def test_redcal_xtalk(self):
antpos = np.array([[0., 0, 0], [1, 0, 0], [2, 0, 0], [3, 0, 0]])
d = {
(1, 2): np.array([[1.]], dtype=np.complex64),
(2, 3): np.array([[1. + 1j]], dtype=np.complex64)
}
x = {
(1, 2): np.array([[0.]], dtype=np.complex64),
(2, 3): np.array([[0. + 1j]], dtype=np.complex64)
}
reds = [[(1, 2), (2, 3)]]
info = Oc.RedundantInfo()
info.init_from_reds(reds, antpos)
m, g, v = Oc.redcal(d, info, xtalk=x, uselogcal=False)
self.assertEqual(g[1][0, 0], 1.)
self.assertEqual(g[2][0, 0], 1.)
self.assertEqual(g[3][0, 0], 1.)
#2D array testing
d = {
(1, 2):
np.array([[1., 2], [3., 4]], dtype=np.complex64),
(2, 3):
np.array([[1. + 1j, 2 + 2j], [3 + 3j, 4 + 4j]], dtype=np.complex64)
}
x = {
(1, 2):
np.array([[0., 2], [2., 3]], dtype=np.complex64),
(2, 3):
np.array([[0. + 1j, 1 + 2j], [2 + 3j, 3 + 4j]], dtype=np.complex64)
}
m, g, v = Oc.redcal(d, info, xtalk=x, uselogcal=False)
self.assertEqual(g[1][0, 0], 1.)
self.assertEqual(g[2][0, 0], 1.)
self.assertEqual(g[3][0, 0], 1.)
self.assertEqual(m['res'][(2, 3)][0][0], 0.)
def test_redcal(self):
#check that logcal give 0 chi2 for all 20 testinfos
for index in range(20):
arrayinfopath = os.path.dirname(
os.path.realpath(__file__)) + '/testinfo/test' + str(
index + 1) + '_array_info.txt'
c = Oc.RedundantCalibrator(56)
c.compute_redundantinfo(arrayinfopath, tol=.1)
info = Oc.RedundantInfo()
info.init_from_reds(c.Info.get_reds(), c.Info.get_antpos())
npz = np.load(testdata % index)
bls = [tuple(bl) for bl in npz['bls']]
dd = dict(zip(bls, npz['vis']))
m, g, v = Oc.redcal(dd,
info,
removedegen=True,
maxiter=50,
stepsize=.2,
conv=1e-5,
uselogcal=True)
calparpath = os.path.dirname(os.path.realpath(
__file__)) + '/testinfo/test' + str(index + 1) + '_calpar.txt'
with open(calparpath) as f:
rawinfo = [[float(x) for x in line.split()] for line in f]
temp = np.array(rawinfo[:-1])
correctcalpar = (np.array(temp[:, 0]) +
1.0j * np.array(temp[:, 1]))
i = list(g.keys())[0]
scalar = correctcalpar[i].real / g[i].real
for i in range(56):
if not i in g: continue
np.testing.assert_almost_equal(
np.abs(correctcalpar[i] - g[i] * scalar), 0, 4)
def setUp(self):
self.info = Oi.RedundantInfoLegacy(filename=redinfo_psa32,
txtmode=True)
self.info2 = Oc.RedundantInfo()
self.info2.init_from_reds([[
(0, 4), (1, 5), (2, 6), (3, 7), (4, 8), (5, 9)
], [(0, 3), (1, 4), (2, 5), (3, 6), (4, 7), (5, 8),
(6, 9)], [(0, 6), (1, 7), (2, 8),
(3, 9)], [(0, 5), (1, 6), (2, 7), (3, 8),
(4, 9)], [(0, 8),
(1, 9)], [(0, 7), (1, 8), (2, 9)],
[(0, 2), (1, 3), (2, 4), (3, 5), (4, 6),
(5, 7), (6, 8), (7, 9)],
[(0, 1), (1, 2), (2, 3), (3, 4), (4, 5),
(5, 6), (6, 7), (7, 8), (8, 9)]],
np.array([[0., 0., 1.], [0., 50., 1.],
[0., 100., 1.], [0., 150., 1.],
[0., 200., 1.], [0., 250., 1.],
[0., 300., 1.], [0., 350., 1.],
[0., 400., 1.], [0., 450., 1.]]))
self.freqs = np.linspace(.1, .2, 16)
self.times = np.arange(4)
self.reds = self.info2.get_reds()
self.true_vis = {}
for i, rg in enumerate(self.reds):
rd = np.array(
np.random.randn(self.times.size, self.freqs.size) +
1j * np.random.randn(self.times.size, self.freqs.size),
dtype=np.complex64)
self.true_vis[rg[0]] = rd
self.true_gains = {
i: np.ones((self.times.size, self.freqs.size), dtype=np.complex64)
for i in self.info2.subsetant
}
self.data = {}
self.bl2red = {}
for rg in self.reds:
for r in rg:
self.bl2red[r] = rg[0]
for redgp in self.reds:
for ai, aj in redgp:
self.data[ai, aj] = self.true_vis[self.bl2red[
ai, aj]] * self.true_gains[ai] * np.conj(
self.true_gains[aj])
self.unitgains = {
ant: np.ones((self.times.size, self.freqs.size),
dtype=np.complex64)
for ant in self.info2.subsetant
}
self.unitdata = {(ai, aj): np.ones((self.times.size, self.freqs.size),
dtype=np.complex64)
for ai, aj in self.info2.bl_order()}
def test_order_data(self):
antpos = np.array([[0., 0, 0], [1, 0, 0], [2, 0, 0], [3, 0, 0]])
reds = [[(0, 1), (1, 2), (2, 3)], [(0, 2), (1, 3)]]
i = Oc.RedundantInfo()
i.init_from_reds(reds, antpos)
dd = {
(0, 1): np.array([[0, 1j]]),
(1, 2): np.array([[0, 1j]]),
(2, 3): np.array([[0, 1j]]),
(2, 0): np.array([[0, 1j]]),
(1, 3): np.array([[0, 1j]]),
}
d = i.order_data(dd)
self.assertTrue(np.all(d[..., 0] == np.array([[0, 1j]])))
self.assertTrue(np.all(d[..., 1] == np.array([[0, 1j]])))
self.assertTrue(np.all(d[..., 2] == np.array([[0, 1j]])))
self.assertTrue(np.all(d[..., 3] == np.array([[0, 1j]]).conj()))
self.assertTrue(np.all(d[..., 4] == np.array([[0, 1j]])))
def test_full_functionality(self):
antpos = np.array([[14.60000038, -25.28794098, 1.],
[21.89999962, -12.64397049, 1.],
[14.60000038, 25.28794098, 1.],
[-21.89999962, -12.64397049, 1.],
[-14.60000038, 0., 1.],
[21.89999962, 12.64397049, 1.],
[29.20000076, 0., 1.],
[-14.60000038, -25.28794098, 1.],
[0., 25.28794098, 1.], [0., -25.28794098, 1.],
[0., 0., 1.], [-7.30000019, -12.64397049, 1.],
[-7.30000019, 12.64397049, 1.],
[-21.89999962, 12.64397049, 1.],
[-29.20000076, 0., 1.], [14.60000038, 0., 1.],
[-14.60000038, 25.28794098, 1.],
[7.30000019, -12.64397049, 1.]])
reds = [[(0, 8), (9, 16)],
[(13, 15), (14, 17), (3, 0), (4, 1), (16, 5), (12, 6)],
[(3, 17), (4, 15), (7, 0), (11, 1), (16, 2), (12, 5), (10, 6),
(14, 10)], [(3, 6), (14, 5)],
[(0, 9), (1, 17), (2, 8), (4, 14), (6, 15), (8, 16), (12, 13),
(11, 3), (10, 4), (9, 7), (15, 10), (17, 11)],
[(3, 8), (11, 2), (9, 5)],
[(3, 9), (4, 17), (12, 15), (11, 0), (10, 1), (8, 5), (13, 10),
(14, 11)], [(0, 13), (1, 16)],
[(0, 4), (1, 12), (6, 8), (9, 14), (15, 16), (17, 13)],
[(0, 5), (3, 16), (7, 12), (17, 2), (11, 8)],
[(0, 10), (7, 14), (10, 16), (11, 13), (6, 2), (9, 4), (15, 8),
(17, 12)],
[(1, 9), (2, 12), (5, 10), (6, 17), (8, 13), (12, 14), (10, 3),
(17, 7), (15, 11)], [(2, 3), (5, 7)],
[(16, 17), (12, 0), (8, 1), (13, 9)],
[(0, 17), (1, 15), (3, 14), (4, 13), (9, 11), (10, 12),
(12, 16), (5, 2), (7, 3), (11, 4), (6, 5), (17, 10)],
[(3, 15), (4, 5), (7, 1), (13, 2), (11, 6)],
[(5, 15), (8, 12), (10, 11), (13, 14), (15, 17), (1, 0),
(6, 1), (4, 3), (12, 4), (11, 7), (17, 9), (16, 13)],
[(0, 15), (1, 5), (3, 13), (4, 16), (9, 10), (11, 12), (15, 2),
(7, 4), (10, 8)],
[(0, 6), (3, 12), (4, 8), (7, 10), (9, 15), (14, 16), (10, 2),
(17, 5)], [(8, 17), (2, 1), (13, 7), (12, 9), (16, 11)],
[(0, 2), (7, 16), (9, 8)], [(4, 6), (14, 15), (3, 1), (13, 5)],
[(0, 14), (1, 13), (6, 16)], [(2, 14), (6, 7), (5, 3)],
[(2, 9), (8, 7)], [(2, 4), (5, 11), (6, 9), (8, 14), (15, 7)],
[(1, 14), (6, 13)], [(0, 8), (9, 16)],
[(13, 15), (14, 17), (3, 0), (4, 1), (16, 5), (12, 6)],
[(3, 17), (4, 15), (7, 0), (11, 1), (16, 2), (12, 5), (10, 6),
(14, 10)], [(3, 6), (14, 5)],
[(0, 9), (1, 17), (2, 8), (4, 14), (6, 15), (8, 16), (12, 13),
(11, 3), (10, 4), (9, 7), (15, 10), (17, 11)],
[(3, 8), (11, 2), (9, 5)],
[(3, 9), (4, 17), (12, 15), (11, 0), (10, 1), (8, 5), (13, 10),
(14, 11)], [(0, 13), (1, 16)],
[(0, 4), (1, 12), (6, 8), (9, 14), (15, 16), (17, 13)],
[(0, 5), (3, 16), (7, 12), (17, 2), (11, 8)],
[(0, 10), (7, 14), (10, 16), (11, 13), (6, 2), (9, 4), (15, 8),
(17, 12)],
[(1, 9), (2, 12), (5, 10), (6, 17), (8, 13), (12, 14), (10, 3),
(17, 7), (15, 11)], [(2, 3), (5, 7)],
[(16, 17), (12, 0), (8, 1), (13, 9)],
[(0, 17), (1, 15), (3, 14), (4, 13), (9, 11), (10, 12),
(12, 16), (5, 2), (7, 3), (11, 4), (6, 5), (17, 10)],
[(3, 15), (4, 5), (7, 1), (13, 2), (11, 6)],
[(5, 15), (8, 12), (10, 11), (13, 14), (15, 17), (1, 0),
(6, 1), (4, 3), (12, 4), (11, 7), (17, 9), (16, 13)],
[(0, 15), (1, 5), (3, 13), (4, 16), (9, 10), (11, 12), (15, 2),
(7, 4), (10, 8)],
[(0, 6), (3, 12), (4, 8), (7, 10), (9, 15), (14, 16), (10, 2),
(17, 5)], [(8, 17), (2, 1), (13, 7), (12, 9), (16, 11)],
[(0, 2), (7, 16), (9, 8)], [(4, 6), (14, 15), (3, 1), (13, 5)],
[(0, 14), (1, 13), (6, 16)], [(2, 14), (6, 7), (5, 3)],
[(2, 9), (8, 7)], [(2, 4), (5, 11), (6, 9), (8, 14), (15, 7)],
[(1, 14), (6, 13)], [(0, 8), (9, 16)],
[(13, 15), (14, 17), (3, 0), (4, 1), (16, 5), (12, 6)],
[(3, 17), (4, 15), (7, 0), (11, 1), (16, 2), (12, 5), (10, 6),
(14, 10)], [(3, 6), (14, 5)],
[(0, 9), (1, 17), (2, 8), (4, 14), (6, 15), (8, 16), (12, 13),
(11, 3), (10, 4), (9, 7), (15, 10), (17, 11)],
[(3, 8), (11, 2), (9, 5)],
[(3, 9), (4, 17), (12, 15), (11, 0), (10, 1), (8, 5), (13, 10),
(14, 11)], [(0, 13), (1, 16)],
[(0, 4), (1, 12), (6, 8), (9, 14), (15, 16), (17, 13)],
[(0, 5), (3, 16), (7, 12), (17, 2), (11, 8)],
[(0, 10), (7, 14), (10, 16), (11, 13), (6, 2), (9, 4), (15, 8),
(17, 12)],
[(1, 9), (2, 12), (5, 10), (6, 17), (8, 13), (12, 14), (10, 3),
(17, 7), (15, 11)], [(2, 3), (5, 7)],
[(16, 17), (12, 0), (8, 1), (13, 9)],
[(0, 17), (1, 15), (3, 14), (4, 13), (9, 11), (10, 12),
(12, 16), (5, 2), (7, 3), (11, 4), (6, 5), (17, 10)],
[(3, 15), (4, 5), (7, 1), (13, 2), (11, 6)],
[(5, 15), (8, 12), (10, 11), (13, 14), (15, 17), (1, 0),
(6, 1), (4, 3), (12, 4), (11, 7), (17, 9), (16, 13)],
[(0, 15), (1, 5), (3, 13), (4, 16), (9, 10), (11, 12), (15, 2),
(7, 4), (10, 8)],
[(0, 6), (3, 12), (4, 8), (7, 10), (9, 15), (14, 16), (10, 2),
(17, 5)], [(8, 17), (2, 1), (13, 7), (12, 9), (16, 11)],
[(0, 2), (7, 16), (9, 8)], [(4, 6), (14, 15), (3, 1), (13, 5)],
[(0, 14), (1, 13), (6, 16)], [(2, 14), (6, 7), (5, 3)],
[(2, 9), (8, 7)], [(2, 4), (5, 11), (6, 9), (8, 14), (15, 7)],
[(1, 14), (6, 13)], [(0, 8), (9, 16)],
[(13, 15), (14, 17), (3, 0), (4, 1), (16, 5), (12, 6)],
[(3, 17), (4, 15), (7, 0), (11, 1), (16, 2), (12, 5), (10, 6),
(14, 10)], [(3, 6), (14, 5)],
[(0, 9), (1, 17), (2, 8), (4, 14), (6, 15), (8, 16), (12, 13),
(11, 3), (10, 4), (9, 7), (15, 10), (17, 11)],
[(3, 8), (11, 2), (9, 5)],
[(3, 9), (4, 17), (12, 15), (11, 0), (10, 1), (8, 5), (13, 10),
(14, 11)], [(0, 13), (1, 16)],
[(0, 4), (1, 12), (6, 8), (9, 14), (15, 16), (17, 13)],
[(0, 5), (3, 16), (7, 12), (17, 2), (11, 8)],
[(0, 10), (7, 14), (10, 16), (11, 13), (6, 2), (9, 4), (15, 8),
(17, 12)],
[(1, 9), (2, 12), (5, 10), (6, 17), (8, 13), (12, 14), (10, 3),
(17, 7), (15, 11)], [(2, 3), (5, 7)],
[(16, 17), (12, 0), (8, 1), (13, 9)],
[(0, 17), (1, 15), (3, 14), (4, 13), (9, 11), (10, 12),
(12, 16), (5, 2), (7, 3), (11, 4), (6, 5), (17, 10)],
[(3, 15), (4, 5), (7, 1), (13, 2), (11, 6)],
[(5, 15), (8, 12), (10, 11), (13, 14), (15, 17), (1, 0),
(6, 1), (4, 3), (12, 4), (11, 7), (17, 9), (16, 13)],
[(0, 15), (1, 5), (3, 13), (4, 16), (9, 10), (11, 12), (15, 2),
(7, 4), (10, 8)],
[(0, 6), (3, 12), (4, 8), (7, 10), (9, 15), (14, 16), (10, 2),
(17, 5)], [(8, 17), (2, 1), (13, 7), (12, 9), (16, 11)],
[(0, 2), (7, 16), (9, 8)], [(4, 6), (14, 15), (3, 1), (13, 5)],
[(0, 14), (1, 13), (6, 16)], [(2, 14), (6, 7), (5, 3)],
[(2, 9), (8, 7)], [(2, 4), (5, 11), (6, 9), (8, 14), (15, 7)],
[(1, 14), (6, 13)]]
freqs = np.linspace(.1, .2, 64)
times = np.arange(11)
ants = np.arange(len(antpos))
info = Oc.RedundantInfo()
info.init_from_reds(reds, antpos)
# Simulate unique "true" visibilities
np.random.seed(21)
vis_true = {}
i = 0
for rg in reds:
vis_true[rg[0]] = np.array(
1.0 * np.random.randn(len(times), len(freqs)) +
1.0j * np.random.randn(len(times), len(freqs)),
dtype=np.complex64)
# Smulate true gains and then remove degeneracies from true gains so that removedegen will produce exact answers
gain_true = {}
for i in ants:
gain_true[i] = np.array(
1. + (.1 * np.random.randn(len(times), len(freqs)) +
.1j * np.random.randn(len(times), len(freqs))),
dtype=np.complex64)
g0 = {i: np.ones_like(gain_true[i]) for i in ants}
_, gain_true, _ = self.removedegen2(info, gain_true, vis_true, g0)
# Generate and apply firstcal gains
fcgains = {}
for i in ants:
fcspectrum = np.exp(2.0j * np.pi * 5.0 * np.random.randn() * freqs)
fcgains[i] = np.array([fcspectrum for t in times],
dtype=np.complex64)
for i in ants:
gain_true[i] *= fcgains[i]
# Generate fake data
bl2ublkey = {bl: rg[0] for rg in reds for bl in rg}
data = {}
for rg in reds:
for (i, j) in rg:
data[(i, j)] = np.array(np.conj(gain_true[i]) * gain_true[j] *
vis_true[rg[0]],
dtype=np.complex64)
# Run logcal, lincal, and removedegen
m1, g1, v1 = Oc.logcal(data, info, gains=fcgains)
m2, g2, v2 = Oc.lincal(data, info, gains=g1, vis=v1)
_g2 = {}
for k in g2:
_g2[k] = g2[k] / fcgains[k]
_, g3, v3 = Oc.removedegen(data, info, _g2, v2)
for k in g3:
g3[k] *= fcgains[k]
#Test that lincal actually converged
np.testing.assert_array_less(m2['iter'], 50 * np.ones_like(m2['iter']))
#Test that chisq is 0 after lincal in the gains/vis/data and in the meta
chiSqBeforeRemoveDegen = self.chisq(data, g2, v2, reds)
np.testing.assert_almost_equal(chiSqBeforeRemoveDegen,
np.zeros_like(chiSqBeforeRemoveDegen),
decimal=10)
np.testing.assert_almost_equal(m2['chisq'],
np.zeros_like(m2['chisq']),
decimal=10)
#Test that chisq is 0 after lincal and remove degen
chiSqAfterRemoveDegen = self.chisq(data, g3, v3, reds)
np.testing.assert_almost_equal(chiSqAfterRemoveDegen,
np.zeros_like(chiSqAfterRemoveDegen),
decimal=10)
#Test that the solution has degeneracies removed properly
Rgains = np.array([np.append(ai, 1) for ai in info.antloc])
Mgains = np.linalg.pinv(Rgains.T.dot(Rgains)).dot(Rgains.T)
ntimes, nfreqs = list(g3.values())[0].shape
for t in range(ntimes):
for f in range(nfreqs):
gainSols = np.array([
g3[ai][t, f] / fcgains[ai][t, f] for ai in info.subsetant
])
np.testing.assert_almost_equal(np.mean(np.abs(gainSols)),
1.0,
decimal=5)
np.testing.assert_almost_equal(Mgains.dot(np.angle(gainSols)),
[0.0, 0.0, 0.0, 0.0],
decimal=5)
#Test that the correct gains and visibilities are recovered
for ai in info.subsetant:
np.testing.assert_array_almost_equal(g3[ai],
gain_true[ai],
decimal=5)
for bl in vis_true.keys():
np.testing.assert_array_almost_equal(v3[bl],
vis_true[bl],
decimal=5)
def setUp(self):
self.legacy_info = Oi.RedundantInfoLegacy(filename=redinfo_psa32,
txtmode=True)
self.info = Oc.RedundantInfo()
self.info.init_from_reds(self.legacy_info.get_reds(),
self.legacy_info.antloc)