def test_sum_different_quantity(self):
p1 = _cartesianparam.new()
p1.quantity = "DBZH"
p1.setData(reshape(array((NO_DATA, 111, 0, 111), uint8), (2, 2)))
p1.nodata, p1.undetect, p1.gain, p1.offset = NO_DATA, UNDETECT, 0.5, -32.5
d1 = _ravefield.new()
d1.addAttribute("what/gain", DEFAULT_DISTANCE_GAIN)
d1.setData(reshape(array((0, 100, 75, 50), uint8), (2, 2)))
d1.addAttribute("how/task", "se.smhi.composite.distance.radar")
p1.addQualityField(d1)
p2 = _cartesianparam.new()
p2.quantity = "TH"
p2.setData(reshape(array((NO_DATA, 111, 0, 111), uint8), (2, 2)))
p2.nodata, p2.undetect, p2.gain, p2.offset = NO_DATA, UNDETECT, 0.5, -32.5
d2 = _ravefield.new()
d2.addAttribute("what/gain", DEFAULT_DISTANCE_GAIN)
d2.setData(reshape(array((0, 100, 75, 50), uint8), (2, 2)))
d2.addAttribute("how/task", "se.smhi.composite.distance.radar")
p2.addQualityField(d2)
obj = _acrr.new()
obj.sum(p1, 200.0, 1.6)
try:
obj.sum(p2, 200.0, 1.6)
self.fail("Expected IOError")
except IOError:
pass
self.assertEqual(True, obj.isInitialized())
self.assertEqual("DBZH", obj.getQuantity())
def test_concatx(self):
obj = _ravefield.new()
obj.setData(numpy.zeros((10,10), numpy.uint8))
obj.setValue(0,1,10.0)
obj.setValue(5,4,20.0)
obj.setValue(9,4,35.0)
obj2 = _ravefield.new()
obj2.setData(numpy.zeros((10,6), numpy.uint8))
obj2.setValue(0,1,15.0)
obj2.setValue(5,4,25.0)
obj2.setValue(0,4,36.0)
obj2.setValue(1,4,37.0)
result = obj.concatx(obj2)
self.assertEqual(16, result.xsize)
self.assertEqual(10, result.ysize)
self.assertAlmostEqual(10.0, result.getValue(0,1)[1], 4)
self.assertAlmostEqual(20.0, result.getValue(5,4)[1], 4)
self.assertAlmostEqual(15.0, result.getValue(10,1)[1], 4)
self.assertAlmostEqual(25.0, result.getValue(15,4)[1], 4)
self.assertAlmostEqual(35.0, result.getValue(9,4)[1], 4)
self.assertAlmostEqual(36.0, result.getValue(10,4)[1], 4)
self.assertAlmostEqual(37.0, result.getValue(11,4)[1], 4)
def test_shiftData_neg(self):
obj = _polarscanparam.new()
f1 = _ravefield.new()
f2 = _ravefield.new()
obj.setData(
numpy.array(
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15]],
numpy.uint8))
f1.setData(
numpy.array(
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15]],
numpy.uint8))
f2.setData(
numpy.array(
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15]],
numpy.uint8))
obj.addQualityField(f1)
obj.addQualityField(f2)
obj.shiftData(-1)
self.assertTrue((numpy.array(
[[4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15], [0, 1, 2, 3]],
numpy.uint8) == obj.getData()).all())
self.assertTrue((numpy.array(
[[4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15], [0, 1, 2, 3]],
numpy.uint8) == obj.getQualityField(0).getData()).all())
self.assertTrue((numpy.array(
[[4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14, 15], [0, 1, 2, 3]],
numpy.uint8) == obj.getQualityField(1).getData()).all())
def test_accumulate_distfield_nodata(self):
distance_field1 = (10, 20, 30, ACRR_NO_DATA)
distance_field2 = (10, 20, 30, ACRR_NO_DATA)
p1 = _cartesianparam.new()
p1.quantity = "DBZH"
p1.setData(reshape(array((111, 111, UNDETECT, NO_DATA), uint8),
(2, 2)))
p1.nodata, p1.undetect, p1.gain, p1.offset = NO_DATA, UNDETECT, 0.5, -32.5
d1 = _ravefield.new()
d1.addAttribute("what/gain", DEFAULT_DISTANCE_GAIN)
d1.setData(reshape(array(distance_field1, uint8), (2, 2)))
d1.addAttribute("how/task", "se.smhi.composite.distance.radar")
p1.addQualityField(d1)
p2 = _cartesianparam.new()
p2.quantity = "DBZH"
p2.setData(reshape(array((111, 111, 111, NO_DATA), uint8), (2, 2)))
p2.nodata, p2.undetect, p2.gain, p2.offset = NO_DATA, UNDETECT, 0.5, -32.5
d2 = _ravefield.new()
d2.addAttribute("what/gain", DEFAULT_DISTANCE_GAIN)
d2.setData(reshape(array(distance_field2, uint8), (2, 2)))
d2.addAttribute("how/task", "se.smhi.composite.distance.radar")
p2.addQualityField(d2)
obj = _acrr.new()
obj.nodata = ACRR_NO_DATA
obj.undetect = UNDETECT
obj.sum(p1, 200.0, 1.6)
obj.sum(p2, 200.0, 1.6)
result = obj.accumulate(0.0, 2, 1.0)
self.assertAlmostEqual(1.0, result.getAttribute("what/prodpar"), 4)
self.assertEqual("ACRR", result.quantity)
self.assertAlmostEqual(ACRR_NO_DATA, result.nodata, 4)
self.assertAlmostEqual(UNDETECT, result.undetect, 4)
refAcrr = [1.0, 1.0, 0.5, ACRR_NO_DATA]
refDist = self.calculateAvgDistanceField(
[distance_field1, distance_field2], DEFAULT_DISTANCE_GAIN)
Acrr = result.getData().flatten()
qfield = result.getQualityFieldByHowTask(
"se.smhi.composite.distance.radar")
self.assertAlmostEqual(0.0, qfield.getAttribute("what/offset"), 4)
self.assertAlmostEqual(1000.0, qfield.getAttribute("what/gain"), 4)
Dist = qfield.getData().flatten()
for i in range(len(refAcrr)):
self.assertAlmostEqual(Acrr[i], refAcrr[i], 2)
self.assertAlmostEqual(Dist[i], refDist[i], 2)
def test_getConvertedQualityValueAtLonLat(self):
obj = _cartesian.new()
obj.projection = _rave.projection(
"gnom", "gnom",
"+proj=gnom +R=6371000.0 +lat_0=56.3675 +lon_0=12.8544")
obj.xscale = 100.0
obj.yscale = 100.0
xy = obj.projection.fwd(deg2rad((12.8544, 56.3675)))
obj.areaextent = (xy[0] - 4 * 100.0, xy[1] - 5 * 100.0,
xy[0] + 6 * 100.0, xy[1] + 5 * 100.0)
param = _cartesianparam.new()
param.quantity = "DBZH"
param.nodata = 255.0
param.undetect = 0.0
a = numpy.zeros((11, 9))
a = numpy.array(a.astype(numpy.float64), numpy.float64)
a = numpy.reshape(a, (11, 9)).astype(numpy.float64)
param.setData(a)
qf = numpy.arange(99)
qf = numpy.array(qf.astype(numpy.float64), numpy.float64)
qf = numpy.reshape(qf, (11, 9)).astype(numpy.float64)
qf2 = numpy.arange(99)
qf2 = numpy.array(qf.astype(numpy.float64), numpy.float64)
qf2 = numpy.reshape(qf, (11, 9)).astype(numpy.float64)
qf2[5][4] = 199.0
field1 = _ravefield.new()
field1.addAttribute("how/task", "se.task.1")
field1.addAttribute("what/offset", 10.0)
field1.addAttribute("what/gain", 2.0)
field1.setData(qf)
param.addQualityField(field1)
field2 = _ravefield.new()
field2.addAttribute("how/task", "se.task.2")
field2.addAttribute("what/gain", 3.0)
field2.setData(qf2)
param.addQualityField(field2)
obj.addParameter(param)
obj.defaultParameter = "DBZH"
#expected = obj.getConvertedValue((4,5))
result = obj.getConvertedQualityValueAtLonLat(
deg2rad((12.8544, 56.3675)), "se.task.1")
self.assertAlmostEqual(10.0 + 2.0 * 49.0, result, 4)
result = obj.getConvertedQualityValueAtLonLat(
deg2rad((12.8544, 56.3675)), "se.task.2")
self.assertAlmostEqual(3.0 * 199.0, result, 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 create_quality_field(self, xsize, ysize, dtype, value, howtask):
obj = _ravefield.new()
data = numpy.zeros((ysize, xsize), dtype)
data = data + value
obj.setData(data)
obj.addAttribute("how/task", howtask)
return obj
def lonlats(self, name=None, level=None):
self.grbs.seek(0)
try:
if name != None and level != None:
latlons = self.grbs.select(name=name, level=level)[0].latlons()
elif name != None and level == None:
latlons = self.grbs.select(name=name)[0].latlons()
else:
latlons = self.grbs.readline().latlons()
lats = _ravefield.new()
lats.setData(latlons[0])
lons = _ravefield.new()
lons.setData(latlons[1])
return (lons, lats)
except ValueError, e:
raise IOError(e)
def testPadding(self):
import numpy
scan = _polarscan.new()
dbzh = _polarscanparam.new()
dbzh.quantity = "DBZH"
l = []
for i in range(360):
l.append([i, i, i])
data = numpy.array(l).astype(numpy.uint16)
dbzh.setData(data)
scan.addParameter(dbzh)
dbzh.quantity = "TH"
scan.addParameter(dbzh)
options = ropo_realtime.options()
options.emitter2 = (-10, 3, 3) # default
newscan, gates = ropo_realtime.PadNrays(scan, options)
newdata = newscan.getParameter("DBZH").getData()
# Test data wrapping
self.assertEqual(newdata[:4].tolist(), data[356:, ].tolist())
self.assertEqual(newdata[364:, ].tolist(), data[:4, ].tolist())
classification = _ravefield.new()
classification.setData(newdata) # bogus probability of anomaly array
unwrapped, classification = ropo_realtime.UnpadNrays(
newscan, classification, gates)
# Test data unwrapping
self.assertEqual(
unwrapped.getParameter("DBZH").getData().tolist(), data.tolist())
def test_apply(self):
distance = _ravefield.new()
distance.setData(numpy.zeros((2, 2), numpy.float64))
distance.setValue(0, 0, 0.1)
distance.setValue(0, 1, 0.2)
distance.setValue(1, 0, 0.3)
distance.setValue(1, 1, 0.4)
distance.addAttribute("what/gain", 10000.0)
distance.addAttribute("what/offset", 0.0)
param = _cartesianparam.new()
param.setData(numpy.zeros((2, 2), numpy.float64))
param.setValue((0, 0), 1)
param.setValue((0, 1), 2)
param.setValue((1, 0), 3)
param.setValue((1, 1), 4)
param.quantity = "ACRR"
param.gain = 10.0
param.offset = 2.0
gra = _gra.new()
gra.A = 1.0
gra.B = 2.0
gra.C = 3.0
result = gra.apply(distance, param)
self.assertAlmostEqual(47.77, result.getConvertedValue((0, 0))[1], 2)
self.assertAlmostEqual(1102.61, result.getConvertedValue((0, 1))[1], 2)
self.assertAlmostEqual(3200.0, result.getConvertedValue((1, 0))[1], 4)
self.assertAlmostEqual(4200.0, result.getConvertedValue((1, 1))[1], 4)
def hacFilter(self, scan, quant="DBZH", enough=100):
NOD = odim_source.NODfromSource(scan)
# If HAC files are missing, then this method will passively fail.
try:
self.readHac(hacFile(scan, lastmonth=True))
if self.hac.getAttribute("how/count") < enough:
raise ValueError, "Not enough hits in climatology for %s" % NOD
hac_data = self.hac.getData()
if hac_data.shape != (scan.nrays, scan.nbins):
print hac_data.shape, (scan.nrays, scan.nbins)
raise IOError, "Scan and HAC have different geometries for %s" % NOD
## Get site-specific threshold!
try:
self.thresh = ARGS[NOD].thresh
except KeyError:
self.thresh = ARGS["default"].thresh
## Got site-specific threshold?
qind = _ravefield.new()
qind.setData(zeros(hac_data.shape, uint8))
qind.addAttribute("how/task", "eu.opera.odc.hac")
qind.addAttribute("how/task_args", self.thresh)
scan.addQualityField(qind)
_odc_hac.hacFilter(scan, self.hac, quant)
except Exception, e:
print traceback.format_exc()
def makeHac(self, fstr, nrays, nbins):
if not os.path.isfile(fstr):
self.hac = _ravefield.new()
self.hac.addAttribute("how/count", 0)
self.hac.setData(zeros((nrays, nbins), uint32))
else:
raise IOError, "HAC file already exists: %s" % fstr
def test_concatx_differentY(self):
obj = _ravefield.new()
obj.setData(numpy.zeros((10,10), numpy.uint8))
obj.setValue(0,1,10.0)
obj.setValue(5,4,20.0)
obj2 = _ravefield.new()
obj2.setData(numpy.zeros((9,6), numpy.uint8))
obj2.setValue(0,1,15.0)
obj2.setValue(5,4,25.0)
try:
obj.concatx(obj2)
self.fail("Expected ValueError")
except ValueError:
pass
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_qualityfields(self):
obj = _cartesianparam.new()
field1 = _ravefield.new()
field2 = _ravefield.new()
field1.addAttribute("what/name", "field1")
field2.addAttribute("what/name", "field2")
obj.addQualityField(field1)
obj.addQualityField(field2)
self.assertEqual(2, obj.getNumberOfQualityFields())
self.assertEqual("field1",
obj.getQualityField(0).getAttribute("what/name"))
obj.removeQualityField(0)
self.assertEqual(1, obj.getNumberOfQualityFields())
self.assertEqual("field2",
obj.getQualityField(0).getAttribute("what/name"))
def createQualityField(self, nrays, nbins, zv, howtask):
qfdata = numpy.zeros((nrays, nbins), numpy.int8)
qfdata = qfdata + zv
qf = _ravefield.new()
qf.setData(qfdata)
qf.addAttribute("how/task", howtask)
return qf
def test_bad_names(self):
obj = _ravefield.new()
BAD_NAMES = ["xyz/is", "what", "is"]
for n in BAD_NAMES:
try:
obj.addAttribute(n, 5)
self.fail("Expected AttributeError")
except AttributeError:
pass
def test_getQualityFieldByHowTask(self):
param = _cartesianparam.new()
param.quantity = "DBZH"
data = numpy.zeros((4, 5), numpy.int8)
param.setData(data)
field1 = _ravefield.new()
field2 = _ravefield.new()
field1.addAttribute("how/task", "se.smhi.f1")
field1.addAttribute("what/value", "f1")
field2.addAttribute("how/task", "se.smhi.f2")
field2.addAttribute("what/value", "f2")
param.addQualityField(field1)
param.addQualityField(field2)
result = param.getQualityFieldByHowTask("se.smhi.f2")
self.assertEqual("f2", result.getAttribute("what/value"))
def test_removeAttributes(self):
obj = _ravefield.new()
obj.addAttribute("what/is", 10.0)
obj.addAttribute("where/is", "that")
obj.addAttribute("how/are", 5)
obj.removeAttributes()
names = obj.getAttributeNames()
self.assertEqual(0, len(names))
def test_findQualityFieldByHowTask(self):
obj = _cartesian.new()
param = _cartesianparam.new()
field1 = _ravefield.new()
field2 = _ravefield.new()
field3 = _ravefield.new()
field1.addAttribute("how/task", "se.task.1")
field1.addAttribute("how/some", "should not be found")
field2.addAttribute("how/task", "se.task.2")
field3.addAttribute("how/notask", "abc")
obj.addQualityField(field1)
obj.addQualityField(field2)
obj.addQualityField(field3)
field4 = _ravefield.new()
field5 = _ravefield.new()
field4.addAttribute("how/task", "se.task.1")
field4.addAttribute("how/some", "should be found")
field5.addAttribute("how/task", "se.task.3")
param.addQualityField(field4)
param.addQualityField(field5)
param.quantity = "DBZH"
param.setData(numpy.zeros((240, 240), numpy.uint8))
obj.addParameter(param)
self.assertEqual(
"should be found",
obj.findQualityFieldByHowTask("se.task.1").getAttribute(
"how/some"))
self.assertEqual(
"se.task.2",
obj.findQualityFieldByHowTask("se.task.2").getAttribute(
"how/task"))
self.assertEqual(
"se.task.3",
obj.findQualityFieldByHowTask("se.task.3").getAttribute(
"how/task"))
self.assertEqual(None, obj.findQualityFieldByHowTask("abc"))
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_addField_no_quantity(self):
obj = _verticalprofile.new()
f = _ravefield.new()
f.setData(numpy.zeros((10, 1), numpy.uint8))
f.addAttribute("how/this", 1.0)
try:
obj.addField(f)
self.fail("Expected AttributeError")
except AttributeError:
pass
def test_setGetNZ(self):
obj = _verticalprofile.new()
self.assertTrue(None == obj.getNZ())
f = _ravefield.new()
f.setData(numpy.zeros((10, 1), numpy.uint8))
f.addAttribute("how/this", 1.0)
obj.setNZ(f)
result = obj.getNZ()
self.assertAlmostEqual(1.0, result.getAttribute("how/this"), 4)
self.assertEqual("nz", result.getAttribute("what/quantity"))
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_getConvertedValue_with_default_values(self): obj = _ravefield.new() obj.setData(numpy.zeros((10,10), numpy.uint8)) obj.setValue(0,1,1.0) obj.setValue(0,2,2.0) obj.setValue(0,3,3.0) obj.setValue(0,4,4.0) self.assertAlmostEqual(1.0, obj.getConvertedValue(0,1)[1], 4) self.assertAlmostEqual(2.0, obj.getConvertedValue(0,2)[1], 4) self.assertAlmostEqual(3.0, obj.getConvertedValue(0,3)[1], 4) self.assertAlmostEqual(4.0, obj.getConvertedValue(0,4)[1], 4)
def test_getQualityFieldByHowTask(self):
obj = _cartesian.new()
field1 = _ravefield.new()
field2 = _ravefield.new()
field3 = _ravefield.new()
field1.addAttribute("how/task", "se.task.1")
field2.addAttribute("how/task", "se.task.2")
field3.addAttribute("how/notask", "abc")
obj.addQualityField(field1)
obj.addQualityField(field2)
obj.addQualityField(field3)
self.assertEqual(
"se.task.1",
obj.getQualityFieldByHowTask("se.task.1").getAttribute("how/task"))
self.assertEqual(
"se.task.2",
obj.getQualityFieldByHowTask("se.task.2").getAttribute("how/task"))
self.assertEqual(None, obj.getQualityFieldByHowTask("abc"))
def test_addField_dev_bird(self):
obj = _verticalprofile.new()
f = _ravefield.new()
f.setData(numpy.zeros((10, 1), numpy.uint8))
f.addAttribute("how/this", 1.0)
f.addAttribute("what/quantity", "dev_bird")
obj.addField(f)
result = obj.getField("dev_bird")
self.assertAlmostEqual(1.0, result.getAttribute("how/this"), 4)
self.assertEqual("dev_bird", result.getAttribute("what/quantity"))
self.assertEqual(10, obj.getLevels())
