def test_two_dim(self):
xa = np.array([1.5, 20.3])
ya = np.array([1.5, 2.0])
interp1 = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
interp2 = Bcinterp( np.arange(1000**2).reshape(1000,1000),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
assert_array_almost_equal(interp1.evaluate(xa, ya), interp2.evaluate(xa, ya))
def test_point_vs_known(self):
interp = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
# the below were generated from a ref implementation, but can also
# be evaluated readily by eye
i = interp.evaluate(1.01, 1.01)
assert_almost_equal(i, 10110.1, decimal=1)
i = interp.evaluate(1.5, 1.5)
assert_almost_equal(i, 15015.0)
i = interp.evaluate(2.3, 2.0)
assert_almost_equal(i, 20023.0)
i = interp.evaluate(20.3, 2.0)
assert_almost_equal(i, 20203.0)
def test_array_vs_known(self):
interp = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
xa = np.array([1.5, 20.3])
ya = np.array([1.5, 2.0])
i = interp.evaluate(xa, ya)
assert_almost_equal( i, np.array([15015.0, 20203.0]) )
def test_trivial(self):
interp = Bcinterp( np.ones(10*10), 1.0, 1.0, 10, 10, 0.0, 0.0 )
interp_vals = interp.evaluate( np.arange(1,9) + 0.1, np.ones(8) + 0.1 )
assert_allclose(np.ones(8), interp_vals)
class TestBcinterp():
def setup(self):
# initialize parameters
dim1 = 100
dim2 = 100
self.vals = np.abs(np.random.randn(dim1 * dim2))
self.x_space = 0.1
self.y_space = 0.1
self.Xdim = dim1
self.Ydim = dim2
self.x_corner = 0.0
self.y_corner = 0.0
self.interp = Bcinterp( self.vals,
self.x_space, self.y_space,
self.Xdim, self.Ydim,
self.x_corner, self.y_corner )
# generate a reference -- need better reference
self.new_x = np.arange(0.0, 8.0, .01) + 0.1
self.new_y = np.arange(0.0, 8.0, .01) + 0.1
def test_for_smoke(self):
ip = self.interp.evaluate(self.new_x, self.new_y)
if np.all( ip == 0.0 ):
print "Interpolator not working, likely cause: OMP failure."
print "Try reinstalling with no OMP: python setup.py install --no-openmp"
raise Exception()
def test_point_vs_known(self):
interp = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
# the below were generated from a ref implementation, but can also
# be evaluated readily by eye
i = interp.evaluate(1.01, 1.01)
assert_almost_equal(i, 10110.1, decimal=1)
i = interp.evaluate(1.5, 1.5)
assert_almost_equal(i, 15015.0)
i = interp.evaluate(2.3, 2.0)
assert_almost_equal(i, 20023.0)
i = interp.evaluate(20.3, 2.0)
assert_almost_equal(i, 20203.0)
def test_two_dim(self):
xa = np.array([1.5, 20.3])
ya = np.array([1.5, 2.0])
interp1 = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
interp2 = Bcinterp( np.arange(1000**2).reshape(1000,1000),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
assert_array_almost_equal(interp1.evaluate(xa, ya), interp2.evaluate(xa, ya))
def test_trivial(self):
interp = Bcinterp( np.ones(10*10), 1.0, 1.0, 10, 10, 0.0, 0.0 )
interp_vals = interp.evaluate( np.arange(1,9) + 0.1, np.ones(8) + 0.1 )
assert_allclose(np.ones(8), interp_vals)
def test_array_vs_known(self):
interp = Bcinterp( np.arange(1000**2),
0.1, 0.1, 1000, 1000, 0.0, 0.0 )
xa = np.array([1.5, 20.3])
ya = np.array([1.5, 2.0])
i = interp.evaluate(xa, ya)
assert_almost_equal( i, np.array([15015.0, 20203.0]) )
