Fix saving lists of arrays with recent versions of numpy
[qpms.git] / oldtests / test_qpms_p.py
blob427594e6da1fbad63b1aba1afb3d0f397758bb70
1 """
2 Unit tests for qpms_p
3 =====================
5 Covered functions
6 -----------------
7 plane_pq_y vs. vswf_yr1
9 Not covered
10 -----------
11 Everything else
13 """
15 import unittest
16 import qpms
17 import numpy as np
18 from numpy import newaxis as ň
19 import warnings
21 # Some constants go here.
22 lengthOrdersOfMagnitude = [2.**i for i in range(-15,10,2)]
24 class PlaneWaveDecompositionTests(unittest.TestCase):
25 """
26 covers plane_pq_y and vswf_yr1
27 """
28 def testRandomInc(self):
29 # The "maximum" argument of the Bessel's functions, i.e. maximum wave number times the distance,
30 # for the "locally strongly varying fields"
31 maxx = 10
32 rfailtol = 0.01 # how much of the randomized test will be tolerated
33 lMax = 80 # To which order we decompose the waves
34 rtol = 1e-5 # relative required precision
35 atol = 1. # absolute tolerance, does not really play a role
36 nsamples = 4 # (frequency, direction, polarisation) samples per order of magnitude and test
37 npoints = 15 # points to evaluate per sample
40 failcounter = 0
41 passcounter = 0
42 for oom in lengthOrdersOfMagnitude:
43 k = np.random.randn(nsamples, 3) / oom
44 ksiz = np.linalg.norm(k, axis=-1)
45 kdir = k / ksiz[...,ň]
46 E_0 = np.cross(np.random.randn(nsamples, 3), k) * oom # ensure orthogonality
47 for s in range(nsamples):
48 testpoints = oom * maxx * np.random.randn(npoints, 3)
49 p, q = qpms.plane_pq_y(lMax, k[s], E_0[s])
50 planewave_1 = np.exp(1j*np.dot(testpoints,k[s]))[:,ň] * E_0[s,:]
51 for i in range(npoints):
52 sph = qpms.cart2sph(ksiz[s]*testpoints[i])
53 M̃_y, Ñ_y = qpms.vswf_yr1(sph, lMax, 1)
54 planewave_2_p = -1j*qpms.sph_loccart2cart(np.dot(p,Ñ_y)+np.dot(q,M̃_y),sph)
55 #self.assertTrue(np.allclose(planewave_2_p, planewave_1[i], rtol=rtol, atol=atol))
56 if not np.allclose(planewave_2_p, planewave_1[i], rtol=rtol, atol=atol):
57 False and warnings.warn('Planewave expansion test not passed; r = '
58 +str(testpoints[i])+', k = '+str(k[s])
59 +', E_0 = '+str(E_0[s])+', (original) E = '
60 +str(planewave_1[i])+', (reexpanded) E = '
61 +str(planewave_2_p)
62 +', x = '+str(np.dot(testpoints[i],k[s]))
63 +'; distance = '
64 +str(np.linalg.norm(planewave_1[i]-planewave_2_p))
65 +', required relative precision = '
66 +str(rtol)+'.')
67 failcounter += 1
68 else:
69 passcounter += 1
70 self.assertLess(failcounter / (failcounter + passcounter), rfailtol,
71 '%d / %d (%.2e) randomized numerical tests failed (tolerance %.2e)'
72 % (failcounter, failcounter + passcounter,
73 failcounter / (failcounter + passcounter), rfailtol))
74 return
76 def testCornerCases(self):
77 pass
80 class SphericalWaveTranslationTests(unittest.TestCase):
81 def testRandom1to1(self):
82 # The "maximum" argument of the Bessel's functions, i.e. maximum wave number times the distance,
83 # for the "locally strongly varying fields"
84 maxx = 10
85 rfailtol = 0.01 # how much of the randomized test fail proportion will be tolerated
86 lMax = 50 # To which order we decompose the waves
87 lMax_outgoing = 4 # To which order we try the outgoing waves
88 rtol = 1e-5 # relative required precision
89 atol = 1. # absolute tolerance, does not really play a role
90 nsamples = 4 # frequency samples per order of magnitude and test
91 npoints = 15 # points to evaluate per frequency and center
93 ncentres = 3 # number of spherical coordinate centres between which the translations are to be made
94 maxxd = 2000 # the center position standard deviation
96 failcounter = 0
97 passcounter = 0
98 my, ny = qpms.get_mn_y(lMax)
99 nelem_full = len(my)
100 nelem_out = lMax_outgoing * (lMax_outgoing + 2)
101 for oom in lengthOrdersOfMagnitude:
102 centres = oom * maxxd * np.random.randn(ncentres, 3)
103 ksizs = np.random.randn(nsamples)
104 for ksiz in ksizs:
105 for i in range(ncentres): # "source"
106 Rs = centres[i]
107 testr = oom * maxx * np.random.randn(npoints, 3)
108 for j in range(ncentres): # "destination"
109 if j == i:
110 continue
111 Rd = centres[j]
112 shift = Rd - Rs
113 shift_sph = qpms.cart2sph(shift)
114 shift_kr = ksiz * shift_sph[0]
115 shift_theta = shift_sph[1]
116 shift_phi = shift_sph[2]
118 A_yd_ys = np.empty((nelem_full,nelem_out), dtype = np.complex_)
119 B_yd_ys = np.empty((nelem_full,nelem_out), dtype = np.complex_)
120 for yd in range(nelem_full):
121 for ys in range(nelem_out):
122 A_yd_ys[yd, ys] = qpms.(my[yd],ny[yd],my[ys],ny[ys],shift_kr, shift_theta, shift_phi, True, 1)
123 B_yd_ys[yd, ys] = qpms.(my[yd],ny[yd],my[ys],ny[ys],shift_kr, shift_phi, shift_phi, True, 1)
124 for r in testr:
125 sph_ssys = qpms.cart2sph(r+Rd-Rs)
126 M_ssys, N_ssys = qpms.vswf_yr1(np.array([ksiz * sph_ssys[0], sph_ssys[1], sph_ssys[2]]), lMax_outgoing, J=1)
127 sph_dsys = qpms.cart2sph(r)
128 M_dsys, N_dsys = qpms.vswf_yr1(np.array([ksiz * sph_dsys[0], sph_dsys[1], sph_dsys[2]]), lMax, J=1)
129 for ys in range(nelem_out):
130 # Electrical waves
131 E_1 = -1j*qpms.sph_loccart2cart(N_ssys[ys], sph_ssys)
132 E_2 = -1j*qpms.sph_loccart2cart(np.dot(A_yd_ys[:,ys],N_dsys)+np.dot(B_yd_ys[:,ys],M_dsys),sph_dsys)
133 if not np.allclose(E_1, E_2, rtol=rtol, atol=atol):
134 failcounter += 1
135 else:
136 passcounter += 1
137 # Magnetic waves
138 E_1 = -1j*qpms.sph_loccart2cart(M_ssys[ys], sph_ssys)
139 E_2 = -1j*qpms.sph_loccart2cart(np.dot(A_yd_ys[:,ys],M_dsys)+np.dot(B_yd_ys[:,ys],N_dsys),sph_dsys)
140 if not np.allclose(E_1, E_2, rtol=rtol, atol=atol):
141 failcounter += 1
142 else:
143 passcounter += 1
144 self.assertLess(failcounter / (failcounter + passcounter), rfailtol,
145 '%d / %d (%.2e) randomized numerical tests failed (tolerance %.2e)'
146 % (failcounter, failcounter + passcounter,
147 failcounter / (failcounter + passcounter), rfailtol))
148 return
150 def testRandom3to1(self):
151 pass
153 def main():
154 unittest.main()
156 if __name__ == '__main__':
157 main()