def test_slicing_and_indexing(self):
# Dummy data
data_a = numpy.array([
[[0, 1], [2, 3], [-1, 0]],
[[1, 0], [3, 2], [+3, 2]],
[[1, 1], [2, 2], [-3, 0]],
],
dtype=float)
x = Axis('x', 0., 1.)
y = Axis('y', 1., 2.)
z = Axis('z', 2., 3.)
cube_a = HyperspectralCube(data=data_a, x=x, y=y, z=z)
# SLICING USING [:,:,:]
cube_t = cube_a[0:1, :, :-1]
data_t = numpy.array([[[0.], [2.], [-1.]]], dtype=float)
self.assertIsInstance(cube_t, HyperspectralCube,
"Truncation result is a HyperspectralCube")
self.assertArrayEqual(data_t, cube_t.data,
"Supports extraction of sub-cube using [:,:,:]")
# SLICING UPDATES METADATA
self.assertEqual(cube_t.z.start, z.start, "Z axis start is unchanged")
self.assertEqual(cube_t.y.start, y.start, "Y axis start is unchanged")
self.assertEqual(cube_t.x.start, x.start, "X axis start is unchanged")
# SLICING [min:max] <=> [min:max,:,:]
cube_t = cube_a[1:2]
data_t = numpy.array([[[1, 0], [3, 2], [+3, 2]]], dtype=float)
self.assertArrayEqual(
data_t, cube_t.data,
"Unspecified axes are defaulted to `:` in slicing")
# SLICING UPDATES METADATA
self.assertEqual(cube_t.z.start, z.start + z.step,
"Z axis start has shifted")
# MUTATING ONE VALUE USING [λ,y,x] INDICES
data_t = numpy.array([
[[0, 1], [2, 3], [-1, 0]],
[[1, 0], [3, 9], [+3, 2]],
[[1, 1], [2, 2], [-3, 0]],
],
dtype=float)
cube_t = cube_a.copy()
cube_t[1, 1, 1] = 9
self.assertArrayEqual(data_t, cube_t.data,
"Supports mutation cube[λ,y,x] = n")
# MUTATING MULTIPLE VALUES USING [:,y,x] INDICES
data_t = numpy.array([
[[0, 1], [2, 9], [-1, 0]],
[[1, 0], [3, 9], [+3, 2]],
[[1, 1], [2, 9], [-3, 0]],
],
dtype=float)
cube_t = cube_a.copy()
cube_t[:, 1, 1] = 9
self.assertArrayEqual(data_t, cube_t.data,
"Supports mutation cube[:,y,x] = n")
def test_slicing_and_indexing(self):
# Dummy data
data_a = numpy.array([
[[0, 1], [2, 3], [-1, 0]],
[[1, 0], [3, 2], [+3, 2]],
[[1, 1], [2, 2], [-3, 0]],
], dtype=float)
x = Axis('x', 0., 1.)
y = Axis('y', 1., 2.)
z = Axis('z', 2., 3.)
cube_a = HyperspectralCube(data=data_a, x=x, y=y, z=z)
# SLICING USING [:,:,:]
cube_t = cube_a[0:1, :, :-1]
data_t = numpy.array([[[0.], [2.], [-1.]]], dtype=float)
self.assertIsInstance(cube_t, HyperspectralCube,
"Truncation result is a HyperspectralCube")
self.assertArrayEqual(data_t, cube_t.data,
"Supports extraction of sub-cube using [:,:,:]")
# SLICING UPDATES METADATA
self.assertEqual(cube_t.z.start, z.start, "Z axis start is unchanged")
self.assertEqual(cube_t.y.start, y.start, "Y axis start is unchanged")
self.assertEqual(cube_t.x.start, x.start, "X axis start is unchanged")
# SLICING [min:max] <=> [min:max,:,:]
cube_t = cube_a[1:2]
data_t = numpy.array([[[1, 0], [3, 2], [+3, 2]]], dtype=float)
self.assertArrayEqual(data_t, cube_t.data,
"Unspecified axes are defaulted to `:` in slicing")
# SLICING UPDATES METADATA
self.assertEqual(cube_t.z.start, z.start + z.step,
"Z axis start has shifted")
# MUTATING ONE VALUE USING [λ,y,x] INDICES
data_t = numpy.array([
[[0, 1], [2, 3], [-1, 0]],
[[1, 0], [3, 9], [+3, 2]],
[[1, 1], [2, 2], [-3, 0]],
], dtype=float)
cube_t = cube_a.copy()
cube_t[1, 1, 1] = 9
self.assertArrayEqual(data_t, cube_t.data,
"Supports mutation cube[λ,y,x] = n")
# MUTATING MULTIPLE VALUES USING [:,y,x] INDICES
data_t = numpy.array([
[[0, 1], [2, 9], [-1, 0]],
[[1, 0], [3, 9], [+3, 2]],
[[1, 1], [2, 9], [-3, 0]],
], dtype=float)
cube_t = cube_a.copy()
cube_t[:, 1, 1] = 9
self.assertArrayEqual(data_t, cube_t.data,
"Supports mutation cube[:,y,x] = n")
def test_wavelength_conversion_roundtrip(self, filepath):
cube = HyperspectralCube.from_fits(filepath)
# For each spectral pixel index
for pixel in range(0, cube.shape[0]):
self.assertEqual(
round(cube.pixel_of(cube.wavelength_of(pixel))), pixel,
"Wavelength conversion from pixel index #%d to λ "
"and back from λ to pixel index #%d." % (pixel, pixel))
def test_shape_property(self):
data_a = numpy.array([
[[0, 1], [2, 3], [-1, +0]],
[[1, 0], [3, 2], [+3, +2]],
[[1, 1], [2, 2], [-3, +0]],
],
dtype=float)
cube_a = HyperspectralCube(data=data_a)
self.assertArrayEqual((3, 3, 2), cube_a.shape,
"cube.shape should return the shape")
def test_wavelength_conversion_roundtrip(self, filepath):
cube = HyperspectralCube.from_fits(filepath)
# For each spectral pixel index
for pixel in range(0, cube.shape[0]):
self.assertEqual(
round(cube.pixel_of(cube.wavelength_of(pixel))),
pixel,
"Wavelength conversion from pixel index #%d to λ "
"and back from λ to pixel index #%d." % (pixel, pixel)
)
def test_wavelength_of(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# With a singe pixel position
self.assertEqual(u.Quantity(6734.7802734375, 'Angstrom'),
cube.wavelength_of(3))
# With a list of pixel positions
self.assertArrayEqual([6734.7802734375, 6737.2802734375],
cube.wavelength_of([3, 5]).value)
self.assertEqual(u.Unit('Angstrom'), cube.wavelength_of([3, 5]).unit)
def test_wavelength_of(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# With a singe pixel position
self.assertEqual(u.Quantity(6734.7802734375, 'Angstrom'),
cube.wavelength_of(3))
# With a list of pixel positions
self.assertArrayEqual([6734.7802734375, 6737.2802734375],
cube.wavelength_of([3, 5]).value)
self.assertEqual(u.Unit('Angstrom'), cube.wavelength_of([3, 5]).unit)
def test_pixel_of(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# With a single wavelength expressed as a simple float
self.assertEqual(3, cube.pixel_of(6734.7802734375))
# With a list of wavelengths
self.assertArrayEqual([3, 5],
cube.pixel_of([6734.7802734375,
6737.2802734375]),
"pixel_of() supports a list as input")
# With a ndarray of wavelengths
self.assertArrayEqual(
numpy.array([3, 5]),
cube.pixel_of(numpy.array([6734.7802734375, 6737.2802734375])),
"pixel_of() supports a numpy array as input")
def test_pixel_of(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# With a single wavelength expressed as a simple float
self.assertEqual(3, cube.pixel_of(6734.7802734375))
# With a list of wavelengths
self.assertArrayEqual(
[3, 5],
cube.pixel_of([6734.7802734375, 6737.2802734375]),
"pixel_of() supports a list as input"
)
# With a ndarray of wavelengths
self.assertArrayEqual(
numpy.array([3, 5]),
cube.pixel_of(numpy.array([6734.7802734375, 6737.2802734375])),
"pixel_of() supports a numpy array as input"
)
def test_write_to_fits_file(self):
cube = HyperspectralCube.from_fits(self.muse_01_filename)
fits_out_filename = os.path.join(self.fits_folder, 'tmp_output.fits')
self.assertFalse(os.path.isfile(fits_out_filename),
"Sanity check : Output FITS file should not exist ; "
"Please remove it by hand and then "
"re-run this test : %s" % fits_out_filename)
cube.to_fits(fits_out_filename)
self.assertTrue(os.path.isfile(fits_out_filename),
"Output FITS file should be created")
# `clobber` option should be false by default
with self.assertRaises(IOError):
cube.to_fits(fits_out_filename)
# clobber without raising
cube.to_fits(fits_out_filename, clobber=True)
os.remove(fits_out_filename) # cleanup
def test_getting_steps(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# Collect the data
steps = cube.get_steps()
# print "STEPS: %s" % steps
# <Quantity 1.25 Angstrom>
# <Quantity 5.55555555556e-05 deg>
# <Quantity -5.55555555556e-05 deg>
# Test the success/failure of the API
self.assertEqual(steps[0], cube.get_step(0))
self.assertEqual(steps[1], cube.get_step(1))
self.assertEqual(steps[2], cube.get_step(2))
# Test unit conversion -- this test is hardwired to muse_01.fits
self.assertEqual(cube.get_step(0).value, 1.25)
self.assertAlmostEqual(
cube.get_step(0).to(u.Unit('micron')).value, 1.25e-4)
def test_write_to_fits_file(self):
cube = HyperspectralCube.from_fits(self.muse_01_filename)
fits_out_filename = os.path.join(self.fits_folder, 'tmp_output.fits')
self.assertFalse(
os.path.isfile(fits_out_filename),
"Sanity check : Output FITS file should not exist ; "
"Please remove it by hand and then "
"re-run this test : %s" % fits_out_filename)
cube.to_fits(fits_out_filename)
self.assertTrue(os.path.isfile(fits_out_filename),
"Output FITS file should be created")
# `clobber` option should be false by default
with self.assertRaises(IOError):
cube.to_fits(fits_out_filename)
# clobber without raising
cube.to_fits(fits_out_filename, clobber=True)
os.remove(fits_out_filename) # cleanup
def test_getting_steps(self):
# This test is hardwired to muse_01.fits
cube = HyperspectralCube.from_fits(self.muse_01_filename)
# Collect the data
steps = cube.get_steps()
# print "STEPS: %s" % steps
# <Quantity 1.25 Angstrom>
# <Quantity 5.55555555556e-05 deg>
# <Quantity -5.55555555556e-05 deg>
# Test the success/failure of the API
self.assertEqual(steps[0], cube.get_step(0))
self.assertEqual(steps[1], cube.get_step(1))
self.assertEqual(steps[2], cube.get_step(2))
# Test unit conversion -- this test is hardwired to muse_01.fits
self.assertEqual(cube.get_step(0).value, 1.25)
self.assertAlmostEqual(
cube.get_step(0).to(u.Unit('micron')).value,
1.25e-4
)
def test_init_with_unfit_filename(self):
with self.assertRaises(IOError):
cube = HyperspectralCube.from_fits('this_is_not_a_fits_file')
def test_init_empty_cube(self):
cube = HyperspectralCube()
self.assertTrue(cube.is_empty())
def test_arithmetic_core(self):
inf = numpy.inf
nan = numpy.nan
# TEST DATA
data_a = numpy.array([
[[0, 1], [2, 3], [-1, +0]],
[[1, 0], [3, 2], [+3, +2]],
[[1, 1], [2, 2], [-3, +0]],
],
dtype=float)
data_b = numpy.array([
[[1, 1], [0, 0], [+1, -1]],
[[0, 0], [3, 1], [+3, -2]],
[[2, 2], [1, 2], [-1, +0]],
],
dtype=float)
data_i = numpy.array([
[[2, 0], [1, 1], [+1, -1]],
], dtype=float)
# BRAIN COMPUTED EXPECTATIONS
data_a_plus_1 = numpy.array([
[[1, 2], [3, 4], [+0, +1]],
[[2, 1], [4, 3], [+4, +3]],
[[2, 2], [3, 3], [-2, +1]],
],
dtype=float)
data_a_plus_b = numpy.array([
[[1, 2], [2, 3], [+0, -1]],
[[1, 0], [6, 3], [+6, -0]],
[[3, 3], [3, 4], [-4, +0]],
],
dtype=float)
data_a_plus_i = numpy.array([
[[2, 1], [3, 4], [+0, -1]],
[[3, 0], [4, 3], [+4, +1]],
[[3, 1], [3, 3], [-2, -1]],
],
dtype=float)
data_a_minus_1 = numpy.array([
[[-1, +0], [1, 2], [-2, -1]],
[[+0, -1], [2, 1], [+2, +1]],
[[+0, +0], [1, 1], [-4, -1]],
],
dtype=float)
data_1_minus_a = numpy.array([
[[+1, +0], [-1, -2], [+2, +1]],
[[+0, +1], [-2, -1], [-2, -1]],
[[+0, +0], [-1, -1], [+4, +1]],
],
dtype=float)
data_a_times_2 = numpy.array([
[[0, 2], [4, 6], [-2, +0]],
[[2, 0], [6, 4], [+6, +4]],
[[2, 2], [4, 4], [-6, +0]],
],
dtype=float)
data_a_times_b = numpy.array([
[[0, 1], [0, 0], [-1, +0]],
[[0, 0], [9, 2], [+9, -4]],
[[2, 2], [2, 4], [+3, +0]],
],
dtype=float)
data_a_div_2 = numpy.array([
[[.0, .5], [1.0, 3. / 2], [-.5, 0]],
[[.5, .0], [3. / 2, 1.0], [3. / 2, 1]],
[[.5, .5], [1.0, 1.0], [-3. / 2, 0]],
],
dtype=float)
data_a_div_b = numpy.array([
[[0, 1], [inf, inf], [-1, -0.]],
[[inf, nan], [1, 2], [+1, -1.]],
[[.5, .5], [2, 1], [+3, nan]],
],
dtype=float)
data_a_pow_2 = numpy.array([
[[0, 1], [4, 9], [+1, +0]],
[[1, 0], [9, 4], [+9, +4]],
[[1, 1], [4, 4], [+9, +0]],
],
dtype=float)
data_2_pow_a = numpy.array([
[[1, 2], [4, 8], [0.5, 1]],
[[2, 1], [8, 4], [8, 4]],
[[2, 2], [4, 4], [1. / 8, 1]],
],
dtype=float)
data_b_pow_a = numpy.array([
[[1, 1], [0, 0], [+1, +1]],
[[0, 1], [27, 1], [27, +4]],
[[2, 2], [1, 4], [-1, +1]],
],
dtype=float)
self.assertArrayEqual(data_a_div_b, data_a_div_b,
"Sanity check with inf and nan")
cube_a = HyperspectralCube(data=data_a)
cube_b = HyperspectralCube(data=data_b)
cube_e = HyperspectralCube() # empty
# CUBE + NUMBER
cube_a_plus_1 = cube_a + 1.
self.assertIsInstance(cube_a_plus_1, HyperspectralCube,
"Addition result is a HyperspectralCube")
self.assertArrayEqual(data_a_plus_1, cube_a_plus_1.data,
"Supports addition of cube and number using +")
# # NUMBER + CUBE
cube_a_plus_1 = 1. + cube_a
self.assertIsInstance(cube_a_plus_1, HyperspectralCube,
"Addition result is a HyperspectralCube")
self.assertArrayEqual(data_a_plus_1, cube_a_plus_1.data,
"Supports addition of number and cube using +")
# CUBE A + CUBE B
cube_a_plus_b = cube_a + cube_b
self.assertIsInstance(cube_a_plus_b, HyperspectralCube,
"Addition result is a HyperspectralCube")
self.assertArrayEqual(data_a_plus_b, cube_a_plus_b.data,
"Supports addition of two cubes using +")
# CUBE A + NDARRAY
cube_a_plus_b = cube_a + data_b
self.assertIsInstance(cube_a_plus_b, HyperspectralCube,
"Addition result is a HyperspectralCube")
self.assertArrayEqual(data_a_plus_b, cube_a_plus_b.data,
"Supports addition of cube and ndarray using +")
# CUBE + IMAGE
cube_a_plus_i = cube_a + data_i
self.assertIsInstance(cube_a_plus_i, HyperspectralCube,
"Addition result is a HyperspectralCube")
self.assertArrayEqual(data_a_plus_i, cube_a_plus_i.data,
"Supports addition of cube and image using +")
# NDARRAY + CUBE
# These are tricky : yields a numpy.ndarray because numpy is flexible
# in what it accepts as the right hand operator, and it accepts our
# HyperspectralCube because it provides the `__array__` method, so our
# __radd__ method is never even called.
cube_b_plus_a = data_b + cube_a
self.assertIsInstance(cube_b_plus_a, numpy.ndarray,
"Addition result is a numpy.ndarray")
self.assertArrayEqual(data_a_plus_b, cube_b_plus_a,
"Supports addition of image and cube using +")
# IMAGE + CUBE
cube_a_plus_i = data_i + cube_a
self.assertIsInstance(cube_a_plus_i, numpy.ndarray,
"Addition result is a numpy.ndarray")
self.assertArrayEqual(data_a_plus_i, cube_a_plus_i,
"Supports addition of image and cube using +")
# ADDITION EXPECTED ERRORS
with self.assertRaises(TypeError):
cube_a + cube_e # right-hand empty cube
with self.assertRaises(TypeError):
cube_e + cube_b # left-hand empty cube
with self.assertRaises(TypeError):
cube_b + 'rock' # left-hand "garbage"
with self.assertRaises(TypeError):
'666.' + cube_a # right-hand "garbage"
# CUBE - NUMBER
cube_a_minus_1 = cube_a - 1
self.assertIsInstance(cube_a_minus_1, HyperspectralCube,
"Subtraction result is a HyperspectralCube")
self.assertArrayEqual(
data_a_minus_1, cube_a_minus_1.data,
"Supports subtraction of number from cube using -")
# NUMBER - CUBE
cube_1_minus_a = 1 - cube_a
self.assertIsInstance(cube_1_minus_a, HyperspectralCube,
"Subtraction result is a HyperspectralCube")
self.assertArrayEqual(
data_1_minus_a, cube_1_minus_a.data,
"Supports subtraction of cube from number using -")
# CUBE * NUMBER
cube_a_times_2 = cube_a * 2
self.assertIsInstance(cube_a_times_2, HyperspectralCube,
"Multiplication result is a HyperspectralCube")
self.assertArrayEqual(
data_a_times_2, cube_a_times_2.data,
"Supports multiplication of cube and number using *")
# NUMBER * CUBE
cube_a_times_2 = 2 * cube_a
self.assertIsInstance(cube_a_times_2, HyperspectralCube,
"Multiplication result is a HyperspectralCube")
self.assertArrayEqual(
data_a_times_2, cube_a_times_2.data,
"Supports multiplication of number and cube using *")
# CUBE A * CUBE B
cube_a_times_b = cube_a * cube_b
self.assertIsInstance(cube_a_times_b, HyperspectralCube,
"Multiplication result is a HyperspectralCube")
self.assertArrayEqual(data_a_times_b, cube_a_times_b.data,
"Supports multiplication of two cubes using *")
# CUBE / NUMBER
cube_a_div_2 = cube_a / 2
self.assertIsInstance(cube_a_div_2, HyperspectralCube,
"Division result is a HyperspectralCube")
self.assertArrayEqual(data_a_div_2, cube_a_div_2.data,
"Supports division of cube and number using /")
# CUBE / CUBE
cube_a_div_b = cube_a / cube_b
self.assertIsInstance(cube_a_div_b, HyperspectralCube,
"Division result is a HyperspectralCube")
self.assertArrayEqual(data_a_div_b, cube_a_div_b.data,
"Supports division of two cubes using /")
# CUBE ** NUMBER
cube_a_pow_2 = cube_a**2
self.assertIsInstance(cube_a_pow_2, HyperspectralCube,
"Exponentiation result is a HyperspectralCube")
self.assertArrayEqual(
data_a_pow_2, cube_a_pow_2.data,
"Supports exponentiation of cube by number using **")
# NUMBER ** CUBE
cube_2_pow_a = 2**cube_a
self.assertIsInstance(cube_2_pow_a, HyperspectralCube,
"Exponentiation result is a HyperspectralCube")
self.assertArrayEqual(
data_2_pow_a, cube_2_pow_a.data,
"Supports exponentiation of number by cube using **")
def test_from_fits(self, filepath):
cube = HyperspectralCube.from_fits(filepath) # should not raise
print cube
def test_from_fits(self, filepath):
cube = HyperspectralCube.from_fits(filepath) # should not raise
print cube
def test_init_with_unfit_filename(self):
with self.assertRaises(IOError):
cube = HyperspectralCube.from_fits('this_is_not_a_fits_file')
def test_init_empty_cube(self):
cube = HyperspectralCube()
self.assertTrue(cube.is_empty())