def test_undefined_attribute(self):
obj = _qitotal.new()
try:
obj.no_such_attribute = 0
self.fail("Expected AttributeError")
except AttributeError:
pass
def test_gain(self):
obj = _qitotal.new()
self.assertAlmostEqual(1.0, obj.gain, 4)
obj.gain = 2.0
self.assertAlmostEqual(2.0, obj.gain, 4)
try:
obj.gain = 0.0
self.fail("Expected AttributeError")
except AttributeError:
pass
self.assertAlmostEqual(2.0, obj.gain, 4)
def test_multiplicative_wGainOffsetAndDatatype(self):
obj = _qitotal.new()
obj.gain = 0.001
obj.offset = -0.01
obj.datatype = _rave.RaveDataType_UCHAR
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.setValue(0, 0, 0.1)
f2.setValue(0, 1, 0.2)
f2.setValue(1, 0, 0.5)
f2.setValue(1, 1, 1.0)
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) #0.2 + 1*0.1 = 0.3
f3.setValue(0, 1, 2) #0.2 + 2*0.1 = 0.4
f3.setValue(1, 0, 3) #0.2 + 3*0.1 = 0.5
f3.setValue(1, 1, 4) #0.2 + 4*0.1 = 0.6
result = obj.multiplicative([f1, f2, f3])
self.assertEqual(_rave.RaveDataType_UCHAR, result.datatype)
self.assertAlmostEqual(0.001, result.getAttribute("what/gain"))
self.assertAlmostEqual(-0.01, result.getAttribute("what/offset"))
# Due to the conversion errors, we can not check on more than 2 decimals
#0,0 = 0.1 * 0.1 * 0.3 = 0.003
self.assertAlmostEqual(0.003,
result.getValue(0, 0)[1] * 0.001 - 0.01, 2)
#0,1 = 0.2 * 0.2 * 0.4 = 0.016
self.assertAlmostEqual(0.016,
result.getValue(0, 1)[1] * 0.001 - 0.01, 2)
#1,0 = 0.3 * 0.5 * 0.5 = 0.075
self.assertAlmostEqual(0.075,
result.getValue(1, 0)[1] * 0.001 - 0.01, 2)
#1,1 = 0.4 * 1.0 * 0.6 = 0.24
self.assertAlmostEqual(0.24,
result.getValue(1, 1)[1] * 0.001 - 0.01, 2)
def test_minimum_with_weights(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.addAttribute("how/task", "se.smhi.f1")
f1.setValue(0, 0, 0.1) # 0.1 * 0.5 = 0.05
f1.setValue(0, 1, 0.2) # 0.2 * 0.5 = 0.10
f1.setValue(1, 0, 0.3) # 0.3 * 0.5 = 0.15
f1.setValue(1, 1, 0.4) # 0.4 * 0.5 = 0.2
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.addAttribute("how/task", "se.smhi.f2")
f2.setValue(0, 0, 0.1) # 0.1 * 0.75 = 0.075
f2.setValue(0, 1, 0.1) # 0.1 * 0.75 = 0.075
f2.setValue(1, 0, 0.5) # 0.5 * 0.75 = 0.375
f2.setValue(1, 1, 1.0) # 1.0 * 0.75 = 0.75
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("how/task", "se.smhi.f3")
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) # (0.2 + 1*0.1) * 0.9 = 0.27
f3.setValue(0, 1, 2) # (0.2 + 2*0.1) * 0.9 = 0.36
f3.setValue(1, 0, 3) # (0.2 + 3*0.1) * 0.9 = 0.45
f3.setValue(1, 1, 4) # (0.2 + 4*0.1) * 0.9 = 0.54
obj.setWeight("se.smhi.f1", 0.5)
obj.setWeight("se.smhi.f2", 0.25)
obj.setWeight("se.smhi.f3", 0.25)
result = obj.minimum([f1, f2, f3])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:minimum",
result.getAttribute("how/task_args"))
#0,0 = min(0.05, 0.075, 0.27) = 0.05 => 0.1
self.assertAlmostEqual(0.1, result.getValue(0, 0)[1], 4)
#0,1 = min(0.1, 0.075, 0.36) = 0.075 => 0.1
self.assertAlmostEqual(0.1, result.getValue(0, 1)[1], 4)
#1,0 = min(0.15, 0.375, 0.45) = 0.15 => 0.3
self.assertAlmostEqual(0.3, result.getValue(1, 0)[1], 4)
#1,1 = min(0.2, 0.75, 0.54) = 0.2 => 0.4
self.assertAlmostEqual(0.4, result.getValue(1, 1)[1], 4)
def test_additive_with_weights(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.addAttribute("how/task", "se.smhi.f1")
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.addAttribute("how/task", "se.smhi.f2")
f2.setValue(0, 0, 0.1)
f2.setValue(0, 1, 0.2)
f2.setValue(1, 0, 0.5)
f2.setValue(1, 1, 1.0)
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("how/task", "se.smhi.f3")
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) #0.2 + 1*0.1 = 0.3
f3.setValue(0, 1, 2) #0.2 + 2*0.1 = 0.4
f3.setValue(1, 0, 3) #0.2 + 3*0.1 = 0.5
f3.setValue(1, 1, 4) #0.2 + 4*0.1 = 0.6
obj.setWeight("se.smhi.f1", 0.5)
obj.setWeight("se.smhi.f2", 0.25)
obj.setWeight("se.smhi.f3", 0.25)
result = obj.additive([f1, f2, f3])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:additive",
result.getAttribute("how/task_args"))
#0,0 = (0.1*0.5 + 0.1*0.25 + 0.3*0.25) = 0.15
self.assertAlmostEqual(0.15, result.getValue(0, 0)[1], 4)
#0,1 = (0.2*0.5 + 0.2*0.25 + 0.4*0.25) = 0.25
self.assertAlmostEqual(0.25, result.getValue(0, 1)[1], 4)
#1,0 = (0.3*0.5 + 0.5*0.25 + 0.5*0.25) = 0.4
self.assertAlmostEqual(0.4, result.getValue(1, 0)[1], 4)
#1,1 = (0.4*0.5 + 1.0*0.25 + 0.6*0.25) = 0.6
self.assertAlmostEqual(0.6, result.getValue(1, 1)[1], 4)
def Xtest_multiplicative_weighted_fields(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.addAttribute("how/task", "se.smhi.f1")
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
f2 = _ravefield.new()
f1.addAttribute("how/task", "se.smhi.f2")
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.setValue(0, 0, 0.1)
f2.setValue(0, 1, 0.2)
f2.setValue(1, 0, 0.5)
f2.setValue(1, 1, 1.0)
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("how/task", "se.smhi.f3")
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) #0.2 + 1*0.1 = 0.3
f3.setValue(0, 1, 2) #0.2 + 2*0.1 = 0.4
f3.setValue(1, 0, 3) #0.2 + 3*0.1 = 0.5
f3.setValue(1, 1, 4) #0.2 + 4*0.1 = 0.6
obj.setWeight("se.smhi.f1", 10.0)
obj.setWeight("se.smhi.f2", 5.0)
result = obj.multiplicative([f1, f2, f3])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:multiplicative",
result.getAttribute("how/task_args"))
#0,0 = 0.1 * 0.1 * 0.3 = 0.003
self.assertAlmostEqual(0.003, result.getValue(0, 0)[1], 4)
#0,1 = 0.2 * 0.2 * 0.4 = 0.016
self.assertAlmostEqual(0.016, result.getValue(0, 1)[1], 4)
#1,0 = 0.3 * 0.5 * 0.5 = 0.075
self.assertAlmostEqual(0.075, result.getValue(1, 0)[1], 4)
#1,1 = 0.4 * 1.0 * 0.6 = 0.24
self.assertAlmostEqual(0.24, result.getValue(1, 1)[1], 4)
def processScan(self, objinfo, scan):
qitotal = _qitotal.new()
qitotal.datatype = QITOTAL_DTYPE
qitotal.gain = QITOTAL_GAIN
qitotal.offset = QITOTAL_OFFSET
qitotalfields = []
for f in objinfo.qifields():
qf = scan.findQualityFieldByHowTask(f.name())
if qf != None:
qitotal.setWeight(f.name(), f.weight())
qitotalfields.append(qf)
if len(qitotalfields) > 0:
if hasattr(qitotal, QITOTAL_METHOD):
method = getattr(qitotal, QITOTAL_METHOD)
result = method(qitotalfields)
scan.addOrReplaceQualityField(result)
def test_minimum_inconsistent_dimensions(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f3 = _ravefield.new()
f3.setData(numpy.zeros((3, 2), numpy.uint8))
try:
obj.minimum([f1, f2, f3])
self.fail("Expected AttributeError")
except AttributeError:
pass
def test_weight(self):
obj = _qitotal.new()
try:
obj.getWeight("se.some.task")
self.fail("Expected AttributeError")
except AttributeError:
pass
obj.setWeight("se.some.task", 2.0)
self.assertAlmostEqual(2.0, obj.getWeight("se.some.task"), 4)
obj.removeWeight("se.some.task")
try:
obj.getWeight("se.some.task")
self.fail("Expected AttributeError")
except AttributeError:
pass
def test_minimum_only_1_field(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
result = obj.minimum([f1])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:minimum",
result.getAttribute("how/task_args"))
self.assertAlmostEqual(0.1, result.getValue(0, 0)[1], 4)
self.assertAlmostEqual(0.2, result.getValue(0, 1)[1], 4)
self.assertAlmostEqual(0.3, result.getValue(1, 0)[1], 4)
self.assertAlmostEqual(0.4, result.getValue(1, 1)[1], 4)
def test_minimum(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.setValue(0, 0, 0.1)
f2.setValue(0, 1, 0.1)
f2.setValue(1, 0, 0.5)
f2.setValue(1, 1, 1.0)
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) #0.2 + 1*0.1 = 0.3
f3.setValue(0, 1, 2) #0.2 + 2*0.1 = 0.4
f3.setValue(1, 0, 3) #0.2 + 3*0.1 = 0.5
f3.setValue(1, 1, 4) #0.2 + 4*0.1 = 0.6
result = obj.minimum([f1, f2, f3])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:minimum",
result.getAttribute("how/task_args"))
#0,0 = min(0.1, 0.1, 0.3) = 0.1
self.assertAlmostEqual(0.1, result.getValue(0, 0)[1], 4)
#0,1 = min(0.2, 0.1, 0.4) = 0.1
self.assertAlmostEqual(0.1, result.getValue(0, 1)[1], 4)
#1,0 = min(0.3, 0.5, 0.5) = 0.3
self.assertAlmostEqual(0.3, result.getValue(1, 0)[1], 4)
#1,1 = min(0.4, 1.0, 0.6) = 0.4
self.assertAlmostEqual(0.4, result.getValue(1, 1)[1], 4)
def test_additive(self):
obj = _qitotal.new()
f1 = _ravefield.new()
f1.setData(numpy.zeros((2, 2), numpy.float64))
f1.setValue(0, 0, 0.1)
f1.setValue(0, 1, 0.2)
f1.setValue(1, 0, 0.3)
f1.setValue(1, 1, 0.4)
f2 = _ravefield.new()
f2.setData(numpy.zeros((2, 2), numpy.float64))
f2.setValue(0, 0, 0.1)
f2.setValue(0, 1, 0.2)
f2.setValue(1, 0, 0.5)
f2.setValue(1, 1, 1.0)
f3 = _ravefield.new()
f3.setData(numpy.zeros((2, 2), numpy.uint8))
f3.addAttribute("what/gain", 0.1)
f3.addAttribute("what/offset", 0.2)
f3.setValue(0, 0, 1) #0.2 + 1*0.1 = 0.3
f3.setValue(0, 1, 2) #0.2 + 2*0.1 = 0.4
f3.setValue(1, 0, 3) #0.2 + 3*0.1 = 0.5
f3.setValue(1, 1, 4) #0.2 + 4*0.1 = 0.6
result = obj.additive([f1, f2, f3])
self.assertEqual("pl.imgw.quality.qi_total",
result.getAttribute("how/task"))
self.assertEqual("method:additive",
result.getAttribute("how/task_args"))
#0,0 = (0.1 + 0.1 + 0.3)/3 = 0.1667
self.assertAlmostEqual(0.1667, result.getValue(0, 0)[1], 4)
#0,1 = (0.2 + 0.2 + 0.4)/3 = 0.2667
self.assertAlmostEqual(0.2667, result.getValue(0, 1)[1], 4)
#1,0 = (0.3 + 0.5 + 0.5)/3 = 0.4333
self.assertAlmostEqual(0.4333, result.getValue(1, 0)[1], 4)
#1,1 = (0.4 + 1.0 + 0.6)/3 = 0.6667
self.assertAlmostEqual(0.6667, result.getValue(1, 1)[1], 4)
def test_datatype(self):
obj = _qitotal.new()
self.assertEqual(_rave.RaveDataType_DOUBLE, obj.datatype)
obj.datatype = _rave.RaveDataType_UCHAR
self.assertEqual(_rave.RaveDataType_UCHAR, obj.datatype)
def test_offset(self):
obj = _qitotal.new()
self.assertAlmostEqual(0.0, obj.offset, 4)
obj.offset = 2.0
self.assertAlmostEqual(2.0, obj.offset, 4)
def test_attribute_visibility(self):
attrs = ['gain', 'offset', 'datatype']
obj = _qitotal.new()
alist = dir(obj)
for a in attrs:
self.assertEqual(True, a in alist)
def test_new(self):
obj = _qitotal.new()
self.assertNotEqual(-1, str(type(obj)).find("QITotalCore"))
