Minor edit to include the influence of day weighting when using the i…

…ntensity categorisation.

improver/wxcode/modal_code.py

@@ -756,7 +756,7 @@ def process(self, cubes: CubeList) -> Cube:
         if len(cube.coord("time").points) == 1:
             result = cube
         else:
-            original_cube = cube.copy()
+            original_cube = self._emphasise_day_period(cube.copy())
             cube = self._consolidate_intensity_categories(cube)
             cube = self._emphasise_day_period(cube)
 

improver_tests/wxcode/wxcode/test_ModalFromGroupings.py

@@ -367,6 +367,13 @@ def test_expected_values_ignore_intensity(
         # least 1/(1+2) * 11 = 3.67 codes must be wet in order to produce a wet code.
         # As 5 codes are wet, a wet code is produced.
         ([1, 1, 1, 1, 1, 17, 17, 17], 2, 2, 4, 7, 8, False, 17),
+        # Mixed precipitation. Day weighting results in the heavy rain (15) codes
+        # doubling. This leads to 4 heavy rain codes vs 4 light rain codes with the
+        # heavy rain codes taking precedence.
+        ([10, 11, 12, 15, 15, 12, 12, 10, 3, 3, 3], 1, 2, 0, 5, 11, False, 15),
+        # Similar to the example above with the intensity categorisation included.
+        # Heavy rain (15) code is still the result.
+        ([10, 11, 12, 15, 15, 12, 12, 10, 3, 3, 3], 1, 2, 0, 5, 11, True, 15),
         # All precipitation is frozen. Ignoring the intensities means that a
         # day weighting of 2 results in 8 sleet codes and 8 light snow shower codes.
         # A wet bias of 2 means that at least 1/(1+2) * 16 = 5.33 codes must be wet