def test_vegetation(self):
# should not raise
niche_vlaanderen.Vegetation()
badveg = "tests/data/bad_ct/bad_vegetation.csv"
with pytest.warns(UserWarning):
with pytest.raises(CodeTableException):
niche_vlaanderen.Vegetation(ct_vegetation=badveg)
def test_borders(self):
soil_code = np.array([3, 3, 3, 3, 3])
mhw = np.array([21, 20, 10, 1, 0])
mlw = np.array([30, 30, 30, 30, 30])
v = niche_vlaanderen.Vegetation()
veg_predict, _ = v.calculate(soil_code, mhw, mlw, full_model=False)
expected = [0, 1, 1, 1, 0]
np.testing.assert_equal(expected, veg_predict[1])
def test_deviation_mlw(self):
v = niche_vlaanderen.Vegetation()
soil_code = np.array([3, 3, 3, 3, 3, -99, 2])
mhw = np.array([5, 5, 5, 5, 5, 5, 5])
mlw = np.array([66, 50, 38, 25, 5, 25, 25])
d = v.calculate_deviation(soil_code, mhw, mlw)
expected = np.array([28, 12, 0, 0, -15, np.nan, np.nan])
np.testing.assert_equal(expected, d["mlw_01"])
def test_deviation_mhw(self):
v = niche_vlaanderen.Vegetation()
soil_code = np.array([3, 3, 3, 3, -99, 2])
mhw = np.array([66, 16, 5, -5, 5, 5])
mlw = np.array([35, 35, 35, 35, 35, 35])
d = v.calculate_deviation(soil_code, mhw, mlw)
expected = np.array([46, 0, 0, -6, np.nan, np.nan])
np.testing.assert_equal(expected, d["mhw_01"])
def test_all_nodata(self):
soil_code = raster_to_numpy("tests/data/small/soil_code.asc")
mlw = raster_to_numpy("tests/data/small/mlw.asc")
mhw = mlw.copy()
mhw.fill(np.nan)
v = niche_vlaanderen.Vegetation()
with pytest.raises(NicheException):
v.calculate(soil_code, mhw, mlw)
def test_one_value(self):
nutrient_level = np.array([5])
acidity = np.array([3])
mlw = np.array([50])
mhw = np.array([10])
soil_code = np.array([140000])
v = niche_vlaanderen.Vegetation()
veg_predict, veg_occurrence = v.calculate(soil_code, mhw, mlw,
nutrient_level, acidity)
correct = [] # no types should match
for vi in veg_predict:
if vi in correct:
np.testing.assert_equal(np.array([1]), veg_predict[vi])
else:
np.testing.assert_equal(np.array([0]), veg_predict[vi])
def test_one_value_simple(self):
mlw = np.array([50])
mhw = np.array([10])
soil_code = np.array([8])
v = niche_vlaanderen.Vegetation()
veg_predict, veg_occurrence = v.calculate(soil_code,
mhw,
mlw,
full_model=False)
correct = [3, 8, 11, 18, 23, 27]
for vi in veg_predict:
if vi in correct:
np.testing.assert_equal(np.array([1]), veg_predict[vi])
else:
np.testing.assert_equal(np.array([0]), veg_predict[vi])
def test_simple_doc_inundation(self):
nutrient_level = np.array([4])
acidity = np.array([3])
mlw = np.array([50])
mhw = np.array([10])
soil_code = np.array([14])
inundation = np.array([1])
v = niche_vlaanderen.Vegetation()
veg_predict, veg_occurrence = \
v.calculate(soil_code, mhw, mlw, nutrient_level, acidity,
inundation=inundation)
correct = [7, 12, 16]
for vi in veg_predict:
if vi in correct:
np.testing.assert_equal(np.array([1]), veg_predict[vi])
else:
np.testing.assert_equal(np.array([0]), veg_predict[vi])
def test_occurrence(self):
nutrient_level = np.array([[4, 4], [4, 5]])
acidity = np.array([[3, 3], [3, 255]])
mlw = np.array([[50, 50], [50, 50]])
mhw = np.array([[31, 30], [10, 4]])
soil_code = np.array([[14, 14], [14, 14]])
inundation = np.array([[1, 1], [1, 1]])
v = niche_vlaanderen.Vegetation()
veg_predict, veg_occurrence = \
v.calculate(soil_code=soil_code, mhw=mhw, mlw=mlw,
nutrient_level=nutrient_level, acidity=acidity,
inundation=inundation)
# check no data propagates nicely
self.assertEqual(255, veg_predict[1][1, 1])
self.assertEqual(1 / 3, veg_occurrence[12])
self.assertEqual(1, veg_occurrence[7])
self.assertEqual(2 / 3, veg_occurrence[16])
def test_testcase(self):
input_dir = "testcase/zwarte_beek/input/"
soil_code = raster_to_numpy(input_dir + "soil_code.asc")
soil_code_r = soil_code
soil_code_r[soil_code > 0] = np.round(soil_code / 10000)[soil_code > 0]
msw = raster_to_numpy(input_dir + "msw.asc")
mhw = raster_to_numpy(input_dir + "mhw.asc")
mlw = raster_to_numpy(input_dir + "mlw.asc")
inundation = \
raster_to_numpy(input_dir + "inundation.asc")
regenlens = raster_to_numpy(input_dir + "nullgrid.asc")
seepage = raster_to_numpy(input_dir + "seepage.asc")
conductivity = raster_to_numpy(input_dir + "minerality.asc")
nitrogen_deposition = \
raster_to_numpy(input_dir + "nitrogen_atmospheric.asc")
nitrogen_animal = raster_to_numpy(input_dir + "nullgrid.asc")
nitrogen_fertilizer = raster_to_numpy(input_dir + "nullgrid.asc")
management = raster_to_numpy(input_dir + "management.asc")
nl = niche_vlaanderen.NutrientLevel()
nutrient_level = nl.calculate(soil_code_r, msw, nitrogen_deposition,
nitrogen_animal, nitrogen_fertilizer,
management, inundation)
a = niche_vlaanderen.Acidity()
acidity = a.calculate(soil_code_r, mlw, inundation, seepage,
conductivity, regenlens)
v = niche_vlaanderen.Vegetation()
veg_predict, veg_occurrence = v.calculate(soil_code_r, mhw, mlw,
nutrient_level, acidity)
for i in range(1, 28):
vi = raster_to_numpy("testcase/zwarte_beek/vegetation/v%d.asc" % i)
# TODO: this is dirty - we apply the same no data filter to the
# original set the new set, as this was done incorrectly in the
# original set.
# this also means that if we predict no data everywhere the test
# also works :-)
vi[(veg_predict[i] == 255)] = 255
np.testing.assert_equal(vi, veg_predict[i])
def test_unique_mlw(self):
badveg = "tests/data/bad_ct/differentmlw.csv"
with pytest.raises(CodeTableException):
niche_vlaanderen.Vegetation(ct_vegetation=badveg)