def test_check_scaling_ranges_int8_vector_unscaled(self):
default_array = DefaultData.create_default_vector(4, np.int8)
default_array[0] = 108
default_array[1] = 109
default_array[2] = -127
default_array[3] = 110
variable = Variable(["y"], default_array)
variable.encoding = dict([('_FillValue', -127)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_uint16_array_ok(self):
default_array = DefaultData.create_default_array(2, 2, np.float32)
default_array[0][0] = 9 # 9
default_array[0][1] = 205.605 # 65535
default_array[1][0] = np.NaN
default_array[1][1] = 14.06
variable = Variable(["y", "x"], default_array)
variable.encoding = dict([('dtype', np.uint16), ('_FillValue', 65535),
('scale_factor', 0.003), ('add_offset', 9)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_int8_vector_ok(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 11.872 # -128
default_array[1] = 12.127 # 127
default_array[2] = np.NaN
default_array[3] = 12.04
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int8), ('_FillValue', -127),
('scale_factor', 0.001), ('add_offset', 12)])
DataUtility.check_scaling_ranges(variable)
def test__get_min_max_missing_type(self):
default_array = DefaultData.create_default_vector(2, np.float32)
variable = Variable(["y"], default_array)
variable.encoding = dict([('_FillValue', -127),
('scale_factor', 0.023)])
try:
DataUtility._get_min_max(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test_check_scaling_ranges_int16_array_ok(self):
default_array = DefaultData.create_default_array(2, 2, np.float32)
default_array[0][0] = 75.534 # 32767
default_array[0][1] = -55.536 # -32768
default_array[1][0] = np.NaN
default_array[1][1] = 14.06
variable = Variable(["y", "x"], default_array)
variable.encoding = dict([('dtype', np.int16), ('_FillValue', -32767),
('scale_factor', 0.002), ('add_offset', 10)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_uint8_vector_ok(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 11.0 # 0
default_array[1] = 16.1 # 255
default_array[2] = np.NaN
default_array[3] = 13.05
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.uint8), ('_FillValue', 255),
('scale_factor', 0.02), ('add_offset', 11)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_int16_valid_min_max_ok(self):
default_array = DefaultData.create_default_array(2, 2, np.float32)
default_array[0][0] = 60 # 25000
default_array[0][1] = 10 # 0
default_array[1][0] = np.NaN
default_array[1][1] = 14.06
variable = Variable(["y", "x"], default_array)
variable.attrs["valid_max"] = 25000
variable.attrs["valid_min"] = 0
variable.encoding = dict([('dtype', np.int16), ('_FillValue', -32767),
('scale_factor', 0.002), ('add_offset', 10)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_uint32_only_NaN(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = np.NaN
default_array[1] = np.NaN
default_array[2] = np.NaN
default_array[3] = np.NaN
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.uint32),
('_FillValue', 4294967295),
('scale_factor', 0.00002),
('add_offset', 14)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_int32_vector_ok(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 21487.83647 # 2147483647
default_array[1] = -21461.83648 # -2147483648
default_array[2] = np.NaN
default_array[3] = 14.04
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int32),
('_FillValue', -2147483647),
('scale_factor', 0.00001),
('add_offset', 13)])
DataUtility.check_scaling_ranges(variable)
def test_check_scaling_ranges_uint8_vector_overflow(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 11.0 # 0
default_array[1] = 16.2 # overflow
default_array[2] = np.NaN
default_array[3] = 13.05
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.uint8), ('_FillValue', 255),
('scale_factor', 0.02), ('add_offset', 11)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test_check_scaling_ranges_uint16_array_overflow(self):
default_array = DefaultData.create_default_array(2, 2, np.float32)
default_array[0][0] = 9 # 0
default_array[0][1] = 205.705 # overflow
default_array[1][0] = np.NaN
default_array[1][1] = 14.06
variable = Variable(["y", "x"], default_array)
variable.encoding = dict([('dtype', np.uint16), ('_FillValue', 65535),
('scale_factor', 0.003), ('add_offset', 9)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test_check_scaling_ranges_int8_vector_underflow(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 11.702 # underflow
default_array[1] = 12.127 # 127
default_array[2] = np.NaN
default_array[3] = 12.04
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int8), ('_FillValue', -127),
('scale_factor', 0.001), ('add_offset', 12)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test__get_add_offset_missing(self):
default_array = DefaultData.create_default_vector(2, np.float32)
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int8), ('_FillValue', -127),
('scale_factor', 0.023)])
add_offset = DataUtility._get_add_offset(variable)
self.assertEqual(0.0, add_offset)
def test_check_scaling_ranges_int16_valid_min_max_underflow(self):
default_array = DefaultData.create_default_array(2, 2, np.float32)
default_array[0][0] = 60 # 25000
default_array[0][1] = 9 # underflow
default_array[1][0] = np.NaN
default_array[1][1] = 14.06
variable = Variable(["y", "x"], default_array)
variable.attrs["valid_max"] = 25000
variable.attrs["valid_min"] = 0
variable.encoding = dict([('dtype', np.int16), ('_FillValue', -32767),
('scale_factor', 0.002), ('add_offset', 10)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test_check_scaling_ranges_uint32_vector_overflow(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 14 # 0
default_array[1] = 85913.4459 # 4294967295
default_array[2] = np.NaN
default_array[3] = 14.01
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.uint32),
('_FillValue', 4294967295),
('scale_factor', 0.00002),
('add_offset', 14)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test_check_scaling_ranges_int32_vector_underflow(self):
default_array = DefaultData.create_default_vector(4, np.float32)
default_array[0] = 21487.83647 # 2147483647
default_array[1] = -21461.93648 # underflow
default_array[2] = np.NaN
default_array[3] = 14.04
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int32),
('_FillValue', -2147483647),
('scale_factor', 0.00001),
('add_offset', 13)])
try:
DataUtility.check_scaling_ranges(variable)
self.fail("ValueError expected")
except ValueError:
pass
def test__get_min_max(self):
default_array = DefaultData.create_default_vector(2, np.float32)
variable = Variable(["y"], default_array)
variable.encoding = dict([('dtype', np.int8), ('_FillValue', -127),
('scale_factor', 0.023)])
min_max = DataUtility._get_min_max(variable)
self.assertEqual(-128, min_max.min)
self.assertEqual(127, min_max.max)
