def test_gauss_fit():
x = np.arange(-1, 1, 0.01)
amplitude = 1
center = 0
sigma = 1
true_val = [amplitude, center, sigma]
y = amplitude / np.sqrt(2 * np.pi) / sigma * np.exp(-(x - center)**2 / 2 / sigma**2)
g = gaussian_model('',
1, 'fixed', [0, 1],
0.1, 'free', [0, 0.5],
0.5, 'free', [0, 1])
result, yfit = fit_engine(g, x, y)
out = result.values
fitted_val = (out['amplitude'], out['center'], out['sigma'])
assert_array_almost_equal(true_val, fitted_val)
def test_lorentzian2_fit():
x = np.arange(-1, 1, 0.01)
area = 1
center = 0
sigma = 1
true_val = [area, center, sigma]
y = (area/(1 + ((x - center) / sigma)**2)**2) / (np.pi * sigma)
m = lorentzian2_model('',
0.8, 'wrong', [0, 1],
0.1, 'free', [0, 0.5],
0.5, 'free', [0, 1])
result, yfit = fit_engine(m, x, y)
out = result.values
fitted_val = (out['area'], out['center'], out['sigma'])
assert_array_almost_equal(true_val, fitted_val)
def test_lorentzian_fit():
x = np.arange(-1, 1, 0.01)
amplitude = 1
center = 0
sigma = 1
true_val = [amplitude, center, sigma]
y = (amplitude/(1 + ((x - center) / sigma)**2)) / (np.pi * sigma)
m = lorentzian_model('',
0.8, 'free', [0, 1],
0.1, 'free', [0, 0.5],
0.8, 'bounded', [0, 2])
result, yfit = fit_engine(m, x, y)
out = result.values
fitted_val = (out['amplitude'], out['center'], out['sigma'])
assert_array_almost_equal(true_val, fitted_val)
def test_quadratic_fit():
x = np.arange(-1, 1, .01)
a = 1
b = 2
c = 3
true_val = [a, b, c]
y = a * x*x + b * x + c
m = quadratic_model('',
a, 'free', [-1, 1],
b, 'free', [-1, 1],
c, 'free', [-1, 1])
result, yfit = fit_engine(m, x, y)
out = result.values
fitted_val = (out['a'], out['b'], out['c'])
assert_array_almost_equal(true_val, fitted_val)
