def test_scaling_equivalence_of_transformations(self):
number = 35
old_range = 0, 100
new_range = 0, 1
new_number = ForcedInvariance.scaling(None, number, old_range, new_range)
number /= 5
new_number /= 5
old = ForcedInvariance.scaling(None, new_number, new_range, old_range)
self.assertAlmostEqual(old, number)
class SmoothingMethod(unittest.TestCase):
def setUp(self):
params = Parameters()
self.fim = ForcedInvariance(params)
def test_smoothing_with_one_window(self):
curve = {0.96: 0.34, 0.51: 0.67, 0.52: 0.87, 0.11: 0.17, 0.18: 0.28}
check_curve = {0.11: 0.17, 0.18: 0.28, 0.51: 0.67, 0.52: 0.87, 0.96: 0.34}
self.fim.smoothing(curve)
self.assertEqual(curve, check_curve)
def unveil_image(self, params):
try:
check.layer(self.imagery_layer)
check.layer(self.index_layer)
check.sizes(self.imagery_layer, self.index_layer)
except (LayersError, SizesError) as e:
log.warning(str(e))
self.view.pushWarningMessage(str(e))
else:
image = Imagery(self.imagery_layer.source(),
index_path=self.index_layer.source())
result = self.view.saveFileDialog(extension=image.extension,
driver=image.driver)
if result:
image.unveiled(result)
log.info(f"Path to unveiled image: {result}")
fim = ForcedInvariance(params)
fim_task = FImTask(self.imagery_layer.name(),
self.index_layer.name())
fim_task.configure(fim, image)
fim_task.taskCompleted.connect(
partial(self.unveiling_finished, result))
fim_task.taskTerminated.connect(self.unveiling_error)
self.view.manager.task_manager.addTask(fim_task)
log.info("Task for FIM starts")
def test_algorithm(self):
try:
check.layer(self.imagery_layer)
check.layer(self.index_layer)
check.layer(self.standard_layer)
check.sizes(self.imagery_layer, self.index_layer, self.standard_layer)
check.test_threshold(self.threshold)
except (LayersError, SizesError, ParametersError) as e:
log.warning(str(e))
self.view.pushWarningMessage(str(e))
else:
imagery = Imagery(self.imagery_layer.source(),
index_path=self.index_layer.source())
unveiled = self.view.saveFileDialog(extension=imagery.extension,
driver=imagery.driver)
if unveiled:
imagery.unveiled(unveiled)
standard = Raster(self.standard_layer.source())
fim = ForcedInvariance(self.view.manager.parameters)
comparator = Comparator(self.threshold)
test_task = TestTask(self.imagery_layer.name(), self.standard_layer.name())
test_task.presenter(self)
test_task.configure(comparator, fim, imagery, standard, unveiled)
self.view.manager.task_manager.addTask(test_task)
log.info("Task for evaluating starts")
def test_correlation_method_basic(self):
stat = {0.93:[0, 0.4],
0.24:[0.39, 0.42, 0.35],
-0.35:[0.2]}
check_curve = {0.93:0.2, 0.24:0.39, -0.35:0.2}
curve = ForcedInvariance.correlation(None, stat)
for i in curve:
curve[i] = round(curve[i], 2)
self.assertEqual(curve, check_curve)
def test_scaling_array(self):
array = np.array([[35, 35],
[35, 35]])
old_range = 0, 100
new_range = 0, 1
check_array = np.array([[0.35, 0.35],
[0.35, 0.35]])
new = ForcedInvariance.scaling(None, array, old_range, new_range)
self.assertTrue(np.array_equal(new, check_array))
class RecalculateMethod(unittest.TestCase):
def setUp(self):
params = Parameters()
self.fim = ForcedInvariance(params)
def test_recalculate_without_mask_and_scaling(self):
band = np.array([[0.79667873, 0.05680854, 0.37823162, 0.76276089, 0.3857816],
[0.14231934, 0.41198452, 0.38253771, 0.18997385, 0.57106195],
[0.28108529, 0.92807336, 0.14973062, 0.78609907, 0.76275981]])
index = np.array([[0.19958716, 0.47928994, 0.06913423, 0.12842111, 0.01005531],
[0.62422291, 0.17398557, 0.01335282, 0.97057436, 0.57872553],
[0.96914121, 0.803315, 0.29172871, 0.97018698, 0.15332311]])
target = 0.4657257932012398
curve = {0.19958716: 0.45194359033603027, 0.47928994: 0.0209933994420185, 0.06913423: 0.46648663805927537, 0.12842111: 0.5039957264445407, 0.01005531: 0.6950645360202986, 0.62422291: 0.5289273409105809, 0.17398557: 0.8603914602720266, 0.01335282: 0.2889846828956172, 0.97057436: 0.1292500940460929, 0.57872553: 0.43467328172539943, 0.96914121: 0.5355264612673543, 0.803315: 0.6322687670285159, 0.29172871: 0.142014264054789, 0.97018698: 0.15467813578386858, 0.15332311: 0.8533351250273246}
check_band = np.array([[0.82097377, 1.26026285, 0.37761472, 0.70484213, 0.25849174],
[0.12531359, 0.22300526, 0.61649523, 0.68453121, 0.61185789],
[0.24444855, 0.68361387, 0.49103105, 2.36689311, 0.41629239]])
self.fim.recalculate(band, curve, index, target, None)
self.assertTrue(np.array_equal(np.around(band, 2), np.around(check_band, 2)))
def test_statistics_without_masks(self):
band = np.array([[0, 0.39, 0.42],
[0.2, 0.4, 0.35]])
index = np.array([[0.93, 0.24, 0.24],
[-0.35, 0.93, 0.24]])
check_stat = {0.93:[0, 0.4],
0.24:[0.39, 0.42, 0.35],
-0.35:[0.2]}
stat = ForcedInvariance.statistics(None, band, index)
self.assertEqual(stat, check_stat)
class TargetValueMethod(unittest.TestCase):
def setUp(self):
params = Parameters()
self.fim = ForcedInvariance(params)
def test_target_value_by_mean_without_coefficient(self):
band = np.array([[0.08366885, 0.33259951, 0.16066737, 0.18176515, 0.11411997],
[0.66369949, 0.09585974, 0.79209454, 0.83440211, 0.37460853],
[0.07946635, 0.5804281, 0.54472177, 0.98930503, 0.51080985]])
check_target = 0.422547756751967
target = self.fim.target_value(band)
self.assertAlmostEqual(target, check_target)
def setUp(self):
params = Parameters()
self.fim = ForcedInvariance(params)
def test_scaling_one_number(self):
number = 35
old_range = 0, 100
new_range = 0, 1
new = ForcedInvariance.scaling(None, number, old_range, new_range)
self.assertEqual(new, 0.35)