit))
+
+    def test_nonequal_calendars(self):
+        # Test that different supplied calendars causes an error.
+        calendar_1 = '360_day'
+        calendar_2 = '365_day'
+        unit = 'days since 2000-01-01'
+        val = [CalendarDateTime(netcdftime.datetime(2014, 8, 12), calendar_1),
+               CalendarDateTime(netcdftime.datetime(2014, 8, 13), calendar_2)]
+        with self.assertRaisesRegexp(ValueError, 'not all equal'):
+            NetCDFTimeConverter().default_units(val, None)
+
 
 class Test_convert(unittest.TestCase):
     def test_numpy_array(self):
         val = np.array([7])
         result = NetCDFTimeConverter().convert(val, None, None)
         np.testing.assert_array_equal(result, val)
 
+    def test_numpy_nd_array(self):
+        shape = (4, 2)
+        val = np.arange(8).reshape(shape)
+        result = NetCDFTimeConverter().convert(val, None, None)
+        np.testing.assert_array_equal(result, val)
+        self.assertEqual(result.shape, shape)
+
     def test_numeric(self):
         val = 4
         result = NetCDFTimeConverter().convert(val, None, None)