def test_marginal_calcs_2d():
"""Tests that marginal_calcs_2d returns the correct values as an array"""
nmean = 1.0
ps = twinbeam_pmf({"n_modes": 1.0, "sq_0": nmean})
res = marginal_calcs_2d(ps, as_dict=False)
expected = np.array([nmean, nmean, 2 + 1 / nmean, 2, 2, 0])
assert np.allclose(res, expected)
def test_twin_correct_stats(sq_0, sq_1, noise_s, noise_i, eta_s, eta_i):
"""Test that dark counts are correctly included by calculating expected g2s and mean photon numbers"""
cutoff = 40
params = {
"n_modes": 2,
"eta_s": eta_s,
"eta_i": eta_i,
"noise_s": noise_s,
"noise_i": noise_i,
"sq_0": sq_0,
"sq_1": sq_1,
}
pmf = twinbeam_pmf(params, cutoff=cutoff)
K = (sq_0 + sq_1)**2 / (sq_0**2 + sq_1**2)
M = sq_0 + sq_1
g2noiseless = 1.0 + 1.0 / K
eps_s = noise_s / (eta_s * M)
eps_i = noise_i / (eta_i * M)
g2s = (g2noiseless + 2 * eps_s + eps_s**2) / (1.0 + 2 * eps_s + eps_s**2)
g2i = (g2noiseless + 2 * eps_i + eps_i**2) / (1.0 + 2 * eps_i + eps_i**2)
marginals = marginal_calcs_2d(pmf)
np.allclose(g2s, marginals["g2_s"])
np.allclose(g2i, marginals["g2_i"])
np.allclose(noise_s + eta_s * M, marginals["n_s"])
np.allclose(noise_i + eta_i * M, marginals["n_i"])
def test_gen_hist_2d_poisson(ns, ni):
"""Test the histograms are correctly generated for pure noise"""
nsamples = 1000000
mat = gen_hist_2d(np.random.poisson(ns, nsamples),
np.random.poisson(ni, nsamples))
n, m = mat.shape
nmax = np.max([n, m])
expected = twinbeam_pmf({"noise_s": ns, "noise_i": ni})[:n, :m]
np.allclose(expected, mat, atol=0.01)
def test_exact_model_2d(n_modes, do_not_vary, threshold):
"""Test that the fitting is correct when the guess is exactly the correct answer"""
sq_n = 0.7 * (0.5**np.arange(n_modes))
noise_s = 0.1
noise_i = 0.15
eta_s = 0.7
eta_i = 0.6
params = {"sq_" + str(i): sq_n[i] for i in range(n_modes)}
params["n_modes"] = n_modes
params["eta_s"] = eta_s
params["eta_i"] = eta_i
params["noise_s"] = noise_s
params["noise_i"] = noise_i
if threshold:
probs = threshold_2d(twinbeam_pmf(params), threshold, threshold)
else:
probs = twinbeam_pmf(params)
fit = fit_2d(probs, params, do_not_vary=do_not_vary, threshold=threshold)
assert np.allclose(fit.chisqr, 0.0)
def test_gen_hist_2d_twin(sq_0):
"""Check that a histogram is constructed correctly for a lossless pure twin-beam source"""
nsamples = 1000000
p = 1 / (1.0 + sq_0)
samples = np.random.geometric(p, nsamples) - 1
mat = gen_hist_2d(samples, samples)
n, m = mat.shape
assert n == m
expected = twinbeam_pmf({"sq_0": sq_0, "n_modes": 1}, cutoff=n)
assert np.allclose(mat, expected, atol=0.01)
def test_two_schmidt_mode_guess_exact(eta_s, eta_i, sq_0, sq_1):
"""Test that one can invert correctly when there are two Schmidt modes
and no dark counts.
"""
pmf = twinbeam_pmf({
"sq_0": sq_0,
"sq_1": sq_1,
"n_modes": 2,
"eta_s": eta_s,
"eta_i": eta_i
})
guess = two_schmidt_mode_guess(pmf)
assert np.allclose(eta_s, guess["eta_s"], atol=1.0e-2)
assert np.allclose(eta_i, guess["eta_i"], atol=1.0e-2)
sq_ns = [sq_0,
sq_1] # We need to sort sq_n1 and sq_n2 so that sq_n1 >= sq_n2
sq_0 = np.max(sq_ns)
sq_1 = np.min(sq_ns)
assert np.allclose(sq_0, guess["sq_0"], atol=1.0e-2)
assert np.allclose(sq_1, guess["sq_1"], atol=1.0e-2)
def model_2d(params, jpd_data):
(dim_s, dim_i) = pd_data.shape
joint_pmf = twinbeam_pmf(params, cutoff=cutoff)[:dim_s, :dim_i]
return joint_pmf - jpd_data
def model_2d(params, jpd_data):
(dim_s, dim_i) = pd_data.shape
joint_pmf = twinbeam_pmf(params, cutoff=cutoff)
return threshold_2d(joint_pmf, dim_s, dim_i) - threshold_2d(
jpd_data, dim_s, dim_i)
