def test_dims_mismatch_error():
"""Test the process function with mismatched dimensions"""
rain, snow = setup_cubes(snow_data=np.array([[0, 0.5]], dtype=np.float32))
with pytest.raises(
ValueError, match="Rain and snow cubes are not on the same grid"
):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def process(*cubes: cli.inputcube, model_id_attr: str = None):
"""
Calculates a snow-fraction field from fields of snow and rain (rate or
accumulation). Where no precipitation is present, the data are filled in from
the nearest precipitating point.
snow_fraction = snow / (snow + rain)
Args:
cubes (iris.cube.CubeList or list):
Contains cubes of rain and snow, both must be either rates or accumulations.
model_id_attr (str):
Name of the attribute used to identify the source model for
blending.
Returns:
iris.cube.Cube:
A single cube containing the snow-fraction data.
"""
from iris.cube import CubeList
from improver.precipitation_type.snow_fraction import SnowFraction
return SnowFraction(model_id_attr=model_id_attr)(CubeList(cubes))
def test_non_coercing_units_error():
"""Test the process function with input cubes of incompatible units"""
rain, snow = setup_cubes()
rain.units = Unit("K")
with pytest.raises(
ValueError, match=r"Unable to convert from 'Unit\('m s-1'\)' to 'Unit\('K'\)'."
):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def test_coercing_units():
"""Test the process function with input cubes of different but compatible units"""
rain, snow = setup_cubes()
rain.convert_units("mm h-1")
result = SnowFraction()(iris.cube.CubeList([rain, snow]))
assert str(result.units) == "1"
np.testing.assert_allclose(result.data, EXPECTED_DATA)
assert rain.units == "mm h-1"
assert snow.units == "m s-1"
def test_wrong_input_names_error():
"""Test the process function with incorrect input cubes"""
rain, snow = setup_cubes()
rain.rename("kittens")
snow.rename("puppies")
with pytest.raises(
ValueError,
match=r"Could not find both rain and snow in \['kittens', 'puppies'\]",
):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def test_masked_data_error(mask_what):
"""Test the process function with masked data points"""
rain, snow = setup_cubes()
mask = [[False, False], [False, True]]
if "rain" in mask_what:
rain.data = np.ma.masked_array(rain.data, mask)
if "snow" in mask_what:
snow.data = np.ma.masked_array(snow.data, mask)
with pytest.raises(ValueError, match=r"Unexpected masked data in input cube\(s\)"):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def test_input_name_matches_both_phases_error():
"""Test the process function with an input cube that has both rain and snow in its name"""
rain, snow = setup_cubes()
rain.rename("its_raining_snowy_kittens")
snow.rename("puppies")
with pytest.raises(
ValueError,
match="Failed to find unique rain and snow cubes from \['its_raining_snowy_kittens', 'puppies'\]",
):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def test_acclen_mismatch_error():
"""Test the process function with mismatched accumulation lengths"""
rain, snow = setup_cubes(name="thickness_of_{phase}fall_amount")
time_coords = construct_scalar_time_coords(
[c.point for c in snow.coord("time").cells()],
(datetime(2017, 11, 10, 1, 0), datetime(2017, 11, 10, 4, 0)),
snow.coord("forecast_reference_time").cell(0).point,
)
_ = [snow.replace_coord(coord) for coord, _ in time_coords]
with pytest.raises(
ValueError, match="Rain and snow cubes do not have the same time coord"
):
SnowFraction()(iris.cube.CubeList([rain, snow]))
def test_basic(cube_name, mask_what, model_id_attr):
"""Run a test with four values, including one with no precip that will trigger
divide-by-zero.
Check data and metadata of result. Check with and without masked arrays (because
divide-by-zero gives a different result, even when input mask is all-False as in
this test) and with and without a model_id_attr value.
"""
rain, snow = setup_cubes(name=cube_name)
if "rain" in mask_what:
rain.data = np.ma.masked_array(rain.data)
if "snow" in mask_what:
snow.data = np.ma.masked_array(snow.data)
expected_attributes = COMMON_ATTRS.copy()
if model_id_attr:
expected_attributes[model_id_attr] = rain.attributes[model_id_attr]
result = SnowFraction(model_id_attr=model_id_attr)(iris.cube.CubeList([rain, snow]))
assert isinstance(result, iris.cube.Cube)
assert not isinstance(result.data, np.ma.masked_array)
assert str(result.units) == "1"
assert result.name() == "snow_fraction"
assert result.attributes == expected_attributes
np.testing.assert_allclose(result.data, EXPECTED_DATA)
def test_missing_cube_error():
"""Test the process function with one cube missing"""
rain, _ = setup_cubes()
with pytest.raises(ValueError, match="Expected exactly 2 input cubes, found 1"):
SnowFraction()(iris.cube.CubeList([rain]))