def test_bounds(self):
from lenstronomy.LensModel.Profiles.spemd_smooth import SPEMD_SMOOTH
profile = SPEMD_SMOOTH()
compute_bool = profile._parameter_constraints(q_fastell=-1,
gam=-1,
s2=-1,
q=-1)
assert compute_bool is False
def test_bounds(self):
from lenstronomy.LensModel.Profiles.spemd_smooth import SPEMD_SMOOTH
profile = SPEMD_SMOOTH()
theta_E, gamma, q, phi_G, s_scale = profile._parameter_constraints(
theta_E=-1, s_scale=0, gamma=3, q=2, phi_G=0)
assert theta_E == 0
class TestSPEMD(object):
"""
tests the Gaussian methods
"""
def setup(self):
from lenstronomy.LensModel.Profiles.spemd_smooth import SPEMD_SMOOTH
self.SPEMD_SMOOT = SPEMD_SMOOTH()
from lenstronomy.LensModel.Profiles.nie import NIE
self.NIE = NIE()
def test_function(self):
phi_E = 1.
gamma = 2.
q = 0.999
phi_G = 1.
s_scale = 0.1
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
x = np.array([1., 2])
y = np.array([2, 0])
values = self.SPEMD_SMOOT.function(x, y, phi_E, gamma, e1, e2, s_scale)
if fastell4py_bool:
values_nie = self.NIE.function(x, y, phi_E, e1, e2, s_scale)
delta_f = values[0] - values[1]
delta_f_nie = values_nie[0] - values_nie[1]
npt.assert_almost_equal(delta_f, delta_f_nie, decimal=5)
else:
npt.assert_almost_equal(values, 0, decimal=5)
def test_derivatives(self):
x = np.array([1])
y = np.array([2])
phi_E = 1.
gamma = 2.
q = 0.7
phi_G = 1.
s_scale = 0.1
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
f_x, f_y = self.SPEMD_SMOOT.derivatives(x, y, phi_E, gamma, e1, e2,
s_scale)
if fastell4py_bool:
f_x_nie, f_y_nie = self.NIE.derivatives(x, y, phi_E, e1, e2,
s_scale)
#TODO test with higher precision, convention of theta_E might be slightly different from NIE with SPEMD
npt.assert_almost_equal(f_x, f_x_nie, decimal=2)
npt.assert_almost_equal(f_y, f_y_nie, decimal=2)
else:
npt.assert_almost_equal(f_x, 0, decimal=7)
npt.assert_almost_equal(f_y, 0, decimal=7)
def test_hessian(self):
x = np.array([1.])
y = np.array([2.])
phi_E = 1.
gamma = 2.
q = 0.9
phi_G = 1.
s_scale = 0.1
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
f_xx, f_yy, f_xy = self.SPEMD_SMOOT.hessian(x, y, phi_E, gamma, e1, e2,
s_scale)
if fastell4py_bool:
f_xx_nie, f_yy_nie, f_xy_nie = self.NIE.hessian(
x, y, phi_E, e1, e2, s_scale)
npt.assert_almost_equal(f_xx, f_xx_nie, decimal=3)
npt.assert_almost_equal(f_yy, f_yy_nie, decimal=3)
npt.assert_almost_equal(f_xy, f_xy_nie, decimal=3)
else:
npt.assert_almost_equal(f_xx, 0, decimal=7)
npt.assert_almost_equal(f_yy, 0, decimal=7)
npt.assert_almost_equal(f_xy, 0, decimal=7)
def test_bounds(self):
theta_E, gamma, q, phi_G, s_scale = self.SPEMD_SMOOT._parameter_constraints(
theta_E=-1, s_scale=0, gamma=3, q=2, phi_G=0)
assert theta_E == 0
def test_is_not_empty(self):
func = self.SPEMD_SMOOT.is_not_empty
assert func(0.1, 0.2)
assert func([0.1], [0.2])
assert func((0.1, 0.3), (0.2, 0.4))
assert func(np.array([0.1]), np.array([0.2]))
assert not func([], [])
assert not func(np.array([]), np.array([]))
class TestSPEMD(object):
"""
tests the Gaussian methods
"""
def setup(self):
from lenstronomy.LensModel.Profiles.spemd_smooth import SPEMD_SMOOTH
self.SPEMD_SMOOT = SPEMD_SMOOTH()
from lenstronomy.LensModel.Profiles.nie import NIE
self.NIE = NIE()
def test_function(self):
phi_E = 1.
gamma = 2.
q = 0.999
phi_G = 1.
s_scale = 0.1
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
x = np.array([1., 2])
y = np.array([2, 0])
values = self.SPEMD_SMOOT.function(x, y, phi_E, gamma, e1, e2, s_scale)
if fastell4py_bool:
values_nie = self.NIE.function(x, y, phi_E, e1, e2, s_scale)
delta_f = values[0] - values[1]
delta_f_nie = values_nie[0] - values_nie[1]
npt.assert_almost_equal(delta_f, delta_f_nie, decimal=5)
else:
npt.assert_almost_equal(values, 0, decimal=5)
def test_derivatives(self):
x = np.array([1])
y = np.array([2])
phi_E = 1.
gamma = 2.
q = 1.
phi_G = 1.
s_scale = 0.1
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
f_x, f_y = self.SPEMD_SMOOT.derivatives(x, y, phi_E, gamma, e1, e2,
s_scale)
if fastell4py_bool:
f_x_nie, f_y_nie = self.NIE.derivatives(x, y, phi_E, e1, e2,
s_scale)
npt.assert_almost_equal(f_x, f_x_nie, decimal=4)
npt.assert_almost_equal(f_y, f_y_nie, decimal=4)
else:
npt.assert_almost_equal(f_x, 0, decimal=7)
npt.assert_almost_equal(f_y, 0, decimal=7)
q = 0.7
phi_G = 1.
s_scale = 0.001
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
f_x, f_y = self.SPEMD_SMOOT.derivatives(x, y, phi_E, gamma, e1, e2,
s_scale)
if fastell4py_bool:
f_x_nie, f_y_nie = self.NIE.derivatives(x, y, phi_E, e1, e2,
s_scale)
npt.assert_almost_equal(f_x, f_x_nie, decimal=4)
npt.assert_almost_equal(f_y, f_y_nie, decimal=4)
else:
npt.assert_almost_equal(f_x, 0, decimal=7)
npt.assert_almost_equal(f_y, 0, decimal=7)
def test_hessian(self):
x = np.array([1.])
y = np.array([2.])
phi_E = 1.
gamma = 2.
q = 0.9
phi_G = 1.
s_scale = 0.001
e1, e2 = param_util.phi_q2_ellipticity(phi_G, q)
f_xx, f_yy, f_xy = self.SPEMD_SMOOT.hessian(x, y, phi_E, gamma, e1, e2,
s_scale)
if fastell4py_bool:
f_xx_nie, f_yy_nie, f_xy_nie = self.NIE.hessian(
x, y, phi_E, e1, e2, s_scale)
npt.assert_almost_equal(f_xx, f_xx_nie, decimal=4)
npt.assert_almost_equal(f_yy, f_yy_nie, decimal=4)
npt.assert_almost_equal(f_xy, f_xy_nie, decimal=4)
else:
npt.assert_almost_equal(f_xx, 0, decimal=7)
npt.assert_almost_equal(f_yy, 0, decimal=7)
npt.assert_almost_equal(f_xy, 0, decimal=7)
def test_bounds(self):
compute_bool = self.SPEMD_SMOOT._parameter_constraints(q_fastell=-1,
gam=-1,
s2=-1,
q=-1)
assert compute_bool is False
def test_is_not_empty(self):
func = self.SPEMD_SMOOT.is_not_empty
assert func(0.1, 0.2)
assert func([0.1], [0.2])
assert func((0.1, 0.3), (0.2, 0.4))
assert func(np.array([0.1]), np.array([0.2]))
assert not func([], [])
assert not func(np.array([]), np.array([]))